et 7 05 - electric todayforums, on-line magazines, industry news, job postings, electrical store,...

PUBLICATION MAIL AGREEMENT # 40051146

Issue 7 Volume 17, 2005Issue 7 Volume 17, 2005


Electrical Buyer’s Guides,Forums, On-Line Magazines,Industry News, Job Postings,Electrical Store, Industry Links


www.electricityforum.com

S E P T E M B E R

2 7 t h - 2 9 t h

20 05

L o u i s v i l l e

K e n t u c k y U S A

World EnergyWorld EnergyA Success Story from Deregulation

ET_7_05 9/16/05 2:46 PM

POWER SYSTEMS

Today’s renewable energy

for tomorrow’s generation

Hubbell Canada LP, Power Systems 870 Brock Road South • Pickering, ON L1W 1Z8

Phone (905) 839-1138 • Fax: (905) 831-6353 www.HubbellPowerSystems.ca

ET_7_05 9/16/05 2:46 PM

3

6 Knowing When and Where to Choose Your Battles

8 New Orleans Tragedy Shuts Down IEEE PES Conference38 A Success Story from Deregulation: World Energy40 Serial Encrypting Transceivers Secure Vital Communications

to Protect Against Cyber Attacks51 Coal-based Fuel Cells: A Giant For Fuel Cell Technology

10 600V High Resistance Grounding for Commercial and Institutional Buildings- Why Not?

15 Advanced Condition-Based Assessment of Medium and High Voltage ElectricalSystems Without requiring an Outage

42 Power Cable Diagnostics Field Application and Case Studies

22 Integration of Protection, Control, Monitoring and Communication intoOne Intelligent Electronic Device for Small Hydro-Power Plants

28 Internal Insulation Failure Mechanisms of HV Equipment Under Service Conditions34 Switch Mode Power Supplies for Electrostatic Precipitators

41 NERC Prepares for FERC Rule on Electric Reliability Organization

52, 53

54

Volu

me

17,

Issue

7, 2

005

in this issueEDITORIAL

PROTECTION AND CONTROL

GROUNDING

TRANSFORMERS

INDUSTRY NEWS

CABLES

NEWS BRIEFS

ADVERTISERS INDEX

Electricity Today Magazine is published 9 times peryear by The Electricity Forum [a division of The HurstCommunications Group Inc.], the conference manage-ment and publishing company for North America’s elec-tric power and engineering industry. Distribution: free of charge to North American electri-cal industry personnel who fall within our BPA requestcirculation parameters. Paid subscriptions are availableto all others. Subscription Enquiries: all requests for subscriptionsor changes to free subscriptions (i.e. address changes)must be made in writing to:

Subscription Manager, Electricity Today215-1885 Clements Road,Pickering, Ontario, L1W 3V4

or on-line at www.electricityforum.com.

Canada Post - Canadian Publications Mail Product SalesAgreement 40051146 ISSN 0843-7343Printed in Canada. All rights reserved. The contents of this publication may not be reproducedin whole or in part without prior permission from thepublisher.

Member of:

President, The Electricity ForumRandolph W. [email protected]

PublisherLynda [email protected]

Associate PublisherCarol [email protected]

Advertising SalesCarol [email protected]

EditorDon [email protected]

Circulation ManagerColleen [email protected]

Visit our Web Site:www.electricityforum.comE-mail: [email protected]

Subscribe on-line:www.electricityforum.com/et/subscribe.htm

PRODUCTS AND SERVICES SHOWCASE

ET_7_05 9/16/05 2:46 PM

Electricity Today Issue 7, 2005 4

editorial board

BRUCE CAMPBELL

BOB FESMIRE

CHARLIE MACALUSO

SCOTT ROUSE

BRUCE CAMPBELL, LL.B., Independent Electricity Market Operator (IMO)Mr. Campbell is responsible for business development, regulatory

affairs, corporate relations and communications, and legal affairs at theIMO. He has extensive background within the industry and, in particular,acted as legal counsel in electricity planning, facility approval and rateproceedings throughout his career in private practice.

BOB FESMIRE, ABBBob Fesmire is a communications manager in ABB's Power

Technologies division. He writes regularly on a range of power industrytopics including T&D, IT systems, and policy issues. He is based in SantaClara, California.

CHARLIE MACALUSO, Electricity Distributor's AssociationMr. Macaluso has more than 20 years experience in the electricity

industry. As the CEO of the EDA, Mr. Macaluso spearheaded the reformof the EDA to meet the emerging competitive electricity marketplace, andpositioned the EDA as the voice of Ontario's local electricity distributors,the publicly and privately owned companies that safely and reliably deliv-er electricity to over four million Ontario homes, businesses, and publicinstitutions.

SCOTT ROUSE, CIPEC ChairmanScott Rouse is a strong advocate for proactive energy solutions. He

has achieved North American recognition for developing an energy effi-ciency program that won Canadian and US EPA Climate ProtectionAwards through practical and proven solutions. As a published author,Scott has been called to be a keynote speaker across the continent fornumerous organizations including the ACEEE, IEEE, EPRI, and CombustionCanada. Scott currently serves as Chair of the Canadian Industry Programfor Energy Conservation (CIPEC) - Energy Manager Network and is a pro-fessional engineer, holds an M.B.A. and is also a Certified EnergyManager.

DAVID W. MONCUR, P.ENG., David Moncur EngineeringDavid W. Moncur has 29 years of electrical maintenance experience

ranging from high voltage installations to CNC computer applications,and has conducted an analysis of more than 60,000 various electrical fail-ures involving all types and manner of equipment. Mr. Moncur has chaireda Canadian Standards Association committee and the EASA OntarioChapter CSA Liaison Committee, and is a Past President of the WindsorConstruction Association.

JOHN McDONALD, IEEE President-ElectMr. McDonald, P.E., is Senior Principal Consultant and Director of

Automation, Reliability and Asset Management for KEMA, Inc. He isPresident-Elect of the IEEE Power Engineering Society (PES), ImmediatePast Chair of the IEEE PES Substations Committee, and an IEEE Fellow.

DAVID W. MONCUR

JOHN McDONALD

ET_7_05 9/16/05 2:46 PM

ET_7_05 9/16/05 2:49 PM


The development of new generationis crucial to the viability of the grid, andnew generation (pardon the pun), cangenerate opposition.

The objections to nuclear are theloudest, especially from those whowould live closest to any proposed plantfearing that their back yards could turninto another Three Mile Island.

Construction of coal-burning plantshas all but been shelved thanks to the air-quality requirements of the K yotoAccord, despite impressi ve advances incleaner burning coal and scrubbing tech-niques.

Large-scale hydro-electric projectsare mostly a thing of the past, onlybecause the majority of potential hydro-electric sites near de veloped areas ha vealready been exploited.

Natural gas? Clean b urning yes, butthe potential for a terrorist attack on sup-ply lines does exist.

Really, the v ery few generationoptions that seem palatable to e veryoneseem to be solar and wind.

Or so we thought.On radio stations in Madison and

Fond du Lac Wisconsin, a group calledthe Advocates for the Horicon Marshwildlife refuge has been running spots inthe hopes of stopping construction ofwind turbines for that state’s largest windfarm, asking Governor Jim Doyle to per-sonally intervene.

Timed to coincide with the go ver-nor’s announcement of an environmentaland energy policy program - a programthat includes a greater state investment inrenewable energy (that includes windpower) - the radio spots e xpress concernthat development of the Forward EnergyCenter will result in the destruction ofbirds at the marsh.

Invenergy is the Chicago-basedcompany that is developing the project -a project that w ould generate four timesthe electricity of all of Wisconsin’s windturbines combined. The wind farm aimsto sell to Wisconsin Public Po wer Inc.;Madison Gas & Electric Co.; Wisconsin

Power & Light Co.; and WisconsinPublic Service Corp.

Paul Baicich of the National WildlifeRefuge Association has stated that, “Ourconcern is that you can be for rene wableenergy and also be responsible. To put(the wind turbines) this close is notresponsible.”

Along with the Horicon MarshSystem Advocates, other groups lik e theWisconsin Audubon Council and theNational Wildlife Refuge Associationwant the wind turbines to be built at leastfour miles from the marsh.

Huh?Admittedly I am not an e xpert on

ornithology, but surely the winged crea-tures that inhabit and use the marsh canavoid the slow turning blades of a windturbine. And the current positioning ofthe turbines still pro vides for a b ufferfrom the marsh.

Wisconsin’s Public ServiceCommission has so far approved placingturbines two miles from the edge of themarsh - a buffer based on the recommen-dations of the U.S. Fish and WildlifeService, which manages the marsh, andWisconsin’s Department of NaturalResources.

If you b uy into what the objectorsare saying, it seems that no w even theblades of wind turbines need the samedistance that nuclear fuel rods and carbondioxide belching smok estacks do fromthe homes of living creatures.

It should be noted that the U.S. Fishand Wildlife Service originally sought abuffer of 3 to 4 miles from the marsh b utlater endorsed the minimum b uffer of 2miles. But it should also be pointed outthat they called for more studies to beconducted to document the impact theturbines have on wildlife. Translation:They aren’t certain the turbines have anyimpact whatsoever.

Indeed, previous studies of similarwind turbine sites sho w that on a veragethere are two bird deaths per turbine peryear. Although the deaths are re grettable(and at the risk of being insensiti ve to

avian lovers everywhere), they seemacceptable.

If I really w anted to ruf fle somefeathers, I could point to the ratio of birddeaths to aircraft propeller blades (bothjet turbine and turboprop). I can safelysay that those numbers are considerablyhigher, and no one is eager to put a halt toair travel any time soon.

Ironically, the fiercest debate may beshaping up between en vironmentalgroups that adv ocate for clean ener gyagainst groups concerned about wind tur-bines’ impact on birds and bats.

According to the Wisconsin PublicInterest Group, the en vironmental bene-fits of wind power should be obvious.

A spokesperson for the group saysthat “if you take a big-picture view of ourenergy choices, if you care about wildlifeand the environment, global warming hasto be the top issue, and that is the biggestthreat.”

Global warming should be a majorconcern especially in Wisconsin, wheremore than 70 per cent of that state’s ener-gy is deri ved from coal (with anotherfour or f ive coal-fired generation plantson the drawing board).

Although global w arming concernsshould be balanced with future genera-tion concerns, the cries from the adv o-cates for the marsh can be lik ened to theludicrous efforts of PETA (People for theEthical Treatment of Animals).

PETA - an or ganization that be ganwith the good intentions of protectingand bettering the li ves of animals - hasbecome so left-wing that even some veg-etarians have begun to question theirmotives and distance themselv es fromthe group.

So too could the actions of theAdvocates for the Horicon Marshwildlife refuge. The downfall of creatinga stir o ver wind f arms harming thewildlife is that when the y raise theirvoice over more serious threats to themarsh, they lose credibility - akin to thestory of the boy who cried wolf. ET

EDITORIAL

KNOWING WHEN AND WHERETO CHOOSE YOUR BATTLES

By Don Horne

ET_7_05 9/16/05 3:01 PM

Do you want better overall economy, as well as longer lifetime for your high voltage switchgears?

Then take a look at an oil that has been tailored for the job. Nyswitcho has a low viscosity and

pour point, even at low temperatures. And it offers great oxidation stability, even for those switches

you use the most. Don’t settle for anything less than the real thing; choose a proper switchgear oil

for your high voltage switches.\\ Expect more

There are many oils out there.But few are tailored for switchgears.

Ehre

nst

råh

le B

BD

O

Nynas Canada, inc. Telephone: (905) 804-8540. E-mail: [email protected] www.nynas.com

ET_7_05 9/16/05 3:01 PM


A message from the President of IEEE PES, Hans B. Püttgen:

Dear Friends,As the reports of the de vastation in New Orleans come in,

our first concerns have been for our man y friends in the areaand, in particular, for the members of the local organizing com-mittee of the IEEE PES Transmission and Distrib utionConference and Exposition. It has now become evident that therecovery process will be prolonged and will require a broadrange of resources, including those of our industry . My col-leagues from the IEEE Po wer Engineering Society Go verningBoard and I have come to the conclusion that the most reason-able course of action, under these dramatic circumstances, is topostpone the event until next spring. We feel that this decisionis in the best interest of the industry we serv e.

The decision has been reached to postpone the e vent untilspring.

If at all possible, we would like to support the great City ofNew Orleans and its w onderful team of volunteers by holdingthe event there next spring.

The venue and dates will be announced within the next fewweeks. We are well aware of the conflicts with other events willendeavor to minimize any inconveniences to the exhibitors andparticipants. Please continue to watch this site for details.

The exact details as to the transition to the ne w dates arestill being worked out. In particular:

- All contracted exhibiting companies will be contactedshortly by

the show management organization to discuss the tran-sition to the

new dates. - We will contact those of you who ha ve already regis-

tered as an attendee with information as to the transition of your

registration.- We will shortly ha ve details regarding your housing

arrangements if you had already made reservations.

Please check with your airline for conditions re gardingrefunds for your New Orleans travel arrangements.

In closing, I would like to ask that you keep our friends inthe devastated areas in your thoughts as their ener gies arefocused to restore their living conditions. Your support is vitalin these difficult times.

We are counting on all of you to attend the e vent nextspring. I am comforted by the strong and broad support wehave received from the manuf acturers during the decisionprocess; I am confident that we can count on their full support

of as exhibitors next spring.Best regards,Hans B. (Teddy) PüttgenPresident, IEEE Power Engineering Society

Editor’s note:Our thoughts and best wishes here at Electricity Today go

out to the people of Ne w Orleans who are enduring unspeak-able hardships in the wake of Hurricane Katrina.

We are confident that the city will bounce back from thistragedy with the resolute pride that e veryone in Louisiana hasshown throughout the past few weeks, and look forward to par-ticipating at next year’s IEEE Conference - hopefully down onthe bayou.

INDUSTRY NEWS

NEW ORLEANS TRAGEDY SHUTS DOWNIEEE PES CONFERENCE

Satellite photos before and after the hurricane.

ET_7_05 9/16/05 3:05 PM

Bell Wireless Data Specialists are helping utilities inspectors work more efficiently.We're setting them up with solutions like the rugged Symbol® handheld runningon our fast, dependable 1X network. Inspectors can submit data once, from thefield, and go directly to the next site instead of going back to the office.

Certain conditions and restrictions apply. Service, features and functionality available in Bell Mobilitycoverage areas where technology exists. SYMBOL is a registered trademark of SYMBOL technologies.

Now there’s a way for inspectors to

SPEND MORETIME ON SITE

and less time in the office. Our Wireless Data Solutions

give them the tools to access up-to-date inventory information

and submit reports from almost anywhere.

To learn more about how we can help your business, call 1 866 838-4107 or visit www.businessonthego.ca/utilities

ET_7_05 9/16/05 3:05 PM


Six-hundred volt and 480V electri-cal power distribution systems withinmost commercial and institutional b uild-ings are solidly grounded by tradition.

In solidly grounded systems, groundfaults can cause do wntime, equipmentdamage, and in se vere cases, arc flashinjury to personnel. Man y industrialfacilities choose high resistance ground-ing for their 600V distrib ution systemsbecause they experience more frequentground faults than commercial and insti-tutional buildings. A high resistancegrounded distribution system respondsmuch more safely to ground f aults, pre-venting downtime, damage, and injury.

Why do commercial and institution-al buildings continue to ha ve solidlygrounded 600V systems when high resis-tance grounding of fers superior protec-tion against ground faults?

• Habit and tradition• A presumption that solid

grounding is safest• 347V lighting requires a 4-wire

solidly grounded source• A preference for a ground f ault

to trip a breaker and de-energized imme-diately

• No maintenance electricians arenot available to locate ground faults

Let’s examine the principal problemfaced in the design of solidly groundedsystems – arcing ground f aults. Highcapacity power systems deli ver tremen-dous energy when electrical faults occur.Seventy to 85 per cent of electrical faultsbegin when an ener gized phase conduc-tor contacts ground, typically through aninsulation failure or human error(dropped tool while w orking on li veequipment). These are known as groundfaults. When the resistance at the pointof the fault is practically zero, a “bolted”ground fault occurs, causing v ery highcurrent to flow until interrupted by a cir-cuit breaker or fuse. Such f aults usuallycause no damage because the y are inter-rupted instantaneously, within 2 – 50 mil-

liseconds (3 cycles or less).Unfortunately, ground faults caused

by insulation f ailures (typically due tomechanical damage or aging and wear)are not bolted f aults; there is an air gapbetween the ener gized conductor andground. In 600V and 480V po wer sys-tems, there is enough v oltage to create asustained electrical arc across the air gap.Called “low level arcing ground f aults”because the arc resistance reduces cur-rent flow, they are not quickly detectedby circuit breakers or fuses. Without spe-cial ground fault protection relays, thesearcs can escalate into phase-to-phase andthree-phase faults, releasing tremendousheat and mechanical ener gy [1].Equipment is damaged and people can beinjured or killed if e xposed to thesefaults. Since the early 1970s both theCanadian Electrical Code and NationalElectrical Code ha ve required groundfault protection on the main break er on600 V or 480 V solidly grounded systemswhere the rated current is 1000 A orhigher. The maximum ground fault pick-up setting allo wed is 1200A, and themaximum time delay allo wed is 1 sec-ond. This measure limits, b ut does notentirely prevent, equipment damage ordowntime – arcing f aults can do muchdamage within one second. To avoid trip-ping the main breaker each time a groundfault occurs, ground fault relays are oftenspecified for branch feeders as well.

Low level arcing ground faults occurmore frequently in industrial f acilitiesthan in commercial or institutional build-ings because there is f ar more electricalequipment, particularly motors within arelatively confined space. In some indus-tries, the en vironments are electricallyhostile with moisture, dust, contaminantsor excessive temperatures. Continuousprocess industries e xperienced intolera-ble equipment damage and do wntimedue to arcing ground f aults in the 1950sand 60s, after they started solidly ground-ing their 600V and 480V distrib ution

systems. Up to the mid-1950s, most low-voltage industrial distrib ution systems(600V and 480V) were ungroundedthree-wire systems, in which high energyarcing ground faults were rare. However,ungrounded systems were sho wn tocause high transient o vervoltages undercertain failure conditions, leading to mul-tiple simultaneous motor f ailures [2].Hence 3-wire 600V and 480V systemsbecame solidly grounded in the 1950s tocure this o vervoltage problem. Solidgrounding solved one problem, but creat-ed another.

The probability that phase-to-groundfaults on 600V and 480V systems willbecome sustained and dangerous arcingfaults is high. On 208V systems, sus-tained arcing faults are rare. A safety haz-ard exists for solidly grounded 600V and480V systems from the se vere flash, arcburning, and blast hazard from a phase-to-ground fault [3]. The phase-to-groundvoltage in 208V systems is not highenough to sustain an arc; in mostinstances the arc self extinguishes in a ?-cycle.

Continuous process industries withhostile electrical en vironments such aspetrochemical, pulp & paper , and steelbegan high resistance grounding their600V and 480V distrib ution systems inthe 1970s to pre vent low level arcingground faults. As a result the y experi-enced fewer arc flash injuries, reducedequipment damage and less do wntimefrom ground faults.

Although hospitals and data centresexperience fewer ground f aults thanindustrial plants, when the y do occur ,downtime is just as intolerable whenpatient care or ATM networks are affect-ed. High resistance grounding eliminatesthis risk from 600V systems.

High resistance grounding is appro-priate for enterprises where:

• downtime due to ground f aults

GROUNDING

600V HIGH RESISTANCE GROUNDING FORCOMMERCIAL AND INSTITUTIONAL BUILDINGS -

WHY NOT?By David Murray, P.Eng., IPC Resistors Inc.

Continued on Page 12

ET_7_05 9/16/05 3:05 PM

ET_7_05 9/16/05 3:09 PM

1-905-686-10401-905-686-1078www.electricityforum.com/forums/reg.htm215-1885 Clements Road Pickering, Ontario L1W 3V4, CANADA

phone: fax:

online:mail:

4easy waysto register

it’s time to go back to school!

don't miss the bus on our new electrical forums

COMING THIS FALL

Plus!BONUS: $100 ELECTRICITY FORUM “POINTS” COUPON • 1.4 CONTINUING EDUCATION UNIT(CEU) CREDITS • FREE ELECTRICITY TODAY MAGAZINE SUBSCRIPTION • PRESENTATIONSIN PAPER FORMAT • CD-ROM CONTAINING PRESENTATIONS • ELECTRICAL HANDBOOK

Electrical Safety In The Workplace Forums

Industrial/CommercialElectrical Grounding

Forums

Industrial ElectricityEfficiency Forum

Electrical Safety In TheWorkplace Forum

Saint John, NB • September 22-23, 2005Hoilday Inn Express St. JohnHalifax, NS • September 26-27, 2005Citadel Halifax HotelSt. John's, NL • September 29-30, 2005Best Western Travellers Inn

These Cross-Canada Forums will focus on recent coderevisions from NEC and NFPA regarding electrical safe-ty, specifically arc flash hazards. As well, electricalinspectors from each province will make special presen-tations about electrical safety codes and standards.Also, various Flame Retardant clothing manufacturerswill discuss the wide range of options available when itcomes to protective clothing and safety equipment.

The registration fee to attend the forum is $649.00 +$45.43 GST ($649 + 97.35 HST for NB, NS, and NL resi-dents). REGISTER AND PREPAY 8 DAYS BEFORE FORUMAND RECEIVE AN EARLY BIRD REGISTRATION FEE OF$599.00 + $41.93 GST ($599 + 89.85 HST for NB, NS,and NL residents) per delegate.

www.electricityforum.com/forums/electrical-safety-east-2005.html

Electrical Grounding ForIndustrial, Commercial andInstitutional Power SystemsToronto, ON • October 5-6, 2005Park Plaza Toronto Airport HotelEdmonton, AB • October 17-18, 2005Cost Terrace InnVancouver, BC • October 19-20, 2005Holiday Inn Vancouver Centre

It has been determined that over 70% of all problems inindustrial, commercial and institutional power systemsare due to poor grounding and wiring errors. Withoutproper grounding, sensitive electronic equipment is sub-jected to destruction of data, erratic equipment opera-tion, and catastrophic damage. This course will give par-ticipants a comprehensive understanding of practicalapplications of proper grounding and wiring practices.

The registration fee to attend the course is $699.00 +48.93 GST. REGISTER AND PREPAY 10 DAYS BEFORETHE COURSE DATE AND RECEIVE AN EARLY BIRD REG-ISTRATION FEE OF $649.00 + 45.43 GST PER DELE-GATE.

www.electricityforum.com/forums/electrical-grounding-2005.html

Industrial Electricity EfficiencyForum

Toronto, ON • October 17-18, 2005Prk Plaza Toronto Airport Hotel

This two-day forum will introduce delegates to the com-petitive value of metering, monitoring and manage-ment. It will discuss what is involved in developing anelectricity monitoring and management program andprovide insight into the associated costs and benefits.

Our Industrial Electricity Efficiency Forum will alsoprovide background information on current market con-ditions in Ontario, including federal and provincial gov-ernment initiatives for promoting energy managementand energy conservation and electricity efficiency.

Ontario’s electricity market continues to provide sig-nificant barriers in taking control of escalating energycosts. The inability to forecast electricity price spikes isjust one example. The resulting impact on budgets, costcentres, potential rebates, etc all need to be betterunderstood, communicated and solutions implemented.

The registration fee to attend the course is $699.00 +48.93 GST. REGISTER AND PREPAY 10 DAYS BEFORETHE COURSE DATE AND RECEIVE AN EARLY BIRD REG-ISTRATION FEE OF $649.00 + 45.43 GST PER DELE-GATE.

Register 3 delegates at the regular price and receive a 4th registration FREE.


is intolerable;• ground faults are located and

repaired once they occur, i.e. they are notignored;

• 347V or 277V lighting is nomore than 10 per cent of the total electri-cal load.

Continuous process industries ha vesuccessfully used high resistance ground-ing for over 30 years increase the a vail-ability of electrical po wer from their600V or 480V distrib ution systems.Data/telecom centres are now using highresistance grounding to squeeze e venmore uptime from their backup genera-tors and UPS systems.

Today, 600V and 480V systems aresolidly grounded for one of tw o reasons:(1) either to pro vide a grounded neutralconductor for 347V or 277V lightingloads, and/or (2) because the solidlygrounded system is still percei ved to bethe safest method of grounding.

347V LIGHTINGContrary to popular belief, 347V or

277V lighting does not require the entire600V or 480V distrib ution system to besolidly grounded, particularly when the347V lighting load is less than 10 percent of the total electrical load. Smalldelta-wye 600V-600/347V isolationtransformers can be locally distributed tosupply 600/347V lighting panels. Thispractice (1) impro ves the safety of the600V system in the rest of the b uildingby allowing high resistance grounding,(2) saves the cost of distributing a neutralconductor throughout a f acility, and (3)provides the added benefit of much lowerfault current at the 600V lighting panels,for better coordination.

Three-hundred-and-forty-seven volt,1-pole, 15A and 20A lighting break ersare available with interrupting ratings of14-kAIC, 18-kAIC, and 25-kAIC only .Lighting panels connected to 600V sys-tems with higher than 25kA of a vailablefault current cannot be fully rated, b utmust use series-rated main and branchbreakers. Under a bolted ground f ault,the series-rated main and branch breakersmust trip simultaneously, to dissipate thefault energy across both breakers. In thisway a faulty lighting ballast can trip the

lighting panel mainbreaker and plunge allof the other branchlighting circuits intodarkness. By insert-ing a small isolationtransformer to supplythe lighting panel, theavailable fault currentdrops to 2kA or less;a faulty ballast tripsthe branch break eronly, and a lessexpensive 14kAIClighting panel may beused. (See figure 1)

It is not necessary to solidly groundan entire 600V distribution system just tosupply 347V lighting. By limiting thesolidly grounded 600V systems to smallislands of lighting circuits, the bulk of the600V distribution system can be highresistance grounded to impro ve person-nel safety, limit do wntime, and reduceequipment damage from arcing groundfaults.

Also, ground f aults cannot tra velthrough a delta winding. Ground f aultson a wye secondary winding are convert-ed to phase overcurrents only on the pri-mary circuit when the primary winding isdelta configured (the norm). Hence aground fault on the secondary cannot tripa ground fault relay on the primary , forbetter coordination.

Are 600V solidly grounded systemsreally safer? Solidly grounded systemsare often percei ved to be safest on themistaken premise that all ground f aultswill quickly trip the affected feeder fuses

or circuit breaker. Unfortunately, no mat-ter how low the impedance of the groundreturn path is back to the neutral of thetransformer, a high impedance at thepoint of f ault itself, such as occurs forarcing ground faults, is enough to reducethe fault current such that the feederbreaker or fuses will not trip without sen-sitive ground fault relays. (See Figure 2)Only bolted ground faults are guaranteedto trip the feeder breaker or fuses withoutsensitive ground fault relays.

Contrary to popular belief, mostground faults are not bolted f aults andcannot be depended upon to reliably tripthe feeder o vercurrent protective deviceeach time, unless the feeders ha ve sensi-tive ground f ault relays. Often only themain breaker or fusible disconnectswitch has ground fault protection, set totrip at a relati vely high le vel of groundfault (1200A) after a time delay of onesecond to avoid nuisance trips. A lot of

GROUNDING

Figure 1 – Improved Co-ordination with Distributed 600V Lighting Transformers

Figure 2 – High Impedance at Point of Fault Fails to Trip Breaker


HIGH RESISTANCEFrom page 10

ET_7_05 9/16/05 3:09 PM

arcing damage can occur at the point ofbranch fault in one second.

To achieve time-current co-ordina-tion between GFP relays, the upstreamrelays must have longer time delays thandownstream relays. The minimum timeco-ordination interval between GFPrelays 0.2s. F or example three GFPrelays in series may ha ve time delays of100ms, 300ms and 500ms for the branch,feeder and main de vices, respectively.Yet the clearing time of arcing f aultsshould ideally be less than 12 c ycles or200ms [3]. Unacceptable damage willstill occur for arcing ground f aults locat-ed upstream in the distrib ution system,where GFP time delay e xceeds 200ms.Time coordination involves the trade-offbetween tolerating more damageupstream and achieving better co-ordina-tion.

Co-ordinated GFP relays and branchphase overcurrent devices reduce arcing

fault damage, but still do not satisfy therigid requirements of personnel arc flashhazard protection [4]. This fact makessolidly grounded 600V or 480V systemsactually less safe than high resistancegrounded systems.

NFPA 70E is a US standard for elec-trical safety in the workplace. It address-es the arc flash hazard pre valent in elec-trical distribution systems. This standard,or similar v ersion of it, will come toCanada in the future. In order to reducethe arc flash hazard, ground f ault break-ers will have to be set with a shorter timedelay. This will reduce coordination, b utincrease safety in the presence of groundfaults.

High resistance grounding prevents afeeder from tripping on a ground fault bylimiting the ground fault current to 5A. Italso prevents an arc flash upon a groundfault. There is less disruption andincreased safety. (See Figure 3)

The neutral point of the LV system isgrounded through a resistor which limitsground fault current to a relati vely lowvalue of 5A or less. When a bolted

ground fault occurs, the line-to-neutralvoltage of the f aulted phase is fullyimpressed across the neutral groundingresistor, injecting a controlled currentinto the fault, in this case 347V÷69 Ω =5A. At such low fault current, the affect-ed feeder need not be de-ener gized untila more convenient time; hence a former-ly un-scheduled outage on solidlygrounded systems is con verted to ascheduled one on high resistance ground-ed systems.

During a ground f ault on a highresistance grounded system, the tw ounfaulted phases rise in v oltage aboveground by 73 per cent to 600V . SeeFigure 4. The neutral point between thethree phases rises to 347V abo ve groundduring a bolted ground fault. If this werea wye winding, the neutral terminalwould become energized to this v oltageduring ground faults. It is for this reasonthat the neutral of ungrounded or highresistance grounded systems cannot bedistributed throughout a facility and used

13Electricity Today Issue 7, 2005



ET_7_05 9/16/05 3:19 PM


GROUNDING

to supply line-to-neu-tral loads – because itbecomes energizedduring ground f aults.To supply 347/600V 4-wire loads, small local-ly distributed isolationtransformers are used,as explained above.

This voltage riseon the tw o unfaultedphases while a ground f ault is presencedoes not hurt the insulation of cables,transformers or motors. These insulationsystems are designed to withstand thisvoltage with no problem.

Two items that must not be used onhigh resistance grounded systems: (1)347/600V “slash-rated” lighting break-ers, and (2) 347V/600V wye-rated tran-sient voltage surge suppressors (TVSS).Lighting breakers are designed to be usedonly on solidly grounded systems, b utthen so is 347V lighting. Only “600V”rated breakers (no “slash”, as in“347/600V”) may be used on high resis-tance grounded systems. Ho wever mostpower panelboards and all switchboardwith molded case circuit break ers rou-tinely use 600V-rated breakers anyway.

TVSS must be 600V “delta-rated”,which means that they can withstand theincreased voltage of the tw o unfaultedphases to 600V. These TVSS are readilyavailable from TVSS manufacturers.

It is a common misconception that ahigh resistance grounded system is notreally a grounded system. Nothing couldbe further from the truth. First, thegrounding and bonding of all non-currentcarrying exposed metal parts (cable trays,conduits, enclosures, racks, etc.) isabsolutely unaffected by the method ofsystem grounding. All such componentsare still bonded directly to ground in ahigh resistance grounded system.Second, the neutral is grounded, e xceptthrough a resistor instead of through asolid conductor. The IEEE Green Booktreats high resistance grounded systemsas grounded systems:

IEEE 142-1991 (GREEN BOOK)Recommended Practice for

Grounding of Industrial and Commercial

Power Systems1.4.1Grounded systems emplo y some

method of grounding the system neu-tral.... These methods can be divided intotwo general categories: Solid groundingand impedance grounding.

1.4.2Numerous advantages are attributed

to grounded systems, including greatersafety, freedom from e xcessive systemover-voltages that can occur onungrounded systems during arcing… andeasier detection and location of groundfaults when they do occur.

When a ground f ault occurs on ahigh resistance grounded system, you fixit just as you would on a solidly ground-ed system. The difference is that the faultwill not ha ve caused an y damage orunscheduled downtime. What’s not tolike?

Many data centres and hospitals nowuse high resistance grounding on their600V distribution systems for increasedservice continuity and safety in the pres-

ence of ground f aults. As the trend forincreased electrical safety in the w ork-place continues, more commercial andinstitutional buildings will begin to real-ize the benef its of high resistancegrounding that industrial plants disco v-ered over thirty years ago. ET

BIBLIOGRAPHY[1] J. R. Dunki-Jacobs, “The

Escalating Arcing Ground-FaultPhenomenon,” IEEE Transactions onIndustry Applications, Vol. IA-22, No. 6,Nov/Dec 1986, pp. 1156-1161.

[2] IEEE Std 141-1993,Recommended Practice for ElectricPower Distribution for Industrial Plants,paragraph 7.2.4, p. 368.

[3] J.R. Dunki-Jacobs, “The Effectsof Arcing Ground Faults on Low-VoltageSystem Design,” IEEE Transactions onIndustry Applications, Vol. IA-8, No. 3,May/June 1972, pp. 223-230.

Figure 3 – High Resistance Grounded Power System


Figure 4 - Voltage Rise of the Neutral and Two Unfaulted Phases

ET_7_05 9/16/05 3:23 PM

INTRODUCTIONPartial discharge (PD) is becoming increasingly vie wed as

the best diagnostic for insulation, particularly for on-line mea-surements. Clearly this applies to insulation, which e xhibits andis degraded by PD activity. However, even for insulation, whichis designed to be PD free, the kno wledge that the system is actu-ally PD free is still a vital part of the diagnostic process. HencePD measurements, which are accurate and reliable, al ways con-tain important information about the high v oltage (HV) equip-ment to which they apply.

With the development of on-line PD methods, se veral prob-lems still exist to be solv ed to yield the optimum diagnostic forinsulation. These are:

• Noise interference from RF radio broadcasts• Noise interference from pulses other than PD pulses


CABLES

ADVANCED CONDITION-BASED ASSESSMENTOF MEDIUM AND HIGH VOLTAGE ELECTRICALSYSTEMS WITHOUT REQUIRING AN OUTAGE

By R.R. Mackinlay, High Voltage Solutions Ltd., and Don Genutis, Hampton Tedder Technical Services

Fig 1 Pulse from a partial discharge in a cable (showingcomputer generated cursors)


ET_7_05 9/16/05 3:32 PM

17th Annual Canadian Power Conference & Trade ShowNovember 21-22, 2005 • Metro Toronto Convention Centre, South Building, Toronto, Ontario • 416-322-6549

200 MWRenewables 3 RFP200 MW

200 MWNiagara Tunnel200 MW

310 MWThunder Bay

Gas Fired Generation

310 MW

250 MWDemand Side Management

& Response

250 MW

500 MWDowntown Toronto500 MW

395 MWRenewables 1 RFP395 MW

2,235 MWClean Energy &

Demand Side Projects

2,235 MW1,000 MW

West GTA1,000 MW

1,000 MWRenewables 2 RFP1,000 MW

1,000 MWCo-generation

1,000 MW

515 MWPickering A Unit 1515 MW

1,540 MWBruce A 1 and 2

1,540 MW

www.appro.orgwww.appro.orgT O R E G I S T E R

ET_7_05 9/16/05 3:34 PM


• Location of PD pulses on cablesand static equipment (e.g. switchgear ,transformers etc)

• Estimation of remaining servicelife for the PD acti vity measured in theHV equipment under test

The remaining service life of equip-ment exhibiting PD acti vity is brieflydealt with for restricted types of equip-ment, otherwise this is left for a futurepaper. This paper addresses the mainissue of making reliable PD measure-ments, namely the identif ication andremoval of noise sources, and the loca-tion of PD e vents. To be able to do thisallows the identif ication of dischar gingequipment, and hence the assessment ofthe risk to reliable service performance.

This paper describes the de velop-ment of methods to distinguish betweennoise and PD activity, and to distinguishbetween several different types of PD.The recognition is done using simplealgorithms, which use the w aveformshape of the PD (and noise) pulses, todiscriminate between the different types.

Partial discharges (PD) in voids andcavities in general produce v ery similarpulse shapes with very fast pulse widths.Values of a fe w tens or hundreds ofpicoseconds are typical. The pulse timesare closely related to the breakdown timeacross the cavity. However, at the termi-nals of the high v oltage equipment (orusing an electromagnetic pickup if theseare used) the pulse shape is dependant onthe parameters of the detection circuit.Often these are determined by the equiv-alent circuit of the po wer equipment atthe detection point. The observer may nothave much control over the detection cir-cuit in these cases. Hence it might appearthat using the w ave shapes of the pulsesto distinguish between noise and PD (andbetween different types of PD) does nothave a bright future.

However, there are tw o classes ofPD pulses, which do retain a reasonableconsistency of beha viour, for both on-line and off-line testing. These are cablePD’s and PD’ s from local equipment(switchgear etc). In this case, local equip-ment means equipment within a fe wmetres of the measurement point. Thealgorithm used in the de velopmentdescribed in this paper, essentially makesthree categories of pulsed signals. Theseare cable PD’ s, local equipment PD’ s,

and all others are regardedas noise. This simplealgorithm gives remark-ably good results in awide variety of circum-stances, and has now beensuccessfully incorporatedinto several commerciallyavailable products.

PD EVENTS FROMHIGH VOLTAGE CABLES

In the special case ofpartial discharge incables, the cavity responsible for the PDdischarges into a real impedance. This isthe surge impedance of the cable and ispurely resistive at the point of launch.The resulting pulse is virtually monopo-lar (see Ref 1) with f ast rise times andvery short pulse width. This pulse travelsoutward from the originating site, andarrives at the detection point wider andsmaller due to attenuation and dispersionon its travel down the cable. The pulse isin general well def ined by the time it isdetected, and in practice retains much ofthe same characteristics as a consequenceof originating in the cable. Figure 1shows a typical cable pulse, with com-puter-generated cursors to measure therise times and pulse properties.

Providing the rise times and widthsfit into sensible v alues for a PD pulse,then the event is cate gorized as a cableevent. The rise times are typicallybetween 50nSec and 1?Sec, with pulsewidths in general being less than 2?Sec.Actually for XLPE cables these v aluesare generally shorter than this, as the lossand dispersion are considerably less forXLPE cables. The pulse rise time andwidth are dependent on both the pulseshape at the cable end, and the detectioncircuit. Hence the neat method of usingthe risetime (or pulse width) as a measureof the location of the pulse cannot be rig-orously used, as the detection circuit isnot known, and for example, may containa large inductance and al ways then giveslow rise times and longer pulses.However, the rise time isnevertheless a valuable toolin the initial localization ofthe pulses, as for on-linePD detection using highfrequency current trans-formers (CT’s) the detec-tion circuits generally havelarge bandwidth (>20MHz)and simple localizationworks reasonably well.

Figure 2 sho ws the current trans-former arrangements for a 33kV XLPEcable, where the CT’ s can be placedaround each of the single cores, abo vethe ground strap take off point.

Normally the current transformersare placed around the ground strap. ThePD pulses traveling down the cable to thetermination, have an equal and oppositepolarity on the conductor and screenrespectively. Hence it does not matterwhether the CT’s are placed in the earthstrap, or the conductor . The importantcriterion is that only one of the ground orconductor currents is intercepted. Inpractice the two signals are very similar,but often the noise content can v arybetween the two detection points.

Figure 3 shows a cable with a spreadof rise time events, where there are clearbands to the rise times recorded. In prin-ciple, a graph could be constructed fromfigure 3, with the ordinate calibrated inmetres. This would assume a relationshipbetween the rise time and the distancetravelled for PD pulses in cables. This isactually reasonably well established(depending on the cable type), b ut themain problem is the unkno wn detectionimpedance at the termination. For exam-ple, if the detecting impedance containeda large inductance, then the rise times ofthe pulses w ould be dominated by thedetection impedance, and not the dis-tance travelled by the PD pulses.


ASSESSMENTFrom page 15

Fig 2 High frequency current transformers (CT’s) aroundcores of 33kV cable for PD measurement (CT’s arearrowed)

Fig 3 Distribution of pulse rise times from a cable

ET_7_05 9/16/05 3:34 PM


In cases lik e this, a relationshipbetween rise time and distance cannot bemade. The response of the detection cir-cuit could be measured, and if possible,the relationship could be correctly estab-lished. This will be the subject of a futurepaper.

The results in f igure 3 were tak enfrom a 33kV paper insulated cable testedon-line. Notice that the risetime of thePD events for this cable are around200nSec. The cable was around 2km inlength, and the locations were all from asingle splice located at 1600m.

The huge advantage of carrying outPD detection using the waveforms of thePD pulses is that it allo ws for correctionof the magnitude of the discharge almostirrespective of the distance do wn thecable the pulse has travelled. Particularlyfor paper-insulated cables where theattenuation is lar ge, making magnitudeonly measurements of PD pulses leads topoor conclusions about the peak magni-tudes. A PD of the same size can easilylook a f actor of 10 or 20 smaller whenseen at some distance do wn the cable.Hence a very severe event remote to themeasurement site can be observ ed asquite a modest PD if the attenuation hasreduced the peak magnitude by a f actorof say 20 or so. Using the w aveform tomeasure the area under the curv e of thePD current, a measure of magnitude canbe made which is much less sensiti ve toattenuation. The measurement of char geis made by simply integrating the chargeunder the current curve:

where ‘const’ is the multiplier whichconverts the current to v oltage. This willinclude the transfer impedance of thecurrent transformer and such f actors asthe cable impedance and amplif ier gain.Hence using this method for calculatingcharge, the values of PD magnitudes canbe measured in Pico Coulombs, providedthat the detection impedance may beassumed to be the cable sur ge imped-ance, or a correction factor used. In prac-tice the variation in the change of sur geimpedance at the ends tends to be lessthan 20% or so, when compared to mea-surements made in the ground straps of

splices in the middle of the cable, wherethe surge impedance will be close to thesurge impedance.

As an e xample of measuring thecharge in this w ay, at 3km on an 11kVpaper insulated mass impre gnated non-draining (MIND) cable, the area measurehad only reduced by a factor 3.

However, the amplitude measure-ment had reduced by factor of 15.

This means that for on-line PD tests,without the need for an y calibration, thecable events can all be expressed in picocoulombs.

PRECISE LOCALIZATION OF PDEVENTS ON CABLES USING A

TRANSPONDERIn the special case of testing high

voltage cables, the usefulness of measur-ing PD acti vity is v astly increased if alocalization of the PD origin can bemade. For short cables, a double-endedmethod of timing the arri val of the PDpulses is the most effective. This requiresa sensor at both ends of the cable, and acalibration of the cable tra vel time forboth legs. For longer cables, getting aradio frequency cable to the f ar end sen-sor is not usually possible. Several meth-ods have been attempted to o vercomethis problem. GPS sensors can be used totie two PD recorders to the same absolutetime. This method takes a lot of care issetting up, and requires a great deal ofmarrying up signals after the data isrecorded.

Alternatively, the timing informationfrom the remote end of the cable can bedirected back down the cable cores them-selves. A simple method has been adopt-ed to achieve this. A transponder consistsof a detector and trigger le vel for PDpulses, linked to a lar ge voltage pulsegenerator, which launches pulses backonto the cable under test. Figure 4 sho wa photograph of the two battery operateddevices. The trigger levels for the inputdevice spans the range 1mV – 1V . Thelinked pulse generator is capable of gen-erating 1?Sec pulses at 200V into anopen circuit. If high frequenc y currenttransformers are used as the detectionand launching sensors, the system can beused successfully with cables of 2 or 3km or more. The accuracy of the de viceis dependant on the rise time of the PDpulses, as the trigger must occur some-where on the rising edge.

Figure 5 shows the results of carry-ing out the locations of a cable PD with

and without the help of a transponder . Inthis case the location of the PD e vent isvery close to the measurement substa-tion, and the v ery large transpondedpulse is clear. The total cable length w asaround 750m.

The transponder developed for local-ization of cable PD’s is battery powered,allowing remote substation location mea-surements to be made, e ven if no localtest power is a vailable (a circumstancewhich is actually quite common). Thismethod for localization has pro ved verysuccessful largely as a result of its simpleand direct application. If the PD pulses

are clear, then the transponder method isvery robust, and locations are al waysclear and unambiguous.

CABLES


Figure 4 the trigger unit and large pulsegenerator of the transponder unit

Fig 5 Location of PD pulses with and with-out transponder location pulses


ET_7_05 9/16/05 3:34 PM

apply to ultrasonic signals and sensors, all of which have muchsmaller frequency content. Other methods apply for the ultra-sonic case.

It is reasonably surprising that such a simple algorithm forrecognizing local PD events is effective. However, pulses withlarger frequency content which are not PD related, are in prac-tice quite rare. The authors have seen some local processingequipment (intelligent switchgear monitors) which producedsimilar high frequency pulses, but with this exception, the vastmajority of pulses with larger frequency content tend to origi-nate from PD’s within switchgear or similar equipment.

NOISE REDUCTION FROM CONTINUOUS WAVE INTERFERENCE

A large problem in PD detection comes in the form ofradio broadcast interference. This is particularly the case foron-line measurements made on cables attached to switchgearwhich is not metal clad, although it can sometimes be a prob-lem for metal clad switchgear.

To overcome this major problem, a spectral subtractionmethod was developed. Again the ideas for this were simple,and are shown schematically in figure 8.

Essentially the largest frequency in the w aveform is cal-culated, and this frequenc y is subtracted from the w aveform,leaving all the pulse data unchanged, b ut without the interfer-ing RF noise. As the RF radio broadcast noise is in general inthe medium or short waveband, the carrier is in general ampli-


PD EVENTS FROM LOCAL EQUIPMENTFor on-line testing local equipment (switchgear , trans-

formers, bushings etc) for PD activity, the most effective sensoris an electromagnetic pickup, placed on the outside of thegrounded metal clad surf ace. Figure 6 sho ws a typical TEVprobe.

These so-called transient earth v oltage (TEV) pickupsdetect the electromagnetic pulses from inside the local equip-ment. They are essentially a capacitive coupling with the metalsurface.

This type of sen-sor is only suitable forlocal PD pickup, asthe sensors do notpass low frequencies.Typical cutoff fre-quencies are in theregion of 2MHz -5MHz, and signalsmust be abo ve thisfrequency to bedetected.

Hence for local PD, if the frequenc y content is typicallyabove this cutoff value, the chances are extremely high that thepulses are from PD’s originating in local equipment. The argu-ment also applies to testing equipment with a high frequenc ycurrent transformer as the pickup; as long as the upper fre-quency cutoff is above say 20MHz, so that the lar ger frequen-cy content of the local PD pulses is observable. Figure 7 showsa typical resonant response from a local PD e vent.

Notice the f ast oscillations denoting the lar ge frequencycontent.

The algorithm for recognizing a local PD event is thereforevery simple. If the frequenc y content of the e vent is above acritical cutoff, (somewhere in the re gion of 2MHz to 5MHz)then the event can be counted as a local event. This method can-not



Figure 6 transient earth voltage probes(TEV’s or capacitive pickups)

Fig 7 Waveform of local equipment event using a capaci-tive sensor

HUBBELL CANADA LP870 Brock Road South

Pickering, ON L1W 1Z8Phone: (905) 839-1138

Fax: (905) 839-6353www.HubbellPowerSystems.ca

HUBBELL POWER SYSTEMS, INC.210 N. Allen

Centralia, Mo 65240

Phone: (573) 682-5521Fax: (573) 682-8714

HUBBELL DE MEXICO,S.A. DE.. CV

Av. Coyoacan No. 1051 Col. Del Valle • 03100 Mexico, D.F.

Phone: (525) 575-2022 Fax: (525) 559-8626

POWER

Fargo® GR SeriesRotating Ground Tool

meets ASTM F855 Standard

...provides a continuouselectrical ground withoutinterfering with the free rotationof the conductor reel.

ET_7_05 9/16/05 3:34 PM


tude modulated. Hence the best results are achie ved if thelength of the waveform is made over a time in which the mod-ulation does not change much.

In practice for noise reduction of a waveform spanning thewhole power cycle, the length of the segments which are noisereduced are typically 100?Sec - 200?Sec. Figure 9 shows wave-forms before and after the noise reduction process.

In figure 9 the reduction in the standard deviation is around7.6 between the tw o cases. The waveform interference is rea-sonably sinusoidal, and the results w ould be e xpected to begood in this case, b ut similar practical cases also yield similarreduction ratios. The double pulse in f igure 9 is now clear andcan be used for location purposes, whereas before the noisereduction, the second pulse is not evident.

OPERATION OF EVENTRECOGNIZER METHOD

The implementation of the e vent recog-niser algorithms starts with recording thecomplete waveform of a single po wercycle of data. This is recorded at fullsample rate (normally at least100MSamples/sec). This ‘longshot’recording is no w RF noise reduced ifrequired.The resulting waveform is then analyzedfor all the pulse data, and separated intopopulations of cable, local equipmentand noise e vents. These categories cannow be treated separately if required.The pulsed data is now really a list of PDevents with properties such as position inthe cycle, channel number (if thisapplies), category of pulse, and all thepulse data from the curve fitting and fre-quency analysis. This per-cycle analysisprovides for v ery powerful techniquesfor continuous monitoring and spot test-ing. The traditional data resulting from

PD measurements was a record of the peak and count values. Inthe new method, the peak and count data can be carried out foreach category of pulses, and hence in essence for each type ofhigh voltage equipment.

Figure 10 shows the results of 2 minutes of data recording,

CABLES


Fig 8 Method for reduction of RF interfer-ence from waveforms

Fig 9 Waveforms from before and after the application of thenoise reduction process

Fig 10 PD distributions across the powercycle in separate categories of cable PDand noise

Figure 11 Waveforms of noise and cable PD respectively from thedata in figure 10


ET_7_05 9/16/05 3:34 PM

which recorded around 6 power cycles ofcomplete waveform data. In this case, thePD pulses were mixed with a large num-ber of different noise sources. The eventrecognizer has separated out the gooddata from the PD recordings, so that allthe cable PD events are correctly identi-fied. This was verified by manual inspec-tion of the individual waveforms. Figure11 shows two typical w aveforms repre-senting noise and cable PD respecti vely.If only peak and count methods had beenused, all of these pulses w ould count asvalid data. Using the e vent recognizertechnique, noise can no w be distin-guished from the cable PD e vents. Therecordings were made on an oscilloscopewith an open PC, so the event recognizersoftware can be run internally on thedevice which records the data.

This improves data transfer ratebetween waveform recorder and the PCby a factor of between 2 and 5 times.

Hence by using this method, a lar gepart of the classic dif ficulties of makingPD measurements can be transformed byclassifying the pulses into useable cate-gories. This means that less e xperiencedoperators can ha ve much more conf i-dence in PD measurements, and thatmore experienced operators can carry outdiagnostics very much more reliablyfrom one set of data.

CONCLUSIONSThe paper has described several new

tools which can help to make PD (partialdischarge) detection for high v oltageequipment much more rob ust and reli-able, by remo ving two of the lar gestobstacles to making good PD measure-ments on high voltage equipment, name-ly:

• Noise reduction of radioInterference from broadcasts and po werline carrier signals

• Separation of noise pulses fromcable PD pulses and local equipment PDpulses

The methods described follo w sim-ple algorithms, and produce reliableresults in practice. All the older tradition-al methods of measuring peak, countsand distributions for PD activity have allbeen retained. The new event recognizertools essentially sit as an add-on to theoriginal data.

However, the clarity and discrimina-tion given by the ne w event recognizermethods will allo w huge impro vements

in the quality of the PD data recorded.This especially applies to on-line PDmeasurements.

For the case of high v oltage powercables, a location method based on atransponder placed at the remote end ofthe cable is described, which allo wslocalization of PD events on cables to bemade. This is particularly useful for on-line PD measurements where it is notori-ously difficult to interpret w aveforms togive good location data.

The tools have been developed intoseveral commercial products, where theresults from the users (not the authors)have been very encouraging. ET

REFERENCES1) Kreuger, F.H. “Partial discharge

detection in high v oltage equipment”,Butterworths, 1989.



ET_7_05 9/16/05 3:39 PM


ABSTRACTThis paper will briefly describe a

new Intelligent Electronic De vice (IED)containing integrated control, protectionand communication elements for smallhydro power plant. This joint projectaims to setup tw o small hydro pilotinstallations at two Swedish utilities. Theconcept requirements and project resultswill be discussed along with results froma Simulink model of the turbine, genera-tor and station control strategies.

(Editor’s note: all references to Swedishmonetary units - Krona - are abbreviatedas SEK. The exchange rate is appr oxi-mately seven Krona to 1 US dollar)

I. BACKGROUND ANDREQUIREMENTS

Sweden has 1000 - 1500 small hydropower stations in operation erected dur-ing the last century. Due to low electrici-ty prices and withdrawn subsidies duringthe 1990s small hydro suffered from lowprofitability. Consequently about 145power stations were closed do wn orpaused during a 10-year period from1989 to 1999 [1]. During this period noor little investment was taking place. Thistrend was unwanted, since small hydro isboth an e xcellent distributed renewableenergy source as well as a part of the cul-tural and industrial heritage. However, inrecent years electricity prices ha ve risensubstantially and electricity certif icateshave been introduced to support hydropower units belo w 1500kW at an addi-tional price le vel of circa 250SEK/MWh. Hence the prof it of, and

interest in running, small hydro hasincreased and the interest to reno vatepower stations in operation and startingup new ones has grown.

Still the relative value of electricityproduced from a 1 MW station is rela-tively low, e.g. an annual production of4000MWh a 500 SEK equals to 2 MSEKor 5000 SEK per day . Therefore prof-itable operation of a small stationrequires a stable system that isautonomous to a certain e xtent and runswith few unscheduled maintenance andsupport visits. Furthermore, the systemsshould have a long life-time with a vail-ability of spare parts, rather than a tailor-made system. A change to a standardizedsolution that is easy to service is prefer-able.

Even for small hydro-po wer plants(< 2 MW) a relati vely large number offunctionalities must be included for ef fi-cient control and protection as comparedto larger power stations. Functionalitiessuch as voltage control, generator super-vision, turbine control, w ater manage-ment, start/stop, e xcitation equipmentand an interf ace to the operator controlcentral or access via a remote PC are nec-essary. Generally the requirements tocontrol all these functionalities set ademand for several building blocks to beincluded in the hydro-po wer plant con-trol system. These requirements drive thecomplexities and costs almost to the levelof larger hydro-power stations.

Although optimized production isimportant, the most vital requirement ofthe station is safe water management anda safe protection system. Moreover, if the

proportion of distributed versus central-ized generation unit connected to the gridwill grow in the future. Also the role ofthe distributed hydro power station as asource for robust grid operation is lik elyto be further strengthened,, which willincrease the requirements for protectionand interface with the grid. Additionally,the storage capacity of small hydro willbe an asset for the grid operator whenbalancing other rene wable sources suchas wind which, in turn, will increase thedemand for fle xible operation of thepower station.

I.1 OBJECTIVE OF THIS PROJECTThe objective of this project is to

show how a standardized IntelligentElectronic Device (IED) can pro videautonomous operation and prioritize safewater management. The IED should becompact and integrated with the controlof electrical and mechanical elementsincluding the flow of water, communica-tion, monitoring and protection systemfor small hydro-po wer plants. The IEDincluding all of the stated functionalitiesabove are therefore provided in one com-pact unit that easily may be carried byjust one single person.

I.2 PROJECT SET-UPThe project is realized together with

one equipment supplier, two electric util-ities and the Swedish Energy Authorityto focus on (1) a solution with the mini-mum functionalities necessary to operatethe station and (2) the most important re-

PROTECTION & CONTROL

INTEGRATION OF PROTECTION, CONTROL,MONITORING AND REMOTE

COMMUNICATION INTO ONE INTELLIGENTELECTRONIC DEVICE FOR SMALL

HYDRO-POWER PLANTSBy Gunnar STRANNE, Urban NYGREN and Lars MESSING, ABB Power Technologies; MagnusCALLAVIK, Lennart BALGÅRD, Bengt ROTHMAN and Per Erik MODÉN of ABB Corporate Research;and Lennart WIKTORSSON and Mats LINDBERG


ET_7_05 9/16/05 3:39 PM

the installed IEDs to provide an optimized electricity produc-tion by regulating an optimum flow of water through the sys-tem.

It is anticipated that using modern control and power elec-tronic equipment will decrease the size of the installation ofthe control, protection and communication equipment sub-stantially.

quirements for reliable andcost-effective operation withthe highest possible availability.Prototypes from ABB areplanned to be installed in tw ohydro-power plants owned byparticipant utilities MälarenergiAB and Tekniska Verken iLinköping AB.

I.3 DESCRIPTION OF THEPOWER STATIONS FOR PILOT

INSTALLATIONThe final prototype con-

cept will be installed in tw opower stations. One pilot isgoing to be installed at the Surahammar po wer station ownedby Mälarenergi AB. This power station was erected in 1927and is rated at 1.0MW , cf. the turbine in A. The second pilotinstallation is aimed at Slattefors po wer station o wned andoperated by Tekniska Verken i Linköping AB. Slattefors waserected in 1962 rated at 0.9MW (see B). Both stations have oneKaplan turbine with a synchronous generator.

II. THE MICROHYDROTM CONCEPTUALAPPROACH

The approach is to inte grate station monitoring, turbineand generator control functionality into one commercial highperformance protection and control IED. Voltage and currentsignals that are used by the protection unit are also made avail-able to the control system without additional wiring.Integration of functionality into one unit is considered as a keyto improve the cost performance.

Communication, on the other hand, is a technology f ieldwith fast development. Therefore the communication moduleis a separate unit, which can be replaced more often than thecontrol and protection device. It is suggested to use a commu-nication block, based on the IEC 61850 standard, with option-al remote communication techniques to a remote PC or theoperator control room. In the future it is anticipated that sever-al hydro-power stations in a common watercourse would have



PLANTSFrom page 22

Figure 1 A and B: (A LEFT) Mälarenergi’s generator at Surahammar power station. (B RIGHT) TekniskaVerken’s Slattefors power station building.

Figure 2 A and B. (A, LEFT) Example of front panel of existing con-trol and protection system at Surahammar hydro power station (B,RIGHT) Example of new station control and protection system.

ET_7_05 9/16/05 3:39 PM


II.1 THE FLEXIBLE CONTROL ANDPROTECTION PACKAGE

The configuration flexibility of therelay terminals enables protection of anygenerator and in particular systemsaccording to the Microhydro conceptwhere even more functions are requiredto be inte grated into one common plat-form. A Generator Protection Guidedescribes the applications of indi vidualprotection functions and ho w they areimplemented in the ABB 500-series ofrelay protection terminals [2]. Typicalsetting guidelines for each protectionfunction are included.

The terminals can be ordered for dif-ferent mounting arrangements (i.e. 19”rack, flush, semi-flush, w all mountingetc.). The 500 series terminals use amicroprocessor-based modular platformhardware. More than 25000 such termi-nals have been delivered worldwide andinstalled in po wer networks with ratedvoltage up to and including 750kV . Themain properties of the modular protec-tion and control terminals are:

• The protection and control terminalis fully numerical de vice (i.e. based onmicro-processor technology) with lo wpower consumption in the measuring cir-cuits;

• Great fle xibility which easilyadapts to user’s practice;

• Protection functions to be includ-ed;

• Number of current and v oltageinputs required (i.e. 10 or 20 analoginputs);

• Number of binary optocouplerinputs required;

• Number of contact outputs needed.Customer requirements for tripping

and external signaling, as well as needsfor light emitting diodes (LEDs) to local-ly indicate status of the unit, starting withprotection function and tripping etc. areeasily met. Protection and control func-tions are a vailable as standard/optionalsoftware modules and presented to theuser as functional blocks. Each indi vid-ual protective terminal can be pro videdwith built-in, multicolor 18-LED panelmodule for visual indicators. Additional40 seconds, inte grated disturbance andevent recorder can be provided for moni-toring/recording up to 10 analogue and48 binary signals per terminal. Alarms

for the follo wing operating conditionscan be provided as voltage free contacts,local LED indication or remote via com-munication:

• Generator frequency deviation (fre-quency outside pre-set band around ratedi.e. ±0,5Hz);

• Generator voltage deviation (volt-age outside pre-set band around rated i.e.±5%);

• Generator loaded/not-loaded;• Generator negative sequence over-

current or over-voltage alarm;• Generator over-excitation alarm;• Generator loss-of-field alarm;• Generator reverse power alarm;• VT fuse failure alarm;• Loss of injection v oltage for rotor

earth fault protection.

II.2 PROTECTION METHODGenerators are designed to run for a

large number of years at a high load f ac-tor and also to permit certain incidencesof abnormal operating conditions. Thesynchronous machine and its auxiliariesare continuously supervised by monitor-ing devices to reduce the e xposure toabnormal operating conditions to anabsolute minimum. Despite monitoring,electrical and mechanical f aults some-times occur, and the generators must beprovided with protecti ve relays that

quickly initiate disconnection of themachine from the system in case of afault and when necessary initiate thecomplete shutdown of the generator.

No international standard e xistregarding the needed configuration of theprotective schemes for dif ferent typesand sizes of generators. Manuf acturersprovide recommendations and there areguides given by v arious organizationStandards. The protective scheme config-urations may vary from country to coun-try and also between power companies inthe same country. There are also differentneeds to consider depending on the spe-cific machine construction. For these rea-sons the generator protective schemes areflexible.

In the Microhydro concept the gen-erator and the step-up transformer areprotected by common equipment. Thereare also back-up protections for f aultsoutside the protected zone. In practice,there are man y mechanical and thermalprotection devices that help a void dam-age to the po wer-generating unit. Theiroutput signals typically come to the pro-tection panel as potential-free contacts ormA or voltage signals and incorporatedin the o verall tripping and alarmingscheme via inputs to the terminal.


PLANTSFrom page 23


Figure 3. Typical single line diagram for generator-transformer unit and list of availableprotection functions.

ET_7_05 9/16/05 3:39 PM

II.4 COMMUNICATION HIERARCHYThe communication box is intended to work as a gateway

between the station netw orks and a WAN to enable secureremote control. Some extra server functionality must be addedto enable a local graphical HMI using just a standard webbrowser as display. The Web browser could be installed in apermanently mounted touch screen or alternatively on a laptopthat the user has when visiting the station. F or electrical pro-tection reason this network is based on optical Ethernet to pro-vide galvanic insulation. The communication solution usesthree networks for three purposes:

i. A station netw ork that can be designed as standardoffice network to enable easy connection to miscellaneousdevices, e.g. local HMI panel, web cameras and other types ofEthernet connected devices. The station network is based onEthernet (CAT5 RJ45) and TCP/IP.

ii. WAN based on Ethernet and TCP/IP to enable connec-tion to an as wide as possible range of communication w aysbetween the station and control center. This network is used toconnect the station to remote control center through a publicor private network, e.g. GSM or rented phone lines. This net-work serves as protection firewall and is VPN built.

iii. IEC61850 network. Real-time network used for con-nection to and between IEDs to the remote control center orthe local HMI.

II.3 CONTROL AND MONITORINGThe functionality required for the control of a small

hydropower station is provided by function blocks as illustrat-ed in the block scheme in. The scope includes:

• control of the water level – with various modes of oper-ation;

• normal operation at specified power, water flow, or guidevane opening;

• logic for the timing of start and stop sequences;• synchronization including required speed control as well

as the actual closing of the breaker;• control of a gate, the guide v anes, and the runner blades

by means of binary increase/decrease signals for each actuator(normally acting on hydraulic valves).

It does not include the direct control of the e xciter, but itdoes include provision of the orders to the excitation control toperform voltage control (during synchronization, until thebreaker is closed) or control of either reacti ve power or powerfactor (during normal operation). Concerning the w ater level,the following modes of operation are provided:

• Continuous le vel control – The level is controlled byvarying the flow through the gate as well as the turbine, b utusing the gate only when the required flo w is higher than theturbine can accept at the moment;

• Intermittent operation – When the water level approach-es the upper and lower limits, the turbine is automatically start-ed and stopped, respectively;

• Level control to within limits - Normally the turbine runsat the preferred operating point. When the water level reachesa limit, le vel control becomes acti ve in order to get it backinside by means of v arying the turbine flo w and if necessaryalso the flow through the bypass gate.

The function blocks have a number of parameter settingsthat allow adaptation to the particular po wer station. In addi-tion to that, one can conf igure the e xact conditions to beincluded in the start sequence by adding simple logic functionblocks as required. Apart from automatic start and stop as inintermittent operation, it is of course possible to manually givestart or stop orders, and to adjust set point v alues and limitsthat govern operation.


Go infrared...without going in the red!

1 800 613 0507www.flir.ca

The NEW ThermaCAM® EX320 provides the same superior resolution as professional infrared cameras costing $50,000or more with zoom and professional measurement tools andreport capabilities. The ThermaCAM® E25 is the most afford-able and feature-rich thermography camera in the world.

CONTACT FLIR SYSTEMS FOR A FREE DEMO.

Servicing Canada for 45 years

The Global Leader in Infrared Cameras

New HighResolution

ThermaCAM®

EX320

ThermaCAM® E25$12,995CDN

Filter capacitors

scanned and avoided

an estimated$500,000

outage.

Facility transformer captured, $200,000 cost averted.

ThermaCAM®

EX320ThermaCAM®

E25

PLANTSFrom page 24


Figure 4. Microhydro interfaces

ET_7_05 9/16/05 3:39 PM


II.5 EXCITATIONThe Microhydro con-

cept should be applicableindependent of the turbineand generator equipmentused, e.g. both for asynchro-nous generators withoutexcitation as well as for syn-chronous generators withexcitation. Since small hydrostations have been installedover a period of more than100 years, man y varioustypes of e xcitation equip-ment exist in the stationstoday. Old synchronous gen-erators usually ha ve a rota-tion device for excitation dri-ven by the mains shaft, aseparate turbine or motor .Brushes and other equip-ment in these solutions maydemand extensive mainte-nance and could discontinueoperation. More recentinstallations usually ha vestatic power electronic exci-tation equipment. The rotat-ing converter has sometimesbeen kept and sometimesbeen bypassed in recentoverhaul. Brushless systemsare uncommon in smallhydro applications inSweden, although these sys-tems would be welcomed tolimit maintenance.

The vast number of dif-ferent solutions that e xisttoday does not make it possi-ble to set one standard e xci-tation solution. The concep-tual aim is to pro vide onesolution with an inte gratedapproach and one solutionwith a clean interf ace to theexcitation hard and software.

II.6. MODEL OF THESTATION AND TURBINE

CONTROLAs a means to verify the

intended control solutionand illustrate its behavior, an


PLANTSFrom page 25

Figure 5. Intermittent operation – turbine at preferred operating level, or closed – the switching gov-erned by the water level reaching its upper or lower limits

Figure 6. Level control to within limits – turbine at preferred operating level as long as water level iswithin limits – combined flow through turbine and bypass gate is varied to control the level when sorequired in order to keep it within limits.

Figure 7. Continuous level control – the flow is varied continuously to control the level to within asmall dead zone around the set point, the dead zone enabling the control to come to rest for certainperiods of timeContinued on Page 27

ET_7_05 9/16/05 3:41 PM

Our Readers Have Their Say...

Lorraine Ilminen, Supervisor Materials Managementat Virginia Public Utilities, Virginia, Minnesota:

“I recently recei ved one of your magazines andobtained access to a very competitive vendor for Arc Flashapparel (EGO1). Thank you.”

Leslie Field, Senior Product Engineer for GeneralElectric International Relays, Corp. in Anasco, PuertoRico:

“I would like more articles on industrial protecti verelays (latest trends in protection schemes and de vices),CTs and Circuit Breakers.”

Gordon Lane, Senior Consultant Engineer with FMGlobal:

“Great Magazine! Enjoyed the Electrical Source arti-cles on Transformer Oil Maintenance and Circuit Break erMaintenance.”

Dick Essig, Senior Analyst with Globe ResearchAssociates in Riverside California:

“We enjoy reading the classified advertising section…need more advertisers there.”

implementation was made in terms of a Simulink model. Areservoir model was added, and simulations were run accord-ing to the different modes of operation – continuous level con-trol, level control to within limits, and intermittent operation.Some of these simulations are shown below. They show a caseof intermittent operation when the inflow of water to the reser-voir remains constant.

Figure 6 sho ws a case of le vel control to within limits,where an artif icial step is disturbing the reserv oir inflow. Infact, there is first a step up and then a step back down, and dur-ing this period the inflo w is lar ger than the maximum flo wthrough the turbine, so the bypass gate has to be opened as wellin order to keep the water level at the limit.

The continuous level control case (Figure 7), is appliedwhen there is no reservoir or a very small one. Here, the inflowgets high enough for a short period of time to require openingof the bypass gate. The simulation model was also extended toallow simulation of start and stop of the turbine.

II.7 BENEFITSWith a standardized integrated concept it is foreseen that

significant cost sa vings and benef its can be achie ved. Anattempt to quantify this is given in Figure 8. ET

III. ACKNOWLEDGEMENTFunding for part of this project by the Swedish Ener gy

Agency is greatly acknowledged.

IV. REFERENCES[1] Widmark, D., Småskalig v attenkraft och kultur-

miljövård, Riksantikvarieämbetet, PM 2002:6[2] ABB Product Guide for RET521 at

www.abb.com/substationautomation


PLANTSFrom page 26

Figure 8. Benefits with a standardized integrated concept

YOU SAID IT

ET_7_05 9/16/05 3:41 PM


SUMMARYThis article presents some consider-

ations in failure mechanisms of oil-barri-er transformer insulation and oil-papercondenser type insulation of instrumenttransformers and HV b ushings on thebasis of service e xperience and specialstudies.

1.TYPICAL FAILURE MODES OFEQUIPMENT

Table 1 summarizes some typicaldefects and f ailure modes of oil-barrierinsulation of transformer , shunt reactor ,condenser type bushings and instrumentcurrent transformer that were suggestedon the basis of service experience [1,2,3]considering a wide range of equipmentdesign and operation conditions.However actual failure-modes and failurecause can be substantially dif ferent fordifferent insulation design. Analysis ofoperation experience with transformerand other equipment in CIS countries forthe last 10 years has shown the followingpicture.

Power TransformersInsulation problems in volve pre-

dominantly impairment of insulationcondition in service.

Excessive moisture in the celluloseinsulation is inherent basically to thetransformers with open-breathing preser-vation system or to those which ha ve aninsufficient sealing. Distrib ution of themoisture in the course of the transformerlife is k ept quite non-uniform. Most ofthe water is stored in so-called “cold thinstructures”, namely in the thin press-board barriers that operate at b ulk oiltemperature.

General “aging” problem is accumu-lation of conductive and polar particles inoil and depositing them on the service.

Insulation surface contaminationhas been observ ed in the forms ofthe adsorption of oil aging productswith cellulose or deposits of con-ducting particles and insolubleaging products in areas of high elec-trical stresses. The surface contami-nation can cause a distortion ofelectrical field and a reduction inthe electrical strength of the insula-tion system.

Relative failure rate of po wertransformers 100 MVA and abo vein CIS countries is about 1%. Therate of major f ailures of insulationis about 0.5%. About 25-30% ofmajor failures are associated withbreakdown of major and minorinsulation being in service over 20-25 years due to accumulation ofmanifold agents of degradation.

The most lik ely failure modesare the following:

– occurrence of critical PD inoil due to penetration of free w aterthrough bad sealing that results inbreakdown between coils typicallyof HV winding and breakdo wn ofoil gap between HV winding andtank;

– flashing o ver HV windingunder effect of switching surge andlighting impulses due to contamina-tion surface likely with conductiveparticles and polar oil aging prod-ucts;

– progressing creeping dis-charges between phases andbetween winding – to ground. Thistype of f ailure mode has beenobserved only with some old designas a result of se vere contaminationof insulation with metal particlesand distortion of insulation struc-ture due to winding buckling.

TRANSFORMERS

INTERNAL INSULATION FAILURE MECHANISMSOF HV EQUIPMENT UNDER

SERVICE CONDITIONSBy A.K. LOKHANIN, T.I. MOROZOVA of the All-Russian Electrotechnical Institute; G.Y. SHNEIDER,Electrozavod Holding Company; V.V. SOKOLOV, Scientific and Engineering Centre ZTZ-ServiceResearch Institute and V.M. CHORNOGOTSKY of the Ukrainian Transformer Research Institute

Table 1. Typical defects and failure modes in insu-lation of oil-immersed transformer equipment.


ET_7_05 9/16/05 3:41 PM

losses, followed by thermal run a way and subsequent ionisa-tion process. Aging of oil, oil/paper b ulk, and in some caseslocalized moisture contamination ha ve been observed as themost probable life-limiting factors.

2. FAILURE MECHANISMS OF OIL-BARRIER INSULATION2.1 Summary of transformer models investigation

Short-term and long-term tests of aged oil-barrier insula-tion on the basis of life models studies ha ve revealed the fol-lowing aging-mode problems:

Reduction of switching surge breakdown voltage due to adeposit of insoluble aging products in areas of high electricalstress and stimulation of surface discharge occurrence [4].

The minimum breakdo wn voltage at switching sur gesmay decrease approximately by 15% after aging. Increasingthe concentration of particles in oil (from 50 cm-3 up to 160cm-3) may decrease switching surge breakdown voltage addi-tionally by 10%.

Long-term dielectric strength tests re vealed two mecha-nisms of insulation breakdown:

- “accidental” breakdown was observed during the f irstperiod of aging due to the influence of particles and moisture;

- “wearing mode” breakdown due to degradation of mate-rials appeared at the last stage of the calculated term of aging.The latter results in decreasing the initial dielectric strength by30-40%

Probable cause for the reduction of electrical insulation

Shunt Reactors 400-750 kVThe relative rate of major failures is about 1%, and about

40% of forced outages are associated with insulation problems.The following failure modes were involved:

– flashover along the winding and traces of dischar ges onthe barrier facing to the winding;

– flashover between pair or several coils;– flashover along the inner surface of low porcelain of the

bushing.In many cases se vere contamination of insulation with

metallic-dust view particles w as found. Wear of the pumpsbearings, aluminium shield attrition and localized oil heatingwere recognized in some cases as sources of particles genera-tion.

HV BushingsBushings cause about 45% of transformers major failures.Aged mode failure occurs predominantly. About 80% of

failures take place after 10-12 years of service and o ver 30%after 20-25 years. Core f ailures happen practically only withunsealed design. Two failure modes ha ve been suggested:ingress of free water resulting in occurrence of critical ioniza-tion at the bottom part of the core, and aging of oil paper bulk,occurrence of excessive dielectric losses, that results in thermalinstability.

Experience has shown a very low rate of core f ailure ofbushing designed with sealed preserv ation system incorporat-ing overpressure bellows (about 10 cases per about 100 000bushings population since 1971). Ho wever experience hasshown also that hermetically sealed preserv ation system ismore sensitive to residual contaminants, and occurrence ofaging by-products. Special requirements shall be specif iedregarding to selection the proper oil, compatibility tests, pro-cessing procedures.

Unusual failures of condenser core ha ve been observ edwith circuit breaker’s bushings of free breading design.

The problem is associated with X-w ax formation at lo wtemperature.

The main f ailure mode of hermetically sealed b ushingsinvolves overflashing along internal lo wer porcelain due topredominantly oil aging and deposit of semiconducti ve stainon the porcelain surface.

Current Transformers 220-750 kVThe failure rate of 220-500 kV CTs accounts for more

than 60 % of the total instrument transformers f ailures.The rate of major failures of insulation is about 0.35%.Two failure-modes have been typically observed:1. Ionisation-mode f ailures that occur predominantly in

wintertime. These are referred to as “cold failures.” They typi-cally happen on the “toroid shape” designs and are caused byde-impregnation of the condenser core, ingress of air duringstorage, and o ver-saturation with air that results in b ubblesevolution after fast cooling in service.

2. Aging –mode f ailures that ha ve occurred after 15-25years of service in the summer are referred to as hot f ailures.Those involved are typically openbreathing CTs and with her-metically sealed units filled with oil having some elevated (14-18%) aromatic content, and are caused by increased dielectric


POWERBehind the CommitmentBehind the Commitment

Your world class source for

Generation & Distribution Transformers

Look to Fortune Electric for other products:

Metalclad SwitchgearMotor Control Center

1965 Shenango Valley Freeway Hermitage, PA 16148-0504 USAPhone: 724-346-2722 Fax: 724-346-1472URL: www.fortune.com.twe-mail:[email protected]

FORTUNE ELECTRIC CO.,LTD

INSULATIONFrom page 28


ET_7_05 9/16/05 3:41 PM


strength is the formation of aging prod-ucts, specifically, of anionic surf actantsand bound w ater, which is not deter-mined by the typical methods.

2.2 Mechanism of the incipient irre-versible failure in oil-barrier insula-tion:

The first stage if incipient f ailure isassociated with critical PD occurrence.

Mechanism of PD occurrence andprogressing depends substantially of theconfiguration of electrical field. Presenceof significant magnitude of tangentialcomponent of electric f ield intensity iscritical to stimulate surface and creepingdischarges.

Typically the f ailure is initiated bythe breakdown of oil gap, which is regis-tered as an apparent char ge in excess of10-8 C that rises rapidly to 10-7-10-6 C.It progresses in surf ace discharge in oil

across the barrier with PDmagnitude over 10–6 C.

“White marks” appear onthe surface due to forcing oiland water out of the press-board pores followed with car-bonised “black marks” on thebarrier. PDs of 10-7-10-6 Cmay cause an irre versibledamage during tens of hours.

Minimum energy requiredto cause incipient carbonizingin the cellulose (heating o ver300 0C) is estimated as 0.1 J,which corresponds to se veralcharge pulses of 10-7-10-6 C

Stable discharge in oilassociated with PD po wer isP>0.4 W.

An average rate of gasgeneration under the ef fect ofstable PDs in oil is 50 ?l/J

Further steps progresseither in the breakdown of theinsulation space or in theoccurrence of creeping dis-charge.

TRANSFORMERS


Table 2. Conditions for creeping discharge occur-rence (service experience and experiments)

Testing and Maintenance Handbook

UPS/Standby Power Handbook

Wire and Cable Handbook

Circuit Breaker and Switchgear Handbook

Electrical Safety Handbook

Variable Frequency Drives Handbook

Programmable Logic Controller Handbook

Industrial Electricity Efficiency Handbook

Electrical Transformer Handbook

Electric Motor Handbook

Power Quality and Electrical Grounding Handbook

Electrical Protection and Control Handbook

DETAILED AND FOCUSED, THESE PROBLEM-SOLVINGhandbooks are a must for your electrical library

Regular Price:$35 (Cdn)/$27.50 (US)

Order Now to receive 30% offregular price

(905) 686 - 1040www.electricityforum.com/estore

Special Offer

Special Offer

to save even more...

Electricity Forum Book of the Month Club

50% off regular price.

$199CDN or $169US plus S&H

The Electricity Forum 15 Harwood Ave. South, Suite 204 Ajax Ontario L1S2B9 tel: (905)686-1040 fax: (905)686-1078

Sale Price!$24.50 (Cdn)/$20.50 (US)

The Electricity Forum 215-1885 Clements Road Pickering Ontario L1W 3V4 tel: (905)686-1040 fax: (905)686-1078


ET_7_05 9/16/05 3:41 PM


2.2.1 Creeping dischargeThis is, likely, the most

dangerous failure mode thattypically results in cata-strophic failures at normaloperating conditions. Thephenomenon occurs in thecomposite oil-barrier insula-tion and progresses in sever-al steps (table2):

- partial breakdo wn ofoil gap;

- surface discharge in oilacross a barrier (an appear-ance of black carbonisedmarks on the barrier);

- microscopic sparkingwithin the pressboard orbetween layers of pressboardsheets where gas e vacuationis limited. The presence ofsome excessive moisturestimulates vapour bubblesforming and de gradation ofmaterial;

- splitting oil moleculesunder the effect of sparking.

The formation of hydro-carbons followed with theformation of carbonisedtraces in the pressboard. Thisprocess is resulted in lo wer-ing magnitudes of PD appar-ent charges to 5·10-10-10-9C however PD energy is stillhigh enough to destruct cellulose.

Creeping process can continue fromminutes to months or e ven years, untilthe treeing conductive path causes shunt-ing of an essential part of transformerinsulation resulting in a powerful arc.

The cellulose destruction while dis-charge progressing within the pressboardis associated with PD q ≈ 10-9 C; PDpower P=0.1-1 W, average rate of gasgeneration of 40-50 ?l/J.

Three critical f actors could beadvised to evaluate probability of creep-ing discharge occurrence:

- Source of initial critical ionisationof high energy to cause carbonised marksof the barrier;

- Specific insulation design configu-ration;

- High enough dielectric stresses.

2.2.2 Surface dischargeOccurrence of surf ace discharge is

associated with increased voltage (transit

overvoltage). Two mechanisms could besuggested on the basis of studies (table3):

1) Oil breakdo wn progressing intoinsulation destruction;

2) Surface discharge as self-f iringphenomenon.

A magnitude of tangential compo-nent of electric field stress that can resultin PD occurrence and forcing oil out ofpressboard (reversible–mode “whitemarks») could be suggested as criterionof dielectric strength across insulationsurface. Investigation of models ofunaged, dry and clean insulation hasshown that surf ace discharge can occurunder action of electric f ield stress 6.5-12.5 kV/mm on condition if ratio of aver-age and maximum f ield intensity in oilgap is 0.4-0.5 or less.

Apparently, contamination of sur-face with conducti ve particles reducesthe value of critical field intensity.

3.FAILURES OF BUSHINGS CAUSEDBY STAINING OFF THE LOWER

PORCELAIN3.1 Failure mode induced by peculiaroil aging and staining of internalporcelain

Discharges across the inner part ofthe transformer end porcelain are out-come of a typical aging-mode phenome-non in the b ushing [3]. The failureprocess is initiated and de veloping with-in the oil channel between the core andlower porcelain.

Electric field intensity in the oilchannel and across the surf aces of core-end components and inner porcelain isestablished both by the b ushing insula-tion design and by disposition of thebushing end relati ve to the groundedparts and the winding.

Formation of semi-conducti veresidue on the inner surf ace of the lowerporcelain has been follo wed and knownto reduce the dielectric safety mar gin ofthe insulation space, causing a flash overacross the inner surface.

The failure process may be intro-duced as the following:

- degradation of the oil and forma-tion of oil decay products, particularlycolloids containing atoms of metals; evo-lution of anionic surf actants and hydrate(bound) water;

- deposits of semi-conducti ve sedi-ment on the surfaces under effect of elec-trical field;

- distortion of electrical field, changein the distrib ution of the v oltage alongthe porcelain; reducing the dielectricstrength of the oil due to transfer ofbound water in dissolved state and parti-cles [5];

Power • UPS • NETA CertifiedField Service Engs • Techs • Mgrs

Design • Vibration • Motor RepairSales • Power Quality • Pred/Prev MaintSince 1980, we've specialized in placing people whotest, repair, maintain or sell electrical apparatus, elec-tronic or electric power equipment. Nationwide. Allfees company paid. Call or send confidential resumeto LISA LINEAL: RecruitingLINEAL Services46 Copper Kettle Road, Trumbull, CT 06611

TOLL FREE 877-386-1091 • Fax: 203-386-9788Email: [email protected] • www.lineal.com

Se Habla Español



Table3 Mechanisms of surface discharge origin andprogressing; HV Power frequency tests of models

ET_7_05 9/16/05 3:47 PM


- surface discharges, gas e volution,and flashover.

Electrical Effect of BushingInstallation

Approach the b ushing end to thegrounded components can increase f ieldintensity by 20-25% and essentially dis-tort the inner field picture.

In some cases (e.g. some design ofshunt reactors) a large stressed volume ofoil is set up, which is very sensitive to oilcontamination. Spare mar gin of dielec-tric strength of the b ushing integritydetermines by dielectric strength of theoil channel between the core and porce-lain.

Reducing dielectric strength of oilcan critically reduce the mar gin andcause partial discharge activity

Electrical field impacts on chemicalreactions in oil and f avours coagulationof colloid.

Electrical field across the stainedbushing surface attracts and picks up theconductive –mode particles out of thetransformer oil.

Thermal Effect of TransformerHeat, which is radiated from trans-

former tank, determines air temperaturearound the air-end of the bushing.

Transformer oil is the main source ofbushing heating.

Another two sources are dielectriclosses in the core and resistance losses inthe central conductor. The latter does noteffect essentially on temperature distrib-ution if bushings current is less than 50%of rated.

The basic heat exchange is expectedto be in the oil channel between the coreand porcelain, specif ically close to thebottom of the mounting tube approxi-mately on the le vel of top oil of trans-former, where principal mass e xchangetakes place. Here con vection flow turnsdown close to the surface of the core. Themaximal temperature of b ushing oil insome place can be equal to the top oiltemperature or even above that.

Peculiar process of bushing oil cool-ing favours formation of colloids.

Effect of oil type and compatibilityof the oil with other materials

Field experience and laboratory testsof numerous bushing models have shown

that significant impairment of b ushingscondition have been occurred practicallyonly with acid treated oil containing rel-atively high aromatic carbon-hydrogencontent (CA=17%), that has sho wn thefollowing properties: high hygroscopic,an elevated content of bound w ater; ani-line point 710C that promotes interactionwith rubber gaskets and especially withplasticizer, inclination to form polaraging products and sludge [6].

It was shown also that oil-proof gas-kets have shown different effect of oilaging acceleration. Particularly, a gasketwith plasticizer can contribute to oil dete-rioration and promote the formation ofsemi-conductive sediment on the porce-lain.

3.2 Failure mode induced by stainingthe outer surface of bottom porcelain

Experience has shown that dielectricwithstand strength of the oil part of HVbushing could be very sensitive to conta-mination of transformer oil with conduc-tive particles.

There have been several documentedcases associated with a deposit of carbonon the lower porcelain, which originatedfrom the localized o verheating of thecore, and with deposits of iron particleson the porcelain surface, which originat-ed from pump bearing wear.

Formation of metal-contained col-loids resulting from aging of oil with rub-ber and transformer metals has beenobserved also as a typical origin ofporcelain contamination

Electrical field around the b ushingstimulates picking up the particles by theporcelain (a trap effect).

Uneven distribution of voltage alongthe porcelain close to the gasket joint andpresence of electrical field in this area isa typical phenomenon for condenser typebushings.

Maximum field intensity is locatedjust in the vicinity of the gasket joint.

Increasing the f ield intensity maystrengthen “a trap”- ef fect and result infurther extension of the staining area.

The following mechanism of b ush-ing deterioration could be suggested:forming of conducti ve-mode residuebecause of comple x aging of the trans-former oil in the presence of a leachingsubstance of rubber gask et and trans-former metals. This is in combinationwith dissolved and suspended metals inthe transformer oil could cause adher-ence under influence of an electrical fieldformed by the bushing and the grounded

parts of transformer.The difference in quality of the oil in

the transformer and that used in the bush-ings and also some difference in oil tem-perature could be the main f actors toexplain occurrence of the formation ofthe stain on the outer surface only.

A local increase of the electricalfield intensity in the “Bushing-T ankWall” space could cause a concentrationof conductive substances in this re gion.Accordingly, that could e xplain the for-mation of the semi-conducti ve streaksalong the porcelain.

The following dangerous f actors ofeffect of the stain should be considered:

- reduction of the porcelain leakagepath (or insulating distance) and accord-ingly reduction of the incipient voltage ofsurface discharge appearance;

- mitigation of the shielding effect ofthe grounded shield (grounding layer)and accordingly strengthening the elec-trical field across the surface close to thegrounding sleeve and particularly near togasket location;

- strengthening the “trap”-ef fect ofthe stained porcelain to attract conduc-tive-mode particles out of transformeroil.

Presence of the conducti ve stain onthe porcelain may be introduced as“extension” of the grounding slee ve,which results in diminishing the shield-ing effect of the grounding layer of thecore. This phenomenon may cause sig-nificant increase maximal field intensity.Accordingly, it could e xplain substantialreduction of the dielectric withstandstrength of the b ushing and possibleoccurrence of PD in the area close to gas-ket joint location of the stained bushing.

3.3 Failure of oil-filed paper insulation of non-sealed bushings ofHV circuit breaker.

When disassembling a b ushing of220 kV oil-f illed circuit break er, failedduring operation, mere has been foundplenty of X-w ax. The failure occurredafter operation for three years approxi-mately. DGA analysis of oil samplesfrom sister-type bushings (more than 20pieces) has shown high concentrations ofhydrogen (up to 3.5%), lower but exceed-ing permissible values of methane (up to0.5%), ethane (up to 0.2%), carbonmonoxide (about 0.5%). The bushingswere put out of operation, and deli veredto the manufacturer for investigations.

TRANSFORMERS



ET_7_05 9/16/05 3:47 PM

occurrence (the sharp edges of electrode), the number of PDpulses (about 10-14 pulses/minute) and their apparent char ge(900 - 1300 pC) are the same as at the room temperature (+20°C).

In the non-sealed v essels, when testing specimens ofpaper- oil insulation which were artif icially moistened, atcooling- warming cycles for oil: +20°C> -20°C> +20°C, thenumber of PD pulses increased to 20-25 pulses/minute but mevalues of their apparent charges were equal to 100-300 pC. Inthese specimens, carbonised traces of PD took place in thearea of uniform field. There were no traces of wax formation.

When testing the specimens, who were naturally moist-ened, the character of PD was the same as at tests of the spec-imens moistened artif icially. However when disassemblingthem, X-was was detected between layers of paper.

Therefore the supposition that moistening of oil in a non-sealed equipment follo wed by X-w ax tonnation can be thecause of PD occurrence in paper -oil insulation in uniformfield, has been confirmed. ET

REFERENCES[1] A.F. Kurbatova, V.V. Sokolov, O.N. Grechk o,

V.P.Mayakov, A.S.Kolesnikov, V.M.Chornogotsky,Development of Diagnostic System of 330-750 kV CurrentTransformer based on service Experience and Endurance Tests(CIGRE, 1998, Paper 12-107)

[2] Victor Sokolov, Vladimir Mayakov, GeorgyKuchinsky, Alexander Golubev, On-Site Partial Discharge

The later tests on a part of these b ushings showed anessential increase in tan ? for the main insulation depending onvoltage, and high level of partial discharges (up to some hun-dreds pC). When doing so, e ven at low- voltage (10 kV), thevalue of tan? exceeded the nameplates values as large as 1.5 -2 times.

All the bushings considered above, have been filled withoil of type GK. In b ushings filled with oil of type T-750, hav-ing a non-sealed construction, with an insulation constructionwith is quite identical to that of the bushings considered, at thesame operational electric f ield strength, w ax formation hasnever been detected. Therefore the cause of PD occurrence inthe bushings considered shall be searched in dif ference inproperties of oils GK and ?-750. The difference in conse-quences of PD action on oils GK and ?-750 are caused by thefact that oil of type GK has the more ability of gas emission (itsrelevant factor is greater by 2.5 fold than that of oil T-75O,therefore creation of conditions for w ax formation process^formation of gas-oil mixture) is easier in oil GK than in oil T-750. Since the value of solubility of atmospheric gases for oilGK is practically equal to that for oil T-750, there are no rea-sons to suppose that the cause of PD occurrence in the bushinginsulation is bound up with formation of gas b ubbles whenchanging the oil temperature.

Experimental investigations ofinsulation specimens

The specimens (two little sheets of cable-insulating paper,each of them is 0.12 mm thick) were the combinations of paperand oil GK, having the different moisture content:

• ‘dry’ paper (Wp = 0.5% - 1.0%) and ‘dry’ oil (W0 - 5g/t.-9 g/t);

• ‘dry’ paper and ‘wet’ oil (WQ = 20 g/t - 40 g/t);• ‘wet’ paper (Wp > 4%) and ‘wet’ oil.The specimens were tested in sealed and non-sealed v es-

sels.A part of specimens was at the room temperature (+20°C)

all the time. The other part of them was being cooled to +5°C,-10°C, -15°C and -20°C. After cooling and e xposure at thistemperature, a voltage was being applied to the specimens, andthey were warming within 1.5 to 2 hours. A voltage was beingapplied for 6 to 8 hours e very day. The total time of the v olt-age action was equal to 50-55 hours. The voltage value corre-sponded to that for the occurrence of steady PD (50 - 100 pC).At the voltage application, the apparent charge and the numberof PD pulses were registered.

Some specimens of ‘dry’ paper and ‘dry’ oil were beingnaturally moistened within 1,5 month, this increased the mois-ture content of the oil to 25 g/t. Some other specimens weremade of paper of a f ailed bushing, which contained moisture(about 20 g/t), wax and oil.

It was revealed that the PD inception v oltage varied littlefor all the specimens, and w as equal to 11 - 13 kV . HoweverPD development character essentially depends on the moisturecontent of specimens, temperature conditions and hermeticsealing of the device.

In the sealed vessel, irrespectively of moisture content ofspecimens, at cooling-warming cycles for oil: +20°C, -20°C,+20°C> the PD de velopment process, the location of their


Dry It On-LineWet Transformer?

Reduce Substation Maintenance Costs!

• Transformer remains in-service during process.

• Reduce transformer dryout costs• Improve oil dielectric values• Improve transformer power factor• Operates continuously unattended• Extend oil service life• Extend transformer service life

Velcon’s Transformer Dryout System is an on-line dryout process designed for safe, unattended, 24-hour operation.

COLORADO SPRINGS, CO, USAPhone (719) 531-5855

FAX (719) 531-5690 VISIT OUR WEB SITE AT: www.velcon.com


ET_7_05 9/16/05 3:48 PM


I. INTRODUCTIONSwitch Mode Po wer Supplies

(SMPS) are being introduced to the elec-trostatic precipitator mark et at pricepoints that are very competitive with lin-ear transformer/rectifier sets and SCR-based controls that have been used in theindustry for many years. The new switchmode power supplies ha ve dramaticallydifferent performance and physical char-acteristics than the linear power suppliesthey will be replacing. As applied to theESP application, the new power supplieswill have a major impact on man yaspects of precipitator design, construc-tion, operation, and maintenance.

Suppliers and users of ESPs need tobecome familiar with this ne w powersupply technology, in order to assess thespecific effect that its introduction willhave on their businesses. This paper pro-vides an overview of the changes that canbe anticipated as SMPS are applied toESP applications.

II. DESCRIPTION OF SWITCH MODEPOWER SUPPLIES

A functional block diagram of a typ-ical switch mode power supply is shownbelow:

The AC/DC block tak es the three-phase input and rectif ies and filters it tocreate a fairly smooth DC bus of approx-imately 650 volts DC. The DC/AC blockconsists of an inte grated gate bipolartransistor (IGBT) full bridge circuitwhich converts the DC b us into a highfrequency AC waveform. The ResonantTank block, combined with the lastAC/DC block, steps up the high frequen-

cy AC, rectifies it, and thus delivers highvoltage DC to the ESP load. This block,which is oil-filled, is the high frequenc yequivalent to a con ventional 60hz. T/Rset.

Difference in Circuit Toplogy OverLinear Power Supplies

Due to the fact that the step-up trans-former operates at high frequenc y, it canbe approximately 1/10 the size andweight of an equi valent 60hz. trans-former. It will also use signif icantly lesscooling fluid. The circuit design requiresthat the AC/DC and DC/AC modules belocated very close to the step-up trans-former. Therefore it is both practical andnecessary for the switch mode po wersupply system to consist of one physical-ly integrated package, unlike current sys-tems in which the control cabinet andtransformer/rectifier set are separateunits, often located a considerable dis-tance from each other.

Since the IGBTs can be quicklyturned off on command and do not w aitfor the line current to decay to zero, thecircuit can turn on and off over 250 timesfaster than a linear po wer supply. Thiswill allow new control algorithms with

particular benefits for intermittent ener-gization.

Because the AC input voltage is rec-tified, filtered, and then switched v eryfast (25 kHz) in small pack ets of energy,the ripple voltage is only 3-5% of the DCvoltage level as compared to 35-45% forlinear 60 hz. power supplies.

The three-phase input results in amuch higher power factor (.94 vs. .63),

and lower power consumed per kilo wattof power delivered to the precipitatorload.

In summary the major differences inelectrical characteristics for a switchmode power supply are:

• Small inte grated control and T/Rpackage

• Faster control response• Higher average output voltage• Less power consumedEach of the impro vements listed

above has a potentially lar ge impact onESP construction, operation, and mainte-nance. The changes in control responseand output v oltage seem lik ely toenhance the actual collection ef ficiency,and hence the o verall performance ofmany precipitators. F or a detailed com-parison of switch mode and conventional60-hertz precipitator po wer supplies,refer to the table in Appendix A whichcompares 19 performance and physicalparameters.

The smaller integrated packages canchange the way ESP power systems aredesigned, installed, and serviced. Thefollowing section of this paper e xamineseach of these issues in more detail.

III. ENHANCED COLLECTIONEFFICIENCY DUE TO SWITCH MODE

Power SuppliesThere is a growing body of evidence

supporting the f act that the ne w switchmode power supplies pro vide improvedcollection for many precipitator applica-tions. This seems to be primarily due totheir ability to deli ver a higher a veragevoltage and current to the ESP load,while operating just belo w the sparkinglevel and/or the back corona le vel forhigher resistivity ashes. This benefit islikely for all ashes: lo w, medium andhigh resistivity.

An EPRI-funded study did field testsusing a prototype NWL SMPS on a slip-stream precipitator at Alabama Power-Plant Miller. The results were v ery

TRANSFORMERS

SWITCH MODE POWER SUPPLIES FORELECTROSTATIC PRECIPITATORS

By David Seitz and Helmut Herder, NWL Inc.


ET_7_05 9/16/05 3:48 PM


encouraging in thatdepending on the testconditions approxi-mately 30% improve-

ment in percentage penetration w as achieved. A conventionallinear T/R set and SCR control was also tested to provide a basereference level.

The table shown in Appendix B details the test conditionsand results. It may be of interest to note that this study also test-ed a SMPS combined with a narro w width (1-3 micro-sec)pulsed power supply. This combined unit pro vided the mostimprovement.

This is also detailed in the results table.Testing conducted at the Lansing Board of Water and Light

Moores Park facility in Lansing, MI is also of interest. This wasa steam generation plant b urning low sulfur (1%) coal.Powerspan Corp. (formerly Zero Emissions Technology) con-ducted tests of a three-phase linear po wer supply system, theoutput of which is similar to what SMPS units will provide. Thetesting showed a higher average power and voltage delivered tothe ESP load, as well as reduced sparking and impro ved col-lection. The results were documented in a paper presented atthe American Power Conference Proceedings, “PerformanceImprovements From the Use of Low Ripple Three Phase PowerSupply for Electrostatic Precipitators.”1

A paper presented at the September 1998 proceedings ofthe ICESP, “A Novel and Versatile Switched Mode Po werSupply for ESPs, ”2 documented precipitator performanceimprovements due to the use of SMPS. The testing was done ona cement kiln ESP under v arious operating conditions. In theconclusions the authors stated that “the prototype test has indi-cated that this SMPS can improve collecting efficiency, both incases of low and moderate dust resistivity. The improvement ishigher in case of occurrence of back corona.”

A second paper presented at the same ICESP conferencealso discussed ESP performance improvements achieved by theuse of SMPS. This paper, “The HVDC Current Source F orESP,”3 concluded that SMPS units pro vided higher a veragevoltage, reduced sparking, and hence improvements to precipi-tator efficiency.

An early adopter of the f irst NWL-built SMPS units hasshown promising results. PPC installed an SMPS unit as areplacement for a 60 hz. power supply on a paper mill ESP. Thefollowing is an e xcerpt from the f ield engineer’s report toNWL:

“The existing power supply was operating in the low 30kvrange and was sparking heavily. When we installed the SMPSwe were able to increase the secondary kv to approximately40kv — a dramatic improvement. The SMPS now runs currentlimited at or near 40kv with a clear stack. ” PPC has installedSMPS units at other sites, and continues to report improved col-lection versus 60 hz. units.

The testing done to date on the use of SMPS can not yet beconsidered to be conclusive proof of enhanced precipitator col-lection under all conditions. The industry is still in the earlystage of understanding the total effects of these new power sup-plies on ESP performance. Ho wever, at this point in time it isreasonable to assume that for man y applications, performanceimprovements can be achieved through the use of SMPS units.As more units are f ield installed, further data will be accumu-lated and analyzed. NWL will continue to publish such infor-mation as it becomes available.

IV. IMPACTS OF A SMALL INTEGRATED PACKAGEThe different physical configuration of an SMPS unit will

have an af fect on at least three major aspects of precipitatordesign, construction, and operation:

1. Physical arrangement of ESP rooftops, control houses,bus ducts, etc.

2. Maintenance and servicing3. Sourcing of the power supply and other related compo-

nents

Physical ArrangementA switch mode power supply is the electrical equivalent of

the control cabinet, T/R set, and current limiting reactor used ina linear 60 hz. ESP po wer supply. The SMPS is contained inone outdoor enclosure, while the con ventional unit consists ofseparate indoor and outdoor enclosures. A quick comparison ofthe key physical characteristics for a 1000 mili-amp ESP powersupply shows the following:

Characteristic SMPS Linear 60 hz.Footprint (square feet) 8 18Weight (lb.) 750 3900Fluid Volume (gallons) 25 140

The size and weight reductions from using SMPS unitsoffer new approaches to ESP design. There is no longer a needfor a large climate-controlled space for control cabinets, and thepower and control wiring to the units will be simpler and easi-



October 25-27, 2005Baltimore Convention CenterBaltimore, MDProfessional Advancement CoursesOctober 23-24, 2005

October 25-27, 2005Baltimore Convention CenterBaltimore, MDProfessional Advancement CoursesOctober 23-24, 2005

PQSA-ET

Part of Presented byCo-located with

REGISTER TODAY! Call 1-800-927-5007 or visitwww.powersystems.com

For Conference information visit our website!

Register NOW and receiveFREE admission to the

Power Quality Exhibit Hall!

Save Money on Equipment Failure Due to Sags, Swells, Transients, Harmonics,

Brownouts, and Power Failures

ET_7_05 9/16/05 3:49 PM


er to install than with conventional powersupplies—resulting in signif icant antici-pated cost savings for new ESP construc-tion.

Another major change with SMPSunits is that their size and weight allo wthe possibility of mounting them directlyonto entrance flanges, thus eliminating asignificant amount of duct w ork andoffering another opportunity for costreduction on ne w precipitator construc-tion.

The reduction in fluid v olume maydecrease the need for e xpensive high-flash-point fluid, or if such fluid is stilldesired, there will be a much lo wer costpremium than for current systems. Thefluid containment issue is now much eas-ier to manage due to the lar ge volumereduction.

Maintenance and ServiceHistory has shown that for a number

of reasons the high voltage step-up trans-former contained within the T/R set isone of the more likely components to failwithin an ESP power supply system.

When such a failure occurs, the T/Rset must be removed from the ESP for atime-consuming repair or replacementprocess. Furthermore, the T/R set is themost expensive component within thesystem, thus preventing many ESP usersfrom stocking complete spare units,thereby creating a difficult problem whena transformer failure does occur.

The situation is dramatically dif fer-ent with an SMPS unit. The T/R set com-ponent of a 1000 ma SMPS unit willweigh roughly 300 lbs. and cost signif i-cantly less than a 60 hz. T/R. This makesit more af fordable for the ESP user tokeep spare T/R sets, as well as making itmuch easier to get them on and of f theESP.

SMPS T/Rs will also ha ve far lessvariation in electrical ratings and physi-cal arrangements than 60 hz. units. Onekey reason for this is that the input to theT/R comes from the in verter, and isalways the same v oltage and frequency,and thus a major T/R set rating variable iseliminated. Less variation in ratings will

TRANSFORMERS



Appendix BChart 1

Summary of Okant Miller Test ResultsHigh Frequency Pulsing Power Supply

Appendix AComparison of SMPS to 60hz. Power Supplies

1000ma Unit

ET_7_05 9/16/05 3:50 PM

required to adapt to SMPS.More importantly, the use of SMPS units will mandate

single point of contact for both the purchase and maintenanceof the ESP power supply system, thus of fering obvious bene-fits to ESP OEMs and users.

CONCLUSIONSThe introduction of switch mode po wer supplies for use

on electrostatic precipitators will have a number of impacts onthe industry:

• Improved collection performance• New options for locating and installing power supplies• Different approaches to maintenance and service• Changes to power supply purchasing strategiesEach of these factors offers potential cost savings for both

new and retrofit installations. Lower power consumption dueto the improved power factor will result in direct cost savings.The full impact of all the potential sa vings has not yet beendefined.

We are still in the early stages of understanding ho w tobest apply SMPS units to ESP applications.

The systems commercially a vailable today are relati velyunproven and offer limited power and voltage options. But aswith all new technology, the systems will continue to improve.Higher power and v oltage ratings and more features andoptions will soon be available. There is little doubt that switch

make it practical for the SMPS supplier to maintain an in ven-tory of replacement T/R sets for quick shipment to a job site.

Furthermore, in emer gency situations, air -freight ship-ment will no longer be prohibitively expensive due to the rela-tively light weight of the T/R set. When an ESP user kno wsthat replacement T/Rs are readily and quickly available, it maythen be practical to fore go the expense of carrying spares onsite. In summary, the user no w has several better options forservicing T/Rs than presently available with linear systems.

SMPS units will also be inherently easier to troubleshoot,since all measurement points are located in an inte grated unitand are not split between a control cabinet and T/R set, whichare often located a considerable distance apart.

Overall, there can be a much different approach to servic-ing and supporting SMPS units on precipitators. This willresult in significant operational benefits for ESP suppliers andusers.

Sourcing of Power SuppliesCurrent ESP power systems consist of three major com-

ponents:1. Voltage control and associated communications hard-

ware and software.2. Control cabinet including SCR, f iring circuit, switch-

gear, metering, etc.3. T/R setThe engineering and manufacturing expertise required for

each of these components is some what diverse, and few sup-pliers excel in all three. Therefore, many ESP suppliers ha vechosen to acquire these parts from dif ferent vendors.Furthermore, in order to of fer exclusive content in their prod-ucts, many ESP OEMs ha ve developed their own proprietaryvoltage controls and communications products.

As a result of these factors, ESP suppliers have had to per-form as the o verall electrical systems inte grator, and resolveincompatibilities between the v arious vendors’ products. Dueto the maturing of these technologies, ho wever, the systemsintegration problems are not as dif ficult as the y were in thepast, but some still e xist today, causing delays and increasedproject costs.

The introduction of SMPS to the ESP mark et will changethe options a vailable to ESP suppliers and users for b uyingpower supplies and communications softw are. The SMPSunits introduced to the mark et to date are of fered only as acomplete package, including the voltage control.

Those ESP OEMs that no w have their o wn proprietaryvoltage control and communications products will soon have achoice to make. It is technically feasible to modify a 60 hz.Voltage control to w ork with SMPS units. NWL has alreadyper formed the ef fort to adapt the NWL 60 hz. control forSMPS use.

However, there will be a signif icant development effortrequired to design an interf ace and ensure compatibilitybetween an alternate voltage control and the rest of the po wersupply system.

Due to other industry trends, there seems to be a mo ve bythe major ESP OEMs away from having their own proprietaryvoltage control and communications systems. The introductionof SMPS units may accelerate such mo ves, due to the ef fort




ET_7_05 9/16/05 3:50 PM

Medium Voltage Cable Terminationsand Splices to 35 kV

For information about theseproducts contact 800 422 6872

or [email protected]

North America’s leading manufacturer offers you an Alternative...

• Suitable for Shielded Power and URD Cable Systems

• Time Proven heat shrinkable technology provides relable sealing and long service life

• Qualifed to IEEE 48 and 404


After deregulation of the electricitymarket in the mid 1990s, a number ofcompanies mushroomed into e xistencetrying their hand at matching b uyers andsuppliers of electricity.

However, by 1999 man y had goneunder, and those fe w that survived havelearned that gathering timely informationis key to not only their success, b ut thatof their customers.

“People want thebest price for electricity,”says Phil Adams, ChiefOperating Officer ofWorld Energy. “We arethe product of dere gula-tion.”

What began origi-nally as Oceanside in1996, they metamor-phosed into WorldEnergy in 1999, with agoal of le veraging thepower of the Internet tostreamline and simplifythe complexities of buy-ing and selling po wer inderegulated markets.

“We avoid theuncertainty of the paperprocess,” says Mr .Adams, “by using thereverse auction platform,we are able to pro videthe lowest price disco v-ery engine for ener gyprocurement.”

One example of thecost savings that ha vebeen achieved using the World Energyauction platform is the General ServicesAdministration’s Energy Center ofExcellence, which sa ved approximately$24 million o ver a three-year periodwhen suppliers competed to meet thecenter’s energy needs.

“By using the online system, the bid-ding is quick er and more certain. Bydoing a lot of up-front w ork, it is ef fi-cient, and the polling of the utility data isformatted in such a w ay so the informa-

tion is readily available.”In areas where retail competition is

now a reality, consumers have the poten-tial to save large sums of money in ener-gy costs through open choice. Lar gecommercial and industrial companiesand government agencies spend an y-where from a fe w hundred thousand toseveral million dollars each year on ener-

gy, and whenpurchased insuch largequantities, asmall changein price canhave a tremen-dous impacton annualprofit margins,often makingthe differencebetween profitand loss.A n o t h e radvantage ofusing theInternet is thelevel ofaccountabilityand trans-parency forthose using it,as they areable to see theentire processin action fromstart to f inishas it happens.“Au toma ted

stop times and time stamps ensure theintegrity of auction e vents and createaudit trails and a self-documenting capa-bility,” points out Mr. Adams. “The newthing we are starting to see is that bigcompanies have to prove they are respon-sible to their shareholders. With thesefeatures, we can provide that integrity.”

Using World Energy’s exchange,qualified electricity and gas supplierscompete against one another to securethe buyer’s business using a structured

and transparent reverse auction platform.Throughout the auction, natural mark et-place competition push prices progres-sively lower, with winning bidders agree-ing to pro vide the po wer needed at theprice determined by the auction’s results.

The flexibility of the e xchange issuch that 9, 12 or 18 contracts can be onauction at any one time, and the y can beoffered collectively or individually.

“For example, we ha ve hadMassachusetts and Texas loads togetherand separately, with various term lengthsand bundles,” says Mr. Adams.

World Energy is one of the fe wnationwide energy procurement compa-nies, boasting a more than 65 per centmarket share in the United States.Running almost 800 auctions in the past18 months, some 11 billion kWh of elec-tricity and 34 billion cubic feet of naturalgas have been procured for some ofNorth America’s largest and most sophis-ticated commercial, industrial and go v-ernment customers, lik e Costco, F ordMotor Company and the United Nations.

The typical customer sa vings in the5-30 per cent range.

“We don’t have a crystal ball, but wecan tell the customer when the prices arebelow the industry’ s benchmark,” saysMr. Adams. “If our client has a contractthat expires in May and the price isfavourable in October, we will encouragethem to rene w immediately. It is all amatter of anticipating the trends.”

With the ever increasing price of oil,the cost of ener gy has come to the fore-front of everyone’s minds.

“It looks like $70 and over for a bar-rel of oil - that’s a 61 per cent increase, ”says Mr. Adams. “That has got e very-one’s attention, and the restaurants,hotels and big box retailers are all notic-ing that their electricity bills ha ve goneup 40 per cent.

“People are just beginning to under-stand that deregulation is good, and the ywant the best price. The real nuance ishow to do it? That’s where we come in.”

ET

INDUSTRY NEWS

A SUCCESS STORY FROM DEREGULATION:WORLD ENERGY

By Don Horne

“Automated stop timesand time stamps ensurethe integrity of auctionevents and create audittrails and a self-docu-menting capability,”points out Mr. Adams.“The new thing we arestarting to see is thatbig companies have toprove they are respon-sible to their sharehold-ers. With these fea-tures, we can providethat integrity.”

ET_7_05 9/16/05 3:52 PM

mode power supply technology willeventually become the standard solutionfor ESP power supplies.

Participants at all le vels of the ESPmarket—new system suppliers, reb uildand repair firms, consultants, and users—should begin the process of understand-ing the impact these ne w power supplieswill have on their businesses. ET

REFERENCES1. Boyle, P., Paradiso, G., Thelen, P.,

“Performance Improvements From Useof Low Ripple Three-Phase PowerSupply for Electrostatic Precipitator ,”Proceedings of the American PowerConference, Vol. 61-1, Illinois Instituteof Technology, Chicago, IL. 1999

2. Reyes, R., Wallgren, B.,Wramdemark, A., “A Novel and VersatileSwitched Mode Po wer Supply F orESPs,” Proceedings of the InternationalConference-Electrostatic Precipitators,September 1998, Kyongju, Korea

3. Chen, Y., “The HVDC CurrentSource For ESP,” Proceedings of theInternational Conference-ElectrostaticPrecipitators, September 1998, Kyongju,Korea


Appendix BTable 1

Test Conditions and DataPlant Miller Tests

High Frequency Pulsing Power Supply


ET_7_05 9/16/05 3:52 PM


Protecting existing and new commu-nications links between control centers,substations, and generating plants fromcyber attacks is crucial to maintainingservice.

Instrumentation and control systemsoperate and monitor critical infrastruc-ture components. Today there are realthreats of malicious physical and c yberattacks on these systems, making theirprotection and security more importantthan ever.

An attacker, for example, can secret-ly monitor sensitive information on datalinks, or send messages to operate equip-ment or change settings. Attackers couldsteal data that would identify biggest cus-tomers, usage patterns, deli vered powerquality, and production rates, or to iden-tify physical targets. Worse yet, imaginethe problems they could cause by com-manding breakers to open or close,machines to stop or start, and gates orvalves to open or close.

The Federal Ener gy RegulatoryCommission and electric utilities areconcerned about attacks on the systemsthat monitor and control the electricalpower grid. Schweitzer EngineeringLaboratories SEL-3021 SerialEncrypting Transceiver is one suchdevice that protects critical infrastructurefrom attack, using encryption and sessionauthentication, and locking out unautho-rized control.

The transceiver has lo w latency,making it ideal for systems that require

time-critical communication, such asSCADA, metering, and protection. This“bump-in-the-wire” solution allows easyinstallation in existing links.

The SEL-3021 is designed to meetFederal Information Processing Standard(FIPS) 140-2 security requirements,meeting or surpassing utility and indus-trial environmental requirements. In fact,the United Telecom Council selected theSEL-3021 as the UTC Telecom 2004Product of the Year. ET

INDUSTRY NEWS

SERIAL ENCRYPTING TRANSCEIVERS SECUREVITAL COMMUNICATIONS TO PROTECT

AGAINST CYBER ATTACKS

Transceivers like these keep intruders locked out of a secure system.

ET_7_05 9/16/05 3:52 PM

NERC is moving quickly to transi-tion to become the Electric ReliabilityOrganization (ERO) envisioned by TheEnergy Policy Act of 2005, culminatingnearly 10 years of work toward this goal.Many tasks, such as the formation of anindependent board and inclusi ve stake-holder representation, have already beenaccomplished.

NERC plans to f ile comments inresponse to the notice of proposed rule-making (NOPR) the Federal Ener gyRegulatory Commission (FERC) issuedon September 1, which proposes criteriafor the establishment of the ERO and theprocedures under which it will operate(Docket No. RM05-30). Comments aredue 30 days after the NOPR is publishedin The Federal Register.

As soon as FERC establishes thefinal requirements of the ER O, NERCwill file an application to become theERO. Under the le gislation, FERC mustfinalize a rulemaking by early February ,within 180 days of the date the presidentsigned the bill into law.

In preparing its application tobecome the ER O, NERC must addressseveral key issues, such as membership,funding, governance, functions andresponsibilities, relationships withCanadian regulatory authorities, and therole of the NERC regions. NERC is seek-ing the active involvement and support ofstakeholders, and is w orking with astakeholder group called the Post-Legislation Steering Committee (PLSC)to assist in coordinating the transition.

The PLSC has been in e xistence fortwo years in anticipation of le gislationand is being reacti vated to continue thework now that legislation has become areality.

ABB establishes Katrina hotlineThe ABB North America Hurricane

Katrina Task Force announced today thatABB has established a special toll-freephone number exclusively for customersdealing with the storm’s aftermath.

ABB customers should call 1-877-

511-4222. ABB Help Desk emplo yeesare available from 8:00 a.m. to 6:00 p.m.EDT to direct customers to the appropri-ate ABB function or contact for assis-tance. After-hour calls will be handledby an outside service, which will use thesame contact sheet and call proceduresas the Help Desk employees.

ABB customers in Me xico, or any-where outside the United States, dealingwith the storm’s fallout can call 1-440-585-8810 and receive the same support.

Quesnelle appoint to OEBKen Quesnelle has become a full-

time member of the Ontario Ener gyBoard.

"Ken's extensive insight and experi-

ence in the electricity distrib ution sectorwill be of great v alue to the OEB," saysBoard Chair, Howard I. Wetston.

Mr. Quesnelle most recently serv edas Vice-President of Woodstock HydroServices and is the past-chair of theElectricity Distributors Association(EDA). He has also serv ed as Vice-President of Info Energy, member of theIndustry Advisory Council to theElectrical Safety Authority and as amember of man y electricity industryrestructuring task forces.

Mr. Quesnelle has serv ed on theEDA's Western District Board ofDirectors since 1995, f illing the role ofChair since 2001. His appointment is fora two-year term.

Electricity Today Issue 7, 2005

NEWS BRIEFS

NERC PREPARES FOR FERC RULE ON ELECTRICRELIABILITY ORGANIZATION

ET_7_05 9/16/05 3:53 PM

THE WUNPEECE DUCT SPACERTHE WUNPEECE DUCT SPACERONE PIECE DOES IT ALL!ONE PIECE DOES IT ALL!

INTRODUCTIONPower cable diagnostics are the

most advanced ‘tool’ in the ‘toolbox’ ofshielded power cable tests. One of themost common questions regarding cabletesting tools is ho w the test will predictthe remaining life of the cable. To beclear, a test that will predict remainingcable life does not e xist. However, thereare cable tests which can help fulfill test-ing requirements, dramatically impro vereliability, and save the cable owner sig-nificant cable replacement costs.

Although standards ha ve offeredguidance, cable o wners and testproviders are lar gely on their o wn todetermine the best cable test tool for theapplication. One of the foremost imped-iments to the decision making process isthe lack of clear , credible information.This short paper attempts to pro videobjective guidance on ho w to best usethe power cable testing toolbox.

The paper then gi ves recommenda-tions on which test is the most ef fectivefor different applications and supportsthis assertion with case studies.

THE TOOLBOX(4)There are 3 major cate gories of

cable tests: high potential withstand,general condition assessment, and partialdischarge (PD).

Each of these tests serves a differentapplication. By comparing the strengthsand weaknesses of each type of test inpractical field application, decision mak-ers can quickly determine the testmethod best suited for their application.In the following section this paper liststhe positive and ne gative attributes foreach of the three cate gories in light ofrecently published IEEE standards andexclusive large scale f ield studies andcomparison tests.

Although these tests are fundamen-tally different and cannot be readilycompared, some clear distinctions can bemade on the basis of practical field appli-cation. A table summarizing the compar-ison can be found at the end of the fol-lowing attribute lists.

High Potential Withstand (HIPOT)The HIPOT withstand test is the

simplest tool in the toolbox. The HIPOTwithstand test can be divided by voltagesource into three cate gories: DirectVoltage (DC), Very Low Frequency(VLF), and Power Frequency. There arepositive and ne gative attributes whichpertain to all three types of voltage.

HIPOT (IN GENERAL)Pros:

• requires simple and relati velyinexpensive equipment

• uses very simple procedures• does not require a trained analyst

to interpret resultsCons:

• does not monitor the ef fect of thetest on the cable during the voltage appli-cation

• relies on ‘pass or fail’, thus expos-ing it to the ‘destructive test’ label

• is blind to certain types of defects• weakens all defects simultaneous-

ly, but fails only one at a time• can initiate test/f ail/repair/test

cycles which are costly and onerous

DC HIPOTPros:

• has a long history of use• uses the most portable, with the

lowest power requirements• is a good HIPO T for conducti ve

type defects (water in laminar cables)Cons:

• has been demonstrated to inducespace charge which aggravates defects inaged extruded cable long after the test’ sconclusion

• is blind to high impedance defectssuch as voids and cuts

• does not replicate service condi-tions and cannot be compared to f actorytests

VLF HIPOTPros:

• is a portable source with relativelylow power requirement

• is a good HIPO T for conducti vetype defect and high impedance defects


POWER CABLE DIAGNOSTICSFIELD APPLICATION AND

CASE STUDIESBy Benjamin T. Lanz, Senior Application Engineer, IMCORP, USA


ET_7_05 9/16/05 3:54 PM

• does not induce as much space char ge as DC whichlessens the aggravation of defects in aged extruded cable

• causes some defects to grow rapidly resulting in short-er test timeCons:

• has been demonstrated to aggra vate defects in agedcable without failing

• does not replicate service conditions and cannot bedirectly compared to factory tests

• is not recommended for aged cable with multipledefects

• does not replicate normal stress distrib ution in insula-tion with wet regions

POWER FREQUENCY HIPOTPros:

• is good for conductive and high impedance defects• does not induce space char ge which minimizes the

aggravation of defects in aged extruded cable• replicates steady state service conditions and f actory

HIPOT testsCons:

• uses the largest source, is the most e xpensive, and hasthe highest power requirement of all the HIPOT tests

• grows certain defects slower than VLF HIPOT testGeneral Condition Assessment (GCA)General Condition Assessment is a term which includes

a long list of cable diagnostic test methods and technologies,all of which determine the overall health of the cable insula-tion.

These tests are one of the more sophisticated diagnostictools in the cable testing toolbox. To name a few of these test-ing techniques, the following list is provided:

• Dissipation Factor/Tangent Delta/Power Factoro 50/60 Hzo 0.1Hz (VLF)• Dielectric Spectroscopy (Time and Frequency Domain)• Depolarization -Return Voltage (Recovery Voltage)• Depolarization -Relaxation Current (Isothermal

Relaxation Current)• Leakage Current• Total Harmonic DistortionEach of the technologies has its own advantages and dis-

advantages which are beyond the scope of this paper . Thereare, however, generalizations which can be made so reason-able comparison to other cable testing techniques can bemade.

GCA DIAGNOSTIC TESTSPros:

• are considered to be nondestructive because most testsare conducted at modest v oltage levels with a short dwelltime

• monitor the overall condition of the cable insulation• are effective in detecting and assessing conduction type

defects• provide information in the form of three cate gories-

critically aged, moderately aged and like newCons:


LaPrairie has been providing theelectrical industry with SadtemInstrument Transfomers formore than 15 years. Sadtem IT’s are designed usingDry Type Insulation, thereforeallowing us to offer compact,lightweight designs.We offer a comprehensive lineof 5-46kV Primary MeteringUnits that comply to CSA orANSI standard. Whether your requirement isfor a Pole Mount, Padmount orVault Mount application, wehave the assembly for you.For any of your PrimaryMetering requirements,LaPrairie is your answer!

LaPrairie Inc.ISO9001-20001206 Ringwell Dr.Newmarket, ON L3Y 8V9

Ph# (905) 830-9975 Ext.#23Fax# (905) 830-9989 e-mail: [email protected]

For further information please contact:

“Primar“Primary Meteringy MeteringAt Its Best”At Its Best”

DIAGNOSTICSFrom page 42


ET_7_05 9/16/05 3:55 PM

a lightbulb?a lightbulb?

how manycalgarianscalgariansdoes it take to screw indoes it take to screw in

ELECTRICAL LEARNING EXPOMA

XB

ELLA

REN

AC

ALG

AR

Y

YOU CAN FIND OUT! NOVEMBER 17, 2005

November 17, 2005 Maxbell ArenaS H O W T I M E 1 0 : 3 0 - 6 : 3 0 / S E M I N A R 1 0 : 0 0 - 6 : 0 0

Alberta Electrical League’s 11th Annual EXPOThe Electrical Learning EXPO provides a forum for electrical com-panies to display, demonstrate and introduce new products, servic-es and innovations to a wide range of consumers, distributors,manufacturers utilities, etc..

The EXPO offers various learning sessions led by a variety of speak-ers and industry specialist, who discuss the growing changes in theelectrical industry, and the direction in which such changes areheaded.

Give your company the leading edge by doing a product demon-stration and or maximize your company visibility by doing a pres-entation. Don’t miss out on this business enhancing opportunityand sign up for your exhibit space today!

For more information on exhibitingor attending please contact:

Alberta Electric LeaguePhone: (403) 514-3085

Tool Free: (800) 642-5508Fax: (403) 514-6169

E-mail: [email protected]

• are not comparable to f actory teststandards

• require a trained analyst to inter-pret measurement

• require equipment that is costlywhen compared to HIPOT equipment

PARTIAL DISCHARGE DIAGNOSTIC TESTSPD diagnostics are the most ef fec-

tive tool but also the most sophisticatedtool in the cable testing toolbox. Thereare 3 types of PD diagnostics: on-linePD testing, on-line PD monitoring, andoff-line PD testing. Continuous on-linePD monitoring technology for cables,although promising, requires furtherdevelopment to mak e it attracti ve. Forthis reason, this paper focuses on on-lineand off-line PD testing. PD diagnosticsare the only cable tests which can locatedefects in shielded power cable.

PD Diagnostic Test (in general)Pros:

• is a nondestructive test• is the only testing technology

which can detect and locate high imped-ance defects such as void, cuts, electri-cal trees, and tracking

• can be performed on-line in limit-ed applications

• is ef fective at locating defects inmixed dielectric systemsCons:

• is limited to cables with a continu-ous metallic shield (time or frequenc ydomain tests)

• requires a trained analyst to inter-pret measurements

• cannot detect or locate conductiontype defects

• becomes comple x, onerous, andloses accuracy in branched netw orkapplications

ON-LINE PD DIAGNOSTIC TESTPros:

• can be performed without switch-ing the circuit out of service

• can detect and locate some acces-sory defects and a few cable defects

• does not require an e xternal volt-age sourceCons:

• will only detect 3% or less of cableinsulation defects in extruded cable

• is not a calibrated test; therefore,the test results are not objective

• cannot be compared to f actorytests or IEEE standards

• has not been determined to beeffective by statistically signif icant datacorrelating results to actual cable systemdefects or failures

69% accurate in large scale studies• have results which are highly tem-

perature dependent in extruded cables• are blind to high impedance

defects such as cuts, v oids, and partialdischarge with less than 10,000pC

• cannot detect singular defects inextruded insulation and require hundredsof large water trees to be present to causethe smallest indication

• are not an ef fective test for mixeddielectric or newly installed cable

• require that prior signature files ofvarious cable types be recorded andcompared

• have a poor economic trade-of f inthat there is no defect location (a cablefound to be ‘bad’ needs to be entirelyreplaced)

• have only been shown to be 50 to




ET_7_05 9/16/05 3:57 PM

• requires access to the cable e veryfew hundred feet depending on the cableconstruction

• requires that manholes be pumpedto access cable

• is only of fered as a service andequipment cannot be purchased

• do not pro vide an onsite report ofthe test results

• cannot be applied to long directlyburied cables

OFF-LINE PD DIAGNOSTIC TESTPros:

• replicates calibrated f actory base-line tests and can be readily compared

• replicates steady state and transientoperating conditions

• can indirectly locate lar ge watertrees associated with electrical trees

• locates all defect sites in one testfrom one end of the cable

• is pro ven to be highly accuratewith statistically signif icant correlationstudies (85-95%)

• detects and locates defects in thecable system during voltage application

• is ef fective with mix ed dielectriccables

• can test up to 1 to 3 miles of cabledepending on the cable construction

• is backed by IEEE 400 as the mosteffective test

• offers equipment that can be pur-chased and used by utility personnel

• provides onsite report of the testresultsCons:

• requires that the cable be switchedout of service for test to be performed

• requires equipment that is e xpen-sive when compared to HIPO T equip-ment

Table –Summary of ComparisonsKey

GCA =General Condition AssessmentPD = Partial Discharge

HIPOT= High Potential WithstandVLF = Very Low Frequency voltage source

(0.1Hz)PWRFRQ = Power Frequency (20-80Hz)

Which is the right test? - An exampleWhich test is the ‘right test’ com-

pletely depends on the application athand. Take an example from the medicalworld. One could use a blood pressuretest to assess the cardiovascular health ofa patient or use a specialized stress testwith an Electrocardiogram (EKG). Both

tests will provide information about thepatient’s cardiovascular health but eithertest could be the wrong test in dif ferentapplications. The stress test, as it is com-monly known, will provide a much bet-ter understanding of the cardio vascularsystem’s health and expected future per-formance than the blood pressure test.However, the stress test requires the pur-chase of specialized equipment and thetraining of a skilled technician. If themeasurement is out of the norm, the

blood pressure test clearly indicates adisposition to cardio vascular problems,yet the simple test lacks the ability to beas predictive as the stress test. Althoughthe blood pressure test has its limitations,it is a useful test because it uses widelyavailable inexpensive equipment anddoes not require a highly skilled techni-cian. The moral of this e xample is thatthe selection of the right ‘tool’ from thecable testing ‘toolbox’ depends on the:




ET_7_05 9/16/05 3:57 PM

• expectation of the test results• operator’s skill level• availability of equipmentThe table abo ve uses these three

attributes to compare all the tests in thecable testing toolbox.

Discussion of ComparisonsEach of the tests in the cable testing

toolbox has strengths and weaknesses.Each test must be selected according tothe application. The following para-graphs are some typical scenarios whichutilities and test providers face while try-ing to select the right test.

If a utility or a test pro vider is look-ing for a direct replacement for the DCHIPOT withstand test that their line crewcan operate without signif icant training,the VLF HIPOT is a good choice.However, the user of such equipmentshould not e xpect to perform predicti vediagnostics with an instrument which isdesigned to perform a simple ‘go, no go’

test.If a utility or a test pro vider would

like to detect conductive type defects andpredict future performance in paper oilinsulated cable, a GCA type test w ouldbe appropriate.

However, if the cable has a lack ofoil, an of f-line PD test w ould be moreappropriate. Both paper oil and e xtrudedinsulation cables that tests ‘good’ with aGCA test can still f ail due to a highimpedance defect such as a crack, v oid,or electrical tree associated with PD. Inthis case, an appropriate, all compassingtest would be a combination of the GCAand off-line PD diagnostics.

CASE STUDIESCase 1: VLF HIPOT Withstand Teston Aged XPLE Cable

THE WRONG TOOL FOR THE JOBAccording to the IEEE 400.2 stan-

dard “Guide for Field Testing of ShieldedPower Cable Systems Using Very LowFrequency (VLF)” Table 2: Usefulness ofVLF testing methods for selected cableand/or insulation conditions, the VLFHIPOT withstand test should not be used

on aged e xtruded cable systems withmultiple defects. This of course presentsa problem to the cable o wner. Whoknows which aged cables ha ve multipledefects? The following case studydemonstrates why the VLF HIPOT with-stand test is a poor choice for testingaged extruded cable systems.

In February of 2004 a utility hadsome doubt whether a simple VLFHIPOT withstand test would be an effec-tive test for sorting out ‘good’ and ‘bad’aged extruded cables. The utility request-ed a 60Hz, off-line PD diagnostic test tobe performed on a 3 phase, 1400ft long,25kV, XPLE cable. This test is a superiorbase-line test for comparisons because itis directly comparable to a cable f actorytest. According to the PD test results,multiple defects in the cable insulationwere present in the cable. The cable wascondemned for replacement on the basisof the utility’ s criteria. There were toomany defects to repair , and it w ould bemore economical to replace the cable.The utility requested a VLF HIPOT with-stand test to be performed on the cable. If


MTE Series A dV/dT filters aredesigned to protect AC motors fromthe destructive effects of peakvoltages facilitated by long cable runsbetween the inverter and motor.Depending on the turn-ontime (risetime) of thepower semiconductorsused in the inverter, andthe size of the motor;cable lengths as short as eight feet

can result in motor peak voltages that exceed the rating ofthe motor’s insulation system.

The MTE dV/dT filter is guaranteed to meet its maximumpeak motor voltage specification (150% of bus voltage)with up to 1,000 feet of cable between the filter and themotor. It is also rated for a maximum dV/dT of 200 voltsper microsecond.

New MTE Series A sine-wave filters provide a sine-waveoutput voltage when driven from PWM inverters withswitching frequencies from 2 kHz to 8 kHz. For driveapplications, these filters eliminate the problem of motorinsulation failures and reduce electromagnetic interference

by eliminating the high dV/dTassociated with inverter outputwaveforms in applications where thedistance between the motor andinverter is up to 15,000 feet.

For alternative energyapplications, such as winddriven generators, where an inverter is used to

return power to the utility distributionsystem through a step-up transformer,these filters meet the requirements of IEEE-519 and permit the use of standard transformers.

Harmonic voltage distortion feeding a transformer at fullload and at 60Hz is 5% maximum. Harmonic voltagedistortion feeding a motor at full load and at 60Hz is 5% typical.

The Sine wave filter is available in 440 – 480 VAC and 550– 600 VAC for motors from 1.5 to 600 HP (440 VAC) and1.5 to 700 HP (550 VAC). For details and our performanceGUARANTEE call:

Stop destructive PWM voltage spikes in motors with MTE dV/dT and Sine Wave Filters

CABLE RUNS TO1,000 Ft.

Series AdV/dT Filters

CABLE RUNS TO15,000 Ft.

Series ASine Wave Filters

800-455-4683

www.mtecorp.comwww.globalspec.com



ET_7_05 9/16/05 3:58 PM


any of the condemned phases survi ved,the PD test w ould be repeated on thosephases to provide comparison data. TheVLF test was performed according to theIEEE 400.2 guidelines for a 25kV classcable, 23kVrms with 0.1Hz VLF for 30minutes. According to the standard, if thecable fails within the 30 minutes HIPO Ttest the cable is ‘bad’ and if it survi vesthe cable is ‘good. ’ If the cable f ails thestandard recommends the repair be madeand the 30 minute test repeated. Eachphase was VLF HIPOT withstand testedindependently. The results were:

• A phase survived 31 minutes.• B phase failed after 20 minutes.• C phase f ailed after a prolonged

dwell time of 37 minutes. (The operatorof the test lost track of time and forgot tostop the HIPOT at 30 minutes.)

In this case study , B phase w as theonly phase that w as found to be ‘bad’according to the IEEE standards. A phasepassed and C phase would have passed if

the test did not accidentally run past the30 minute time limit.

Immediately after the VLF HIPOTwithstand test w as concluded, the PDdiagnostic test w as repeated. The only

cable which could be retested w as Aphase, since the other tw o phases failedduring the test. A phase w as known tohave 3 cable PD (defect) sites (see Figure



Figure 1A Phase Off-line 60Hz Partial Discharge Test Results Before and After VLF HIPOTWithstand Test

ET_7_05 9/16/05 3:58 PM

2 for an example) from the previous test (top graph of Figure1). In addition to these 3 sites, the second PD test sho wed 7new sites after the 30 minute VLF HIPOT test (lower graph ofFigure 1). The three original PD sites, circled, appeared atlower test voltage after the VLF HIPOT. This is indicative thatthe defects have worsened without failing. Both PD tests werehighly sensitive despite high levels of ambient noise and werecalibrated to be accurate to 10 picoCoulombs (pC) accordingto IEC and AEIC calibration standards.

This case study demonstrates ho w using the wrong tool

for the job can mak e a poor situation w orse, unbeknownst tothe test operator . In this case, according to the IEEE 400.2standard, the operator w ould have to go through the c ycle oftest, fault locate, repair , and retest se veral times before thecable could be declared defect free. This is a time consuming,expensive and onerous proposition on aged cables with multi-ple defects. The VLF HIPOT withstand might be a greatreplacement for the DC HIPO T withstand test, which can beoperated by line crews, but, users of this test should not expectto diagnose aged extruded cables and predict their future per-formance.

Case 2: On-line PD Diagnostic Test Applied to AgedExtruded Cable.

THE WRONG TOOL FOR THE JOBIn January of 2005, a utility w as interested in comparing

Off-line PD Diagnostics with On-line PD Diagnostics. The on-line test had the ob vious advantage of not ha ving to take theline out of service. The utility first requested a 60Hz, Off-linePD Diagnostic test performed on a 1200ft long, 15kV class,extruded cable. The test w as calibrated to 5pC sensiti vityaccording to AEIC and IEC standards. The results of the testshowed that there were se veral defects in the cable insulationand one defect in a joint. The defect in the joint w as found tohave PD at only 80% of the operating v oltage stress le vel.After the off-line PD test, the cable was put back in service.

The On-line PD Diagnostic test technician applied thesensors at the terminations of the cable. Using customdesigned sensors, a spectrum analyzer, and a special noise f il-tering process the mobile on-line testing unit w as not able todetect even the largest levels of PD in the cable. A report stat-ed that the cable was free from defect and should be left in ser-vice.

This case study demonstrates how using the wrong ‘tool’can give an inaccurate reading on a potentially imminent f ail-ure. Although the on-line PD Diagnostic test is v ery conve-nient, it has a number of limitations. These limitations are list-ed in the comparison section (abo ve) of this paper . This casestudy suggests that the on-line test did not ha ve enough sensi-tivity to detect the defects in this cable. Since on-line tests arenot calibrated in the f ield, as the of f-line tests are, the y areinherently susceptible to missing defects. Some of the seriouscable insulation defects found by the of f-line test w as notdetected by the on-line test because the PD sites were notactive at the operating voltage stress level. According to Figure4, 3% or less of all defects in e xtruded cable associated withPD are active at operating v oltage. One plausible reason that




Figure 2Typical Defect (PD Site) from the Utility’s Cable System

Figure 3Typical Defect (PD Site) from the Utility’s Cable System

ET_7_05 9/16/05 3:59 PM

geted for cable replacement, the f ailure rate of the cables w assteadily rising at an a verage rate of 4% per year . By incorpo-rating off-line PD testing to their replacement program theutility has been able to maintain the same cable replacementbudget but dramatically improve the reliability of their w orstperforming cables.

Since the utility has tak en action on the basis of the of f-line PD cable diagnostics, the utility has not only reversed theincreasing failure rate trend, the f ailure rates ha ve dropped

online test do not typically detect defects in e xtruded cableinsulation is that insulation defects typically last only minutesto hours under continuous PD conditions. This means the win-dow of opportunity for observing these defects with PD atoperation voltage is very short and not likely to be active dur-ing the few minutes of test duration.

The graph in Figure 4 is deri ved from 1,555 miles ofextruded cable and 960 PD sites located in the cable e xtrudedcable insulation.

On-line PD Diagnostic tests are useful if there is no w ayto take the cable system out of service. They are useful atdetecting some cable accessory (joints and terminations)defects; especially in short cables where the PD signals are notgreatly attenuated. As the case study implies an On-line PDDiagnostic test should not be e xpected to perform a compre-hensive test on a cable system.

Case 3: 60Hz PD Diagnostic Test Applied to Aged XPLECable

THE RIGHT TOOL FOR THE JOBIEEE 400 standard “Guide for Field Testing and

Evaluation of the Insulation of Shielded Po wer CableSystems” states the most effective test is the off-line PD diag-nostic test which replicated factory test conditions. Section 7.4states that “if the cable system can be tested in the field to showthat its partial discharge level is comparable with that obtainedin the factory tests on the cable and accessories, it is the mostconvincing evidence that the cable system is in e xcellent con-dition.” The following case study demonstrates ho w off-linePD diagnostic testing is the right tool for testing an aged XLPEcable system.

During the period from January 2001 to December 2004,a large utility in southern United States used 60Hz, Off-line PDdiagnostic testing to direct their cable replacement program.This program uses the results of the tests to determine whetherto repair, replace, or defer action on the utility’s directly buriedfeeder cable. The worst performing cables were 430 miles ofdirectly buried cable ranging in age from 20 to 35 years old.Prior to utilizing cable diagnostics, the utility w as using thecommon practice of replacing these cables on the basis of fail-ure history and vintage. Although millions of dollars were bud-



Figure 4(5)Cumulative Probability of pC Magnitude vs. PD Inception Voltage

Figure 5Number of Failures vs. Year


ET_7_05 9/16/05 4:00 PM

Owned & Managed by: Flagship Media Sponsor:

m0508PGI1pa59.indd 1 8/9/05 12:06:44 PM

ET_7_05 9/16/05 4:01 PM


The Department of Ener gy recentlyannounced the first two projects selectedunder the Department’ s new Fuel CellCoal-Based Systems program. The pro-jects will be conducted by tw o researchteams — one led by General ElectricHybrid Power Generations Systems andthe other by Siemens WestinghousePower Corporation — and they share thesame goal: to de velop the fuel cell tech-nology required for central po wer sta-tions to produce af fordable, efficient,environmentally friendly electricity fromcoal.

The new program leverages knowl-edge gained in DOE’s Solid State EnergyConversion Alliance (SECA), extendingcoal-based solid oxide fuel cell technolo-gy to large central power generation sta-tions.

The benefits of de veloping coal-based fuel cell systems are signif icant.Coal is a cost-ef fective energy sourceand, with approximately 250 years ofreserves, it is America’s most ab undantfossil fuel. Our impro ved ability to usethis domestic energy supply reduces ourdependence on foreign mark ets andincreases our ener gy security. In addi-tion, fuel cells are one of the most attrac-tive power generating technologies forthe future. Fuel cell systems do not relyupon combustion; consequently, theyproduce nearly none of the air pollutantsassociated with conventional power sys-tems.

Given these adv antages, advancesmade under the Fuel Cell Coal-BasedSystems program are expected to becomekey enabling technologies forFutureGen, a DOE demonstration ofadvanced power systems that emit near -zero emissions, ha ve double-today’selectric generating efficiency, co-producehydrogen, and capture and sequester car-bon dioxide.

The two teams will research, de vel-op, and demonstrate fuel cell technolo-gies that can support po wer generationsystems larger than 100 me gawattscapacity. Key system requirements to be

achieved include:- At least 50 per cent

overall efficiency in con vert-ing the ener gy contained incoal to grid electrical power.

- Capture of 90 percent ormore of the system’ s carbondioxide emissions.

- Cost of $400 per kilo-watt, exclusive of the coalgasification unit and carbondioxide separation subsys-tems.

Projects will be conduct-ed in three phases. DuringPhase I, the teams will focuson the design, cost analysis,fabrication, and testing oflarge-scale fuel cell stacksfueled by coal synthesis gas.Central to the Phase I ef fortwill be the resolution of tech-nical barriers with respect tothe manufacture and perfor-mance of lar ger-sized fuelcells.

Phases II and III willfocus on the f abrication ofaggregate fuel cell systemsand will culminate in proof-of-concept systems to be fieldtested for a minimum of 25,000 hours.These systems will be sited at existing orplanned coal gasification units, potential-ly at DOE’s FutureGen facility.

The Office of F ossil Energy’sNational Energy Technology Laboratorywill manage the Fuel Cell Coal-BasedSystems program and projects. The newprojects are described below:

• Solid Oxide Fuel Cell Coal BasedPower Systems — General ElectricHybrid Power Generation Systems willpartner with GE Ener gy, GE GlobalResearch, the Pacific Northwest NationalLaboratory, and the Uni versity of SouthCarolina to develop an integrated gasifi-cation fuel cell system that mer ges GE’sSECA-based solid oxide fuel cell, gasturbine, and coal gasif ication technolo-gies. The system design incorporates a

fuel cell/turbine hybrid as the mainpower generation unit. (DOE Phase Iaward: $7.5 million; Phase I duration: 36months).

• Coal Gas Fueled Solid Oxide FuelCell/Gas Turbine Hybrid Power Systemwith CO2 Separation — SiemensWestinghouse Power Corporation is part-nering with ConocoPhillips and AirProducts and Chemicals Inc. to de veloplarge-scale fuel cell systems based ontheir in-house gas turbine and SECA-modified tubular solid oxide fuel celltechnology. ConocoPhillips will pro videgasifier expertise, while the baselinedesign will incorporate an ion transportmembrane (ITM) oxygen separation unitfrom Air Products. (DOE Phase I a ward:$7.5 million; Phase I duration: 36months).

INDUSTRY NEWS

COAL-BASED FUEL CELLS: A GIANT LEAP FORFUEL CELL TECHNOLOGY

By David Anna, DOE National Energy Technology

A diagram of Siemens-Westinghouse 25 KW tubularsolid oxide fuel cell.

m0508PGI1pa59.indd 1 8/9/05 12:06:44 PM

ET_7_05 9/16/05 4:01 PM



FLIR SYSTEMSHawkIR SightGlass InfraredWindowsFLIR Systems Ltd.offers a completeline of HawkIR SightGlass InfraredWindows for Highand Medium Voltage

applications. The Infrared Sightglassis ULC approved for retrofit on enclo-sures. The IP65 viewing windowallows for safe and quick infraredscanning on indoor an outdoor MCC’sand Electrical cabinets.The gasketedSightglass lens comes in sizes 1” to4” and provides safety locks.

For further details contactFLIR Systems Ltd. 1-800-613-0507 X30 [email protected]

For information on Canadian 2004 Infrared TrainingCenter courses visit:www.infraredtraining.com

INFRARED IMAGING SOLUTIONS INC.

Canadian company that provides pre-dictive maintenance services to the util-ity, industrial and commercial marketsthroughout the Americas and the MiddleEast. InfraRed Imaging Solutions Inc.performs all inspections using certifiedthermographers and state of the arttechnology.This coupled with complete-ly unbiases reporting that can be tai-lored to meet your specific needs. 24hours emergency service.

For additional information, contact:

Tel: 416-529-4733

FLEX-CORECurrent

Transformers,Transducer &Accessories

Catalog 2011-170 page catalog fea-turing split-core and solid-core currenttransformers, potential transformers,electrical transducers for AC and DCapplications, signal conditioners, ana-log and digital panel and switchboardmeters, current relays/switches,shunts, shunt switches, shorting termi-nal blocks, multi-function power metersand accessories.

For more information contact:Flex-Core6625 McVey BlvdColumbus, OH 43235Tel: (614) 889-6152 Fax: (614) 876-8538 Email: [email protected]

! " #$ %#

& %'( )!

& * ' )!

& + )!

& ,+,) '$

& "%

&

-#' . '' #)

ACCESS CONTROL SALES3-Phase Monitors Catalogue

Time Mark Corporation’s expandedversion of the PM-310 specialty cata-logue contains information on the com-pany’s popular line of phase loss mon-itors. The 24 page book containsdescriptions, photos, dimensionaldrawings, specifications and typicalapplication diagrams for each device.Also included is a cross referencechart of available models, and answersto commonly asked questions.For more information, contact:Access Control Sales Ltd.Tel: (905) 564-1472Fax: (905) 564-3349Email: [email protected]

ET_7_05 9/16/05 4:04 PM


Surplus New - Never UsedERMCO Polemount Transformers 25 kVA, 12,470Y – 480Y GE Pad Mount Transformers 300 kVA, 12,470Y - 208YABB Bushings2900 Amp, 15 KV

Global Power Supply has a wide array ofexcess equipment available for sale, includ-ing motors, pumps, process controls,protective relays, switchgear, trans-formers, turbines... and other infra-structure equipment. We purchaseequipment and we find buyers for yourequipment. We have numerous require-ments from utilities and industrial plantsfor maintenance parts and expan-sion projects.

www.globalpwr.net805-683-3828 Ext. 310

IEEE C62.11 Tested

PDV-100 OPTIMA Distribution Class

Polymer Surge Arrester

• Improved Isolator Reliability Down to 1 AMP• ANSI Heavy-Duty Rated• New Universal Temporary Overvoltage• Field-tested Sealing System

POWER SYSTEMS,INC.

For more information contact:Hubbell Canada LP

Tel: 905-839-1138Fax: 905-831-6353

Email: [email protected] visit our website

http://www.HubbellPowerSystems.ca

Test CT’s and PT’sHigh performance at low cost. Model CT/PT is easy touse and super fast. Just connect and enter or retrievethe stored program. RCF and Phase will appear likemagic. Prints certificates and multi-point reports.

Special Features:• low cost• high accuracy• high speed, easy to use• high power, to 8000A• multi-point testing• meets ANSI and IEC• prints reports and certificates• all solid state

OPTIMIZED DEVICES, INC.220 Marble Ave. Pleasantville, New York 10570

Phone: 914 760 6100 Fax: 914 769 6102 Email: [email protected]


G.T. WOOD COMPANY LIMITEDHigh Voltage Electrical Contractor

- High Voltage Preventative Maintenance

- Transformer and Switchgear Repair- Infra Red Thermography

Inspections- Engineering Design Studies- Substation Design and Build- Harmonic Analysis/Power Factor

Correction- Cable Locates and Repairs- Sales of New and Pre-owned

Equipment- Supply of Temporary Substations- PCB Management24 HOUR EMERGENCY SERVICE

For additional information contact:Toll free:1-800-305-2036Tel: (905) 272-1696Fax: (905) 272-1425Email:[email protected]

PHOENIX MANUFACTURING LTD.Insulating Boots For Indoor And

Outdoor ApplicationsPhoenix manufactures a comprehen-

sive range of reusable, flexible, form fit-ting insulating boots for indoor and out-door applications for service voltagesup to 38 KV. Specialists in the ‘Fluidisedbed’ process to fully or partially encap-sulate copper and aluminium busbars,connections and components with highgrade dielectric strength epoxy andvinyl insulation in the voltage range 600volts to 38 KV.For more information contact:Phoenix Manufacturing Ltd.141 Fulton St.Milton, Ontario L9T 2J8Tel: (905) 878-2818 Fax: (905) 878-0051Email: [email protected]

ET_7_05 9/16/05 4:04 PM

NEHRWESS, INC.For all your conduitstandoff and meter

mounting needs

1408 E. State Rd. 250Brownstown, IN 47220

Tel: 812-358-4593 Fax: 812-358-4594Email: [email protected]

Hurst Communications

Rectangle


ADVERTISERS’ INDEX

Tp

ADVERTISER PAGE CONTACT INFO ELECTRICITYFORUM WEB PAGEAccess Control Sales 52 www.accesscontrolsales.com www.electricityforum.com/products/access.htmAlberta Electric League 44 www.elecleague.ab.caAPPrO 2005 16 www.appro.orgBell 9 www.businessonthego.ca/utilitiesComverge 5 www.comverge.comCandura Instruments 47 www.candura.comDSG Canusa 37 www.dsgcanusa.com www.electricityforum.com/products/dsg.htmThe Electricity Forum 30 www.electricityforum.comFlex-Core 52 www.flex-core.com www.electricityforum.com/products/flex.htmFLIR Systems Ltd. 25, 52, 56 www.flir.com www.electricityforum.com/products/flir.htmFluke Electronics 48 www.morethanjust.ca www.electricityforum.com/products/fluke.htmFortune Electric Co. Ltd. 29 www.fortune.com.tw www.electricityforum.com/products/fortune.htmGlobal Power Supply 53 www.globalpwr.netGT Wood Company Ltd. 53 www.gtwood.com www.electricityforum.com/products/gtwood.htmHigh Voltage Supply 49 www.HighVoltageSupply.com www.electricityforum.com/products/High_Voltage_Supply_(HVS).htmlHubbell Power Systems 2, 19, 53 [email protected] www.electricityforum.com/products/hubbell.htmIncon 27 www.incon.com www.electricityforum.com/products/incon.htmInfraRed Imaging Solutions 52 www.iristhermography.comIPC Resistors 11 www.ipc-resistors.comLaPrairie Inc. 43 www.laprairieinc.com www.electricityforum.com/products/laprairie.htmLineal 31 www.lineal.com Lizco Sales 52 www.lizcosales.com www.electricityforum.com/products/lizco.htmManitoba HVDC Research Centre 45 www.pscad.com www.electricityforum.com/products/manitoba_hvdc.htmManitoba Hydro 15 www.midasmetering.com www.electricityforum.com/newsprod/midas.htmlMorgan Schaffer 55 www.morganschaffer.com www.electricityforum.com/products/morgan.htmMTE Corporation 46 www.mtecorp.com www.electricityforum.com/products/mte.htmNehrwess 53 www.nehrwess.com www.electricityforum.com/products/nehrwess.htmNynas 7 www.nynas.comOptimized Devices Inc. 53 www.optdev.com www.electricityforum.com/products/optimized_devices.htmPhoenix Manufacturing Ltd. 53 www.phnxmfg.com www.electricityforum.com/products/phoenix.htmPower-Gen 50 www.poewer-gen.comPowerSystems World 35 www.powersystems.comSediver 13 www.sediver.frSynergy 21, 52 www.synergyinc.com www.electricityforum.com/products/SYNERGY_ENERGY_INC.htmlTamini Group 23 www.tamini.comUnderground Devices 41 www.udevices.com www.electricityforum.com/products/underground.htmVelcon Filters 33 www.velcon.com www.electricityforum.com/products/velcon_filters.htmThe VON Corporation 42 www.voncorp.com www.electricityforum.com/products/von.htmWire Services 39 www.wireservices.ca www.electricityforum.com/products/wire_services.htm

from 132 in 2001 to 30 in 2004 (a 77%reduction in failures). Figure 5 shows thedramatic change in f ailure rate after of f-line PD diagnostic tests were used todirect the cable replacement program.

This successful reduction in f ailurerate was enabled by using the right toolfor the application. The utility alsoreports that during this 3 year periodthere have been no f ailures in cableswhich had defects located, were repaired,and put back in service. With a 5 yeardeferment on the cost of replacementcapital alone, the cost benefit of this pro-ject is up to 4 times the initial cable test-ing and cable repair e xpense. Whetherthis case study is viewed from the aspectof reliability or economics, the 60Hz,Off-line PD Diagnostics test is clearly theright tool for the job.

CONCLUSIONThroughout this paper the tools from

the ‘toolbox’ of shielded po wer cabletests have been compared using IEEEstandards and e xclusive field case stud-ies. These studies are a sample of the case

studies that are a vailable but were notincluded in this paper for the sak e ofbrevity. The IEEE standards quoted andthe field studies presented pro vide evi-dence that making an informed decisionis imperative to achie ving the e xpectedreliability results with cable testing.

Throughout this paper the Of f-linePD Diagnostic test is sho wn to be themost effective test in the field and highlyrecommended by IEEE specif ications.The VLF HIPOT withstand test, althougha good replacement for the DC HIPO Tand is simple to use, needs to be usedwith caution because its limitations couldbe detrimental to the reliability of f ieldaged cable. While the on-line PD diag-nostic test is con venient, the test is onlyrecommended if a cable cannot be tak enout of service, and cable o wner canaccept the risk of the test missing cableinsulation defects. When the right test isused in the correct application and theexpectations are inline with reality , theresults are predictable.

BIBLIOGRAPHY1. IEEE 400, Guide for Field Testing

and Evaluation of the Insulation ofShielded Power Cable Systems,

Copyright 2001, IEEE.2. IEEE 400.2, Guide for Field

Testing of Shielded Po wer CableSystems Using Very Low Frequency(VLF), Copyright 2004, IEEE.

3. Mashikian, M., et al, “MediumVoltage Cable Testing by P artialDischarge Location.” Jicable 99, B6.3.

4. Lanz, Ben. “Cable TestComparisons 2004” December 10, 2004http://www.imcorptech.com/images/TestComparisons.

5. Mashikian, M “Pre ventiveMaintenance Testing of Shielded Po werCable Systems” IEEE Transactions onIndustry Applications, Vol. 38, NO.3,May/June 2002.

Benjamin Lanz r eceived a BS inElectrical Engineering fr om theUniversity of Connecticut in 1999. Since1997 he has work ed with IMCORP inStorrs, Connecticut and now holds theposition of Senior Application Engineer.He supported the de velopment of theIMCORP Cable Dia gnostic Technologyand has extensive field testing experiencein North America and Europe.


ET_7_05 9/16/05 4:06 PM

http://www.electricityforum.com/t&d/company.htm























Transformer problems?

This is where your Calisto should have been.

Don’t be left picking up the pieces. Catastrophic transformer failures can easily be preventedwith the Calisto on-line, dissolved hydrogen and water monitor.

Calisto monitors hydrogen production and detects transformer faults at their earliest stagesthereby reducing the number of unforeseen outages and preventing catastrophic failures.

• Designed to withstand the harshest environments

• Precise monitor readings are as reliable as lab results

• Increases the life expectancy of your transformer by monitoring the moisture content in the transformer oil

• Download specifications and software at www.morganschaffer.com

Calisto prevents catastrophic failures – one transformer at a time.

5110 de Courtrai AvenueMontreal, QuebecCanada H3W 1A7

Tel.: +514.739.1967Fax: +514.739.0434E-mail: [email protected]

Transformers - The Inside View VISIT OUR WEBSITE AT: www.morganschaffer.com

ET_7_05 9/16/05 4:06 PM

Go infrared...without going in the red!

ThermaCAM®

EX320ThermaCAM®

E25

TM

The Global Leader in Infrared Cameras

1 800 613 0507 www.flir.ca

The NEW ThermaCAM® EX320 provides the same superior resolution as

professional infrared cameras costing $50,000 or more with zoom and

professional measurement tools and report capabilities. The ThermaCAM® E25

is the most affordable and feature-rich thermography camera in the world.

CONTACT FLIR SYSTEMS FOR A FREE DEMO.

Filter capacitors scanned and

avoided an estimated

$500,000 outage.

Facility transformercaptured,$200,000 cost averted.

ThermaCAM® E25$12,995

New HighResolution

ThermaCAM®

EX320

CDN

Servicing Canada for 45 years

ET_7_05 9/16/05 4:06 PM

et 7 05 - electric todayforums, on-line magazines, industry news, job postings, electrical store,...

Documents