IEEE Transactions on Power Systems, Vol. PWRS-1, No. 1, February 1986

TURBINE FAST VALVING TO AID SYSTEM STABILITY:BENEFITS AND OTHER CONSIDERATIONS

Report of a panel discussion sponsored jointly by theIEEE Discrete Supplementary Controls Working Group* andthe ASME/IEEE Power Plant/Electrical System InteractionWorking Group

Panelists: L. Edwards (Chairman), J. D. Gregory,D. L. Osborn, J. H. Doudna, B. M. Pasternack, W. G. Thompson

ABSTRACT

Fast valving of steam turbines is a means ofrapid reduction of mechanical power on turbine-generators to prevent rotor angle instability when asignificant power-load imbalance is sensed. Fastvalving and early valve actuation (EVA) are equiva-lent terms. It is a potentially useful aid to systemstability that has not been widely applied for sev-eral reasons. The four "short note" reports describeimplementations of several methods of fast valvingand discuss application considerations.

INTRODUCTION

A panel discussion on turbine fast valving waspresented at the 1981 Joint Power Generation Confer-ence, with panelists from utilities who had imple-mented or considered implementation of fast valving.Subsequently, it was decided to condense the materialinto this paper. In the following sections, theindividual contributions of panelists are reported.

Introductory information on fast valving andother discrete controls is presented in a DiscreteSupplementary Controls Working Group paper [1]. Asdiscussed by the individual contributions and in thereferences, there are a number of considerations inapplying fast valving including:

o Momentary (temporary) fast valving versus sus-tained fast valving.

o Which valves (control, intercept, or both) toclose, and then reopen or reposition.

o Method of initiating fast valving.

o Effect on boiler or steam generator safety/reliefvalves.

o Effect of pressure and thermal stress on boiler/steam generator, moisture-separator, reheater,and turbine.

o Need for boiler/steam generator control adjust-ment.

o Generator tripping as an alternative to fastvalving. (The Discrete Supplementary ControlsWorking Group has sponsored two panel discussionson generator tripping and is preparing a similarpaper reporting the presentations.)

85 CM 50iS-4 A oaoer recomrmended and aoorovedby the hEE;£ Power Systemn Ericineerinn Committee ofthe l : owe-v £'na ineerinq bocietv for oresentattiona- the IEEE/cES 1985 Surnmner meetinrn Vancouver.

ans :knr., July 1X4 - 19 t985. Manuscriot s;eb-mittec- Decerlr'1 1984; made available for prinrt-inr: prik e6, d1985.

o Methods, such as single-pole switching or neutralresistors to reduce frequency of fast valving op-eration§. (The Discrete Supplementary ControlsWorking Group has also sponsored panel discus-sions on single-pole switching and is preparing apaper reporting the presentations.)

The presentation on sustained fast valving by LeoEdwards of Tennessee Valley Authority (TVA) is notreported in this paper. Rather, comprehensive de-scriptions of TVA's fast valving implementations areavailable in references 2, 3 and 4. References 2 and3 include field test results to verify the power decaycontrols installed on a large cross-compoundfossil-fired once-through unit. Reference 4 de-scribes sustained fast valving on a large nuclearunit. TVA is applying sustained fast valving on allnew large turbine-generators.

A. EARLY VALVE ACTUATION EXPERIENCE ON SOUTHERNELECTRIC SYSTEM (J. R. Woodal I and J. D. Gregory,Southern Company Services, Inc., Birmingham, AL)

The bulk of experience- with EVA on Southern'ssystem was gained from testing and operation of theGeorgia Power Company Plant Harllee Branch units. Thefirst investigation for application of EVA wasinitiated in 1966 for Plant Branch which has aninstalled capacity of approximately 1600 MW made up offour coal fired units, as summarized in Table A.l.

TABLE A.l.

PLANT BRANCH UNITS 1-4

Unit 1 Unit 2 Unit 3 Unit 4GeneratorManufacturer GERating (MVA) 320ExcitationResponseRatio 0.5

TurbineRating (MW) 250.8Pressure

(PSIG) 2400Temperature 1000/1OCSpeed/Load

Governor MHC

Steam GeneratorManufacturer B&WType Drum

GE422.4

GE W640 640

1.5 1.5 1.5

319 480.8 490

2400 3500 350000 1000/1000 1000/1000 1000/1000

MHC MHC EHC

Riley B&WDrum Once

Through

B&WOnce

Through

*Paper prepared by Discrete Supplementary Controlslorking Group. P.A.E. Rusche, Consumers Power Com-

pany; C. Chen, Stone & Webster Engineering Corpora-tion; and C. W. Taylor, Bonneville Power Administra-tion assisted in paper preparation.

0885-8950/86/0001-0143$01.00O©1986 IEEE

143

144

The plant is connected to the system with a230-kV ring bus substation and nine 230-kV lines.Initial transient stability studies demonstrated thatstable operation could not be maintained at the plantfor a bolted three-phase fault and breaker failure.Tripping units at the plant would maintain stableoperation for the postulated dual contingency; how-ever, this solution was considered regressive and notin the best interest of system operation. State ofthe art stability aids were simulated on subsequentstudies which range from high speed breaker failurerelaying schemes to braking resistors. None of theseaids would singularly satisfy the requirements.

During these initial studies, General ElectricCompany (G.E.) presented a paper [5] demonstratingimprovements which could be realized in unit trans-ient stability performance by momentarily closing theturbine steam valves. This concept was reviewed withG.E. and subsequently Westinghouse Electric Corpora-tion for application on the turbine-generators atPlant Branch. The manufacturers were requested tomake an indepth review of the turbine-generator de-signs for EVA operation and on-line field tests forthe units. Preliminary stability studies were madeemploying the manufacturer's calculated turbine powerdecay EVA characteristic for each unit. These re-sults showed considerable promise. Figures A.1 andA.2 illustrate, respectively, the calculated EVApower decay characteristics for G.E. Unit 2 andWestinghouse Unit 4.

The EVA was limited to operation of the interceptvalves. This eliminated complexities of unit speed/load and plant controls associated with the governorvalves. The turbine-generator manufacturers found nodesign limitations for EVA operation of the units andthe decision was made to apply EVA controls on PlantHarllee Branch Units 2, 3, and 4.

Steam Generator EVA Limitations

During review of the turbine energy control paper[5] with G.E., Southern was advised that one primaryobjective of the paper was for generating discussionby the steam generator manufacturers; however nonewas received. Southern then requested the domesticsteam generator manufacturers to review the equipmentdesign and advise if EVA operation of the PlantBranch units had any limitations on the equipmentdesign and operation. Several responses were notclear, although all indicated that periodic operationof EVA was acceptable. A review was made with eachmanufacturer to clear the issue. The basic criticalissue resolved into how soon the initial steam flowwould be essentially restored to the reheater. Theproblem was one of reheater tube heating when thesteam flow was shut-off and subsequently restoredduring an EVA operation. For these conditions thesteam generator time limitation, to essentiallyreestablish reheater steam flow, was obtained forboth recent and approved unit additions on theSouthern electric system. These are:

Babcock and Wilcox - Once-throughsuper critical

- 12 seconds

Lifting of the steam safety valves can be expected tooccur at high loads for an EVA operation. Only expe-rience will determine if these safety valves willproperly reseat. Also, the response of the steamgenerator controls should be investigated.

1.0

0.9

O 0.8

< 0.7-

8a 0.6

1r 0

0.4

0 0.2 0.4 0.6 0.8 1.0

TIME (SECONDS)

Figure A. 1 Calculated power decay suppliedby General Electric for HarlleeBranch Unit 2.

500

0a.

zI

ui

0 0.2 0.4 0.6 0.8 1.0

TIME (SECONDS)

Figure A.2 Calculated power decay suppliedby Westinghouse for HarlleeBranch Unit 4.

- Natural circu-lation

- 3 seconds

Natural circu- - 12 seconds

lation

CombustionEngineering

- Once-throughsuper critical

- 10 seconds

EVA Control Logic Design

There were unknowns involving a U.S. Patent asto fault detection and initiation of an EVA oper-ation. As a result, Southern had the turbine-gener-ator manufacturer provide the detecting initiatinglogic as an extension of a method employed to limitoverspeed due to load rejection. This logic dependson the comparison of mechanical input power and elec-tric output power. The former is measured by reheat

Foster Wheeler

Riley

2.0

steam bowl pressure; the latter is monitored by ahigh speed power relay or transducer depending on thedesign of the unit speed/load control system. Whenthe mechanical input power exceeds the output powerby a set amount the intercept valves close rapidly.

The setting applied to the EVA control scheme forthe tests and for initial operation allows the con-trols to function when the steam flow is above the 75percent rating and the electrical output drops below25 percent of rating. EVA operation is initiated assoon as sufficient imbalance exists between mechani-cal input power and electrical output power. Thereis no waiting period for acceleration of the unitrotor or for breaker failure to become evident. TheEVA controls are not actuated by external protectiverelays or other devices. Each turbine-generator soequipped is individually restrained when sufficientacceleration power occurs for any reason.

This system avoids the exposure to which a commonEVA logic control system is subjected at a multi-unitplant where failure or misoperation would encumberall units. The control system must be immune toelectrical power backswings experienced for EVA oper-ation initiated by a system fault. EVA operation maybe experienced when not required due to any threephase fault, heavy system swings, and by the neces-sity of-having the logic detection set with some overreach to assure coverage of the critical fault zone.The scheme is designed for detection of only a threephase fault. EVA operation is not required for faultsof less severity with a failed breaker.

Field Test - General Electric Turbine-Generator Units

Units 2 and 3 were tested jointly on-line underload by engineers of the utility and G.E. Dry runtests (no steam) were made on April 5, 1970. On-linefield tests were successfully made on April 10-11,1970. For brevity only the highlights of results fortests on Unit 2 will be described, since those onUnit 3 had no significant differences.

Three tests were made on Unitload conditions:

Test 1Test 2Test 3

2 for the following

- 360 MW, 5% Over Pressure- 373 MW, 5% Over Pressure- 260 MW

TIME -SECONDS

Fig. A.3 Plant Harilee Branch Unit 2 EVA Test 2.Initial load 373 MW, 5% over pressure.

145

The results of these tests have been reported earlier[6]. Each test was initiated by opening the gener-ator bus potential transformer 120 volt circuits sup-plying the EVA watt sensing relays in the MHC controlsystem. This simulated a three-phase fault in thecritical zone.

Figure A.3 shows Unit 2 test results. Both interceptvalves start to close approximately 0.07 secondsafter test initiation and go through full stroke inapproximately 0.11 seconds. After remaining closedfor approximately 1.23 seconds, they are reopenedthrough full stroke in approximately 2.6 seconds.This easily meets the 12 second time established bythe steam generator manufacturer. The generator ter-minal electrical megawatt power lags the closing ofthe intercept valves as power is removed from themachine rotor by valves closing. Inertial electricalpower oscillations are exhibited while the valves areclosed followed by the power recovery back to fullload as the valves are moved to the full open posi-tion. Good repeatability was obtained on each test.Figure A.4 shows comparison of the transient stabil-ity simulation results with the test results for thiscase. Good correlation is shown for the generatorterminal megawatt output reduction.

1.0 T -s; 1

0 .2 .4 .6 .8 1.0(12) (24) (36) (48) (60)

TIME -SECONDS(CYCLES)

Fig. A.4 Plant Harllee Branch Unit 2 EVA Test2--Comparison with Simulation. Oneper unit equals 373 MW.

There were no noticeable changes observed inunit operation (noise and vibration) by observers inthe plaht other than that producted by escaping steamwhen the reheat safety valves lifted during thetests. The safety valves lifted and reseated prop-erly during test 1 and 2, but were not observed tofunction on test 3 due to less reheater pressurebuild-up at the lower load.

146

Field Test - Westinghouse Turbine-Generator Unit

Unit 4 was tested on-line under load jointly byengineers of the utility and Westinghouse on June 6,1970. Several tests were made, with three tests forturbine-generator load conditions of 530 MW. Eachtest was initiated by opening the generator bus po-tential transformer 120 volt circuits supplying theEVA watt transducer in the Analog Electro-Hydrauliccontrol system. This simulated a three-phase faultin the critical zone.

Figure A.5 shows Unit 4 test 2 results at 530 MWinitial load. The intercept valves started to closeapproximately 0.18 seconds after test initiation andgo through full stroke in approximately 2 seconds.After remaining closed for approximately 2 seconds, avalve on the right side started to crack open fol-lowed by the other valves on both sides. It took inexcess of 27 to 28 seconds for each valve to reachthe full open position. However, I.P. inlet pressure(not shown) returns to essentially the initial valuein about 9 seconds due to the non-linear flow charac-teristic with valve position. This meets the 12 sec-ond time established by the steam generator manufac-turer. Good repeatability was obtained on each test.

600 -

500 -

400 -

300 -

200 -

100-

OPEN

GENERATOR POWER RETURNS TOINITIAL VALUE AT APPROXIMATELY 11 SECONDS

INTERCEPT VALVES FULL OPENAT APPROXIMATELY 28 SECONDS

V \ GENERATOR ELECTRICALMEGAWATTS

INTERCEPT VALVES START TO OPEN

INTERCEPT VALVE POSITION 7/

CLOSED

0 0.5 1.0 1.5 2.0 2.5 3.0

TIME -SECONDS

3.5 4.0 4.5

Fig. A.5 Plant Harllee Branch Unit 4 EVA Test 2.Initial load 530 MW.

In comparison with the previous tests on Units 2and 3, tests on Unit 4 were dramatic in higher noiselevel and longer duration of blow down when the re-heat safety valves lifted. This was due in part tothe slow reopening time of the intercept valves. Thereheat safety valves reseated properly during thetests. There were no other noticeable changes inunit operation.

The recorded generator terminal megawatt powerhas the characteristic expected for the turbine EVAtests. However, the results did not meet the pre-dicted values calculated by Westinghouse. The inter-cept valves must start to close faster to meet theWestinghouse calculated EVA performance character-istic. This has an impact on the stability of PlantBranch in that, with this EVA characteristic, un-stable operation would occur for a very close inthree-phase fault when at full load. Southern re-quested Westinghouse to advise what would be donesubject to their analysis of the EVA test data. Thefinal fix list was developed after several iterationswith Southern Company-Services and Georgia Power Com-pany. These items were:

1. Supply completely new servo actuators foreach intercept valve.

2. Replace the present intercept valve closingspring with two closing springs.

3. Modify the spring housing by making agroove to act as a guide for the new springs.

4. Alter the intercept valve piston rings.

5. Replace the Hall effect Watt transducer andthe turbine cross-over pressure transducer with threeunder power relays in series with the pressureswitch. This is for a complete modification of theEVA control logic to eliminate 0.08 second delay.

6. Make a minor change to the Analog EHCcontroller.

Shipment of these components and completion offield modifications on Unit 4 have been accom-plished. Since a decision had been previously madeto provide EVA controls on all new unit additions andseveral on-line units on the Southern electricsystem, these modification were completed on theseunits also.

Conclusion

EVA tests on the Southern electric system andsystem simulations have proven EVA to be a practicalstability aid. Although there were no adverse ef-fects found during and after the EVA tests on Units2, 3 and 4, EVA operation has been questioned ashaving the possibility of degrading unit reliabil-ity. Comments from the equipment manufacturers onthis subject would be of interest.

B. APPLICATION OF NEUTRAL RESISTORS AND FASTVALVING AT THE GERALD GENTLEMEN STATION(J. H. Doudna and D. L. Osborn*, Nebraska PublicPower District, Columbus, Nebraska)

The Nebraska Public Power District's (NPPD)Gerald Gentlemen Station (GGS) is located on theSutherland Reservoir approximately 32 km (20 miles)west of North Platte. The plant consists of two 600MW coal-fired units. The turbine-generator for Unit1 was manufactured by Brown-Boveri, and that for Unit2 was manufactured by General Electric Company. Theboilers for Units 1 and 2 were manufactured by FosterWheeler and Babcock & Wilcox, respectively. Bothboilers are drum-type.

During the summer. periods, when there is exten-sive use of electricity for irrigation and air condi-tioning purposes, the load center on NPPD's system isnear Grand Island, Nebraska (274 km or 170 miles fromGGS). During fall, winter, and spring, the electri-cal load center is near Lincoln, Nebraska (386 km or240 miles from GGS). Because of the large distancesand the high cost of transmission facilities, it wasnecessary to design GGS to operate with minimum oftransmission system additions.

The transmission system which resulted from thestudies, and which was in service at the time thefirst unit at GGS went in service, tied GGS to NorthPlatte and Ogallala by 230-kV lines and to GrandIsland by a new 345-kV line. Powerflow studiesshowed this transmission configuration to be adequatefor an intact system condition for carrying the en-tire 600 MW output of Unit 1. However, stabilitysimulations using this transmission system demon-strated the transient behavior of Unit 1 to be mar-ginal. To enhance transient stability, options otherthan further transmission additions were investi-gated. These included optimizing the step-up trans-former impedance, evaluating reclosing schemes, andselecting the preferred transmission voltage to whichto connect the unit.*Presently with ASEA, Inc.

n I:u

After examination of the foregoing items, it wasfound that Unit 1 could still be unstable for three-phase and close in single line-to-ground faults whichremoved the GGS-Grand Island 345-kV tie fromservice. The unit was also unstable for nearbysingle line-to-ground faults, particularly in connec-tion with a failed breaker which removed an addi-tional transmission element. These studies also de-monstrated that fast valving the turbine-generatorwas the most effective and least costly means of sta-bilizing GGS Unit 1 for these disturbances.

Since line-to-ground type faults comprise themajority of all faults, other options were investi-gated which would minimize the use of fast valvingfor severe single line-to-ground faults. Methodsinvestigated concentrated on limiting fault currentsby the insertion of a reactor or a resistor in theneutral of the generator step-up transformer. A re-actor, while limiting the fault current, did not pro-vide significant stabilizing effects on the gener-ator. Even with a neutral reactor, the transmissionlines leading away from GGS were not able to carrysufficient amounts of power due to the collapse involtage resulting from the fault. Thus the imbalancebetween turbine mechanical power and net electricalpower output of the generator went into the accelera-tion of the rotor. A neutral resistor of the sameohmic value as a reactor, however, provided a signif-icant stabilizing effect. Although the voltage col-lapse still did not allow the transmission network tocarry sufficient power away from GGS, the fault cur-rent flowing in the neutral resistor created realpower losses and thus provided an outlet for excesselectrical power which was produced by the turbine-generator during the fault. The power dissipation inthe neutral resistor was significant enough to elimi-nate the need for fast valving for all but singleline-to-ground faults which occurred in conjunctionwith a failed breaker and which also caused the lossof the Gentleman-Grand Island 345-kV line.

147

second 765-kV line, 155 km (97 miles) in length, willbe constructed from Rockport to a new station sitecalled Sullivan, which is adjacent to the existingBreed Plant in west-central Indiana. The Sullivanstation will be connected to the existing Breed345-kV system via a 765/345-kV 3000 MVA station. Inthe first stage of development when Unit 1 comes inservice, only the Rockport-Jefferson 765-kV line willbe in service. The Rockport-Sullivan line will beplaced in service by December, 1985. Therefore,there will be a period of over one year when thefirst 1300 MW unit will be operated on a single line.

To maximize the availability and reliability ofthe 765-kV outlets from the Rockport plant, single-phase switching will be installed on each line. Thisswitching feature provides the means to isolate tem-porary line-to-ground faults by tripping and thenreclosing only the faulted phase. The remaining twounfaulted phases remain in service to provide a pathfor synchronizing power from the Rockport plantduring this temporary disturbance.

While the single-phase switching feature willenhance the stability performance of the Rockportunits for a number of operating conditions, certainmultiple transmission outage conditions involving theRockport outlets would necessitate curtailing unitoutput to maintain adequate stability performance.To further enhance stability performance of theRockport units, temporary fast turbine valving willbe installed on each of the Rockport 1300 MW units.Results of simulation studies showed that the combin-ation of the single-phase switching and fast turbinevalving will enhance the operating flexibility of the2600 MW Rockport plant and maintain the desired levelof system reliability without resorting to operator-directed curtailments in anticipation of criticalcontingencies.

Fast Turbine Valving Concept

The benefits derived from the neutral resistorinstalled on GGS Unit 1 prompted NPPD to investigatethe installation of a resistor in the neutral of theUnit 2 step-up transformer. To assure that Unit 2would receive adequate stabilizing benefits from aneutral resistor, it was necessary to install a re-sistor in the neutral of its step-up transformer.Two factors dictated this: (1) there may be periodsof time when Unit 1 and its associated neutral resis-tor are out of service; (2) GGS Units I and 2 step-upto different voltage levels and are isolated by a345/230-kV autotransformer. Any ground current re-sulting from a line-to-ground fault and not flowingthrough a resistor would not cause power dissipationand thus greater rotor acceleration would result.Stability simulation for the present day system con-firm the benefits derived from neutral resistors.

C. TEMPORARY FAST TURBINE VALVING AT AEP's ROCKPORTPLANT (B. M. Pasternack, American Electric PowerService Corporation, Columbus, Ohio)

The Rockport generating plant comprises two 1300MW coal-fired super critical once-through units. Theboilers were manufactured by Babcock and Wilcox Com-pany and the cross-compound turbine-generators weremanufactured by Brown Boveri Corporation. The firstunit is scheduled for operation in September, 1984and the second in September, 1988.

Two 765-kV outlets will be utilized to tie thisplant into the AEP stations closest to Rockport. One765-kV line will be constructed from Rockport to theexisting Jefferson 765-kV station located in south-east Indiana, a distance of 180 km (112 miles). The

Fast turbine valving can be classified into twogeneral categories--temporary and sustained. To ac-complish either type of fast valving, the controland/or intercept valves must be closed rapidly. Bothtypes of fast valving produce a very rapid reductionin mechanical driving power following a critical sys-tem fault. Even though the rate of reduction is notfast enough to precisely match the reduced electricalload, the accumulated effect can significantly arrestthe swing of the generator, thereby improving trans-ient stability performance.

With temporary valving, the control and/orintercept valves are permitted to reopen to theiroriginal operating position, allowing the drivingpower to return to the prefault value very shortlyafter a predetermined minimum is achieved. For sus-tained valving, on the other hand, after the initialclosure of the control valves, the reopening of thevalves is adjusted so that the post-fault drivingpower is reduced to a new unit load level. In addi-tion, boiler controls must be adjusted to maintainsteam pressure and temperature within acceptablelimits. Sustained valving generally requires morecomplicated controls than temporary valving and isusually associated with long duration faults or sys-tem disturbances that result in permanent outagesthat significantly weaken the transmission system.

To initiate action of the control and/or inter-cept valves requires fast sensing of pre-definedelectrical disturbances and appropriate controls topartially close the valves in the shortest possibletime. The initiating signal can be developed by thesame relays that detect transmission network faults

148

or by measuring any one of several quantities, in-cluding generator acceleration or depressed bus volt-age.

Temporary Fast Valving for Rockport Plant

The steam flow path for a typical 1300 MW Brown-Boveri cross-compound turbo-generator is shown inFigure C.l. The steam from the boiler enters thehigh pressure (HP) turbine through the main controland stop valves. The exhaust steam from the HP tur-bine goes to the reheater for a temperature boost andenters the intermediate pressure (IP) turbine sectionthrough the reheat, intercept, and reheat stopvalves. The exhaust steam from the IP turbine goesto the low pressure (LP) turbine sections of eachshaft before being exhausted to the condenser forrecycling.

ro OM;H P STEAMIRlz,

2 l~~~~~~~~~~~~~~~~~~ELFCTNONtC ACCEL. ELECTROI FASt: l V^LtSJ t Sl$TlS ~LlIMIT CONTrROL VAL 1¢NCONROL

<VALVf _ $t{D/LO^D AC~~~~~~~~~~CELl>216 *..'SC C rSEgCODiZO

Fig.C.lSteam Flow Path and~~~~~Speed ConTrol fOrDIIa

T cNE TRyIPBIBNG 0OGIC

$^"z *s if"ctrT} . _ _ ~~~~~~~~~~~~~STEP-UPing operation, Kth me a d g p _rRaNSFORiER

several seconds later b aN re i P andS L

r \ ,.J~~~~~CONDENSFR

- - 7 h~~~~~~~~~~~OTWELLl..FCEED LP k

wit{"t HEATERS A^TR CObohDISATIo

Fig. CdinSteam Flow Path and Speed Control for aTypical BBC 1300 MW Unit.

By closing the control valves during a fast valv-ing operation, the Rechanical driving power appliedto the HP turbine wit be reduced rapidly, fol owedseveral seconds later by a reduction in IP and LPturbine driving power. Since the intercept valvescontrol over 70 percent of the total unit power, theuse of intercept valve stroking to supplement controlvalve action in the Rockport appl ication providess ign if i cant stab ili ty benef its by c aus ing a rapi ddecrease of intermediate and low pressure turbinepower.

Control valve closure must be carefully analyzedto avoid imposing severe transient stresses on theboiler system as well as other problems, such aslifting of boiler safety valves and turbinestresses. In developing the turbine valving charac-teristic for Rockport plant, a multi-disciplinaryapproach involving interaction among AEP's systemplanners, electrical and mechanical engineers andBrown-Boveri machine designers and control expertswas required. From preliminary data supplied by theturbine manufacturer, transient stability studieswere run to define the acceptable range of fast valv-ing characteristics that satisfied both the systemrequirements and the mechanical and thermodynamicconstraints of the boiler and turbo-generator. Basedupon extensive analysis, it was determined that thevalve stroke characteristics should conform to cer-tain specifications in order to meet the system per-formance requirements for the Rockport plant whiletaking into account equipiment restrictions. Thesespecifications are:

1. Control and intercept valves should close asrapidly as possible.

2. Control valve should close to 40 percentand intercept valves to 25 percent of full stroke.

3. Time delay before reopening valves shouldbe as short as possible (0.3 seconds or less).

4. Control valves should reopen to originalposition in approximately six seconds, and interceptvalves reopen to original position in approximatelyseven and one-half seconds.

Automatic reopening of both sets of valves tothe pre-fault position within 8 seconds from theinitiation of the fast valving is expected to mini-mize mechanical stresses and the possibility of lift-ing safety valves or causing turbine fatigue. Inorder to avoid thermal and pressure stresses in theboiler and turbine, the valves were allowed to immed-iately begin reopening after reaching the predeter-mined set point (25 percent for intercepts and 40percent for control valves). This reopening occursat the natural rate under speed-governor control.Furthermore, in order to minimize the impact of theboiler transients that follow a fast valving opera-tion, a second operation will only be permitted aftera delay period of approximately 10 minutes. The tur-bine fast valving characteristic is illustrated inFigure C.2.

The turbine power reduction response followingthe actuation of temporary fast turbine valving isinherently delayed due to the time lag involved infault detection, conversion of the electrical signalsto a hydraulic control signal and overcoming the in-ertia of the large valve masses. The 12-cycle timedelay shown in the fast valving characteristic be-tween fault initiation and the start of rapid drivingpower reduction consists of two components. Thefirst, which cannot be reduced substantially, is dueto valve inertia and hydraulic control fluid andsteam transport delay. These factors account forapproximately 7 to 8 cycles. The remainder of the12-cycle interval relates to system relay and controldelays, which may vary depending upon the fault con-ditions. In some cases, only 1 cycle of relay timemay be required, whereas for evolving faults orbreaker failure conditions, up to 5 cycles (orlonger) may be necessary.

At the Rockport. Plant, temporary fast turbinevalving will be initiated using the 765-kV line pro-tection system as the principal source of fast valv-ing signals.

,00

t A

60"-A

z - 00

TIME IN SECONDS

Fig. C.2 Relationship Between MechanicalDriving Power and Valve Position

Rockport Plant Transient Stability Performance

The Rockport Plant outlets are planned on thebasis of double contingency criteria. These criteriarequire that the Rockport Plant remain stable fol-lowing a disturbance, even with a critical transmis-sion facility out prior to the disturbance. Fol-lowing the permanent outage of a transmission facil-ity, the generation on the system will be readjusted,if necessary, to insure that all facilities are oper-ated within their normal ratings. Realistic contin-gencies were selected by applying the specific plan-ning criteria outlined in Table C.l.

Table C.A

ROCKPORT PLANT TRANSIENT STABILITY TESTING CRITERIAPRE-FAULT SYSTEM

I. All facilities in service

II. All facilities in service

III. One Rockport Plant trans-mission outlet out ofservice

IV. Most critical transmissionfacility remote from theRockport plant out ofservice

FAULT CONDITION

Permanent three-phasefault on eitherRockport outlet(cleared in primarytime)Permanent single-phase-to-groundfault on eitherRockport outletwith singlebreaker pole fail-ure (cleared inback-up time)Temporary single-phase-to-groundfault on the remain-ing Rockport trans-mission outlet(cleared by single-phase switching)Permanent single-phase-to-groundfault on eitherRockport outlet

The transient stability performance was evaluatedfor a multi-phase fault or a slow clearing (14cycles) single phase-to-ground fault on eitherRockport outlet with all facilities in service, andfor a single phase-to-ground fault on either Rockportoutlet with the remaining outlet or another criticalfacility remote from the Rockport Plant out of serv-ice. Primary fault clearing was assumed to occur 3cycles after fault initiation.

Table C.2 presents the results of those selectedtransient stability tests which illustrate the needfor stability enhancements for the two-unit/two-outlet configuration. As indicated, to maintain ade-quate stability margins for a number of possibleoperating conditions, particularly when one of thetwo 765-kV outlets is out of service and Rockport isoperating at or near its full capacity of 2600 MW,operator-directed unit curtailment of up to 500 MWwould be required in anticipation of the next criti-cal transmission contingency.

The transient stability performance of the plantwas reevaluated with fast turbine valving to quantifythe potential benefits. The last column of Table C.2highlights the results of this analysis. As indi-cated, actuation of fast turbine valving eliminatesthe need for pre-fault generation curtailment in allbut one case. The 100 MW operator-directed curtail-ment at Rockport for the case with the Rockport-Jefferson 765-kV line out of service is a reductionfrom the 500 MW curtailment required if no fast tur-

149

bine valving were employed. This required modestcurtailment is primarily due to the relatively weak345-kV system in the Breed area, where the second765-kV outlet from Rockport Plant is to be termin-ated. However, the transient stability simulationscarried out for this study did not consider the ef-fect of plant auxiliary motor load damping. The aux-iliary load requirements at Rockport total about 160MW, which has a significant damping effect on thegenerating units. Subsequent studies taking thisauxiliary load requirement into account indicate thatthese dynamics will eliminate the need for even the100 MW curtailment.

Table C.2

ROCKPORT PLANT STABILITY PERFORMANCE

PRE-FAULTCONDITION

All facilities inservice

All facilities inservice

Rockport-Sullivan765-kV line outof service

Rockport-Jefferson765-kV line out ofservice

Breed-Casey 345-kVline out ofservice

PRE-FAULTCURTAILMENT FOR

STABLE OPERATIONCONTINGENCY OF THE PLANT

TEMPORARY FASTTURBINE VALVING:

NO YESPermanent three- 100 MW Nonephase fault on.Rockport-Jefferson765-kV line

Permanent single- 300 MW Nonephase-to-groundfault on Rockport-Jefferson 765-kVline with breakerfailure

Temporary single- 200 MW Nonephase-to-groundfault on Rockport-Jefferson 765-kVline (cleared bysingle-phaseswitching)

Temporary single- 500 MW 100 MWphase-to-groundfault on Rockport-Sullivan 765-kVline (cleared bysingle-phaseswitching)

Permanent single- 200 MW Nonephase-to-groundfault on Rockport-Jefferson 765-kVline

Conclusion

Temporary fast valving at the Rockport Plantwill provide several benefits, including avoided unitcurtailments and reduced electro-mechanical stressesresulting from system disturbances. These benefitswill be achieved at minimum expense without makingchanges in the boiler control systems.

The development of the temporary fast valvingcontrol scheme for Rockport considered three key fac-tors: speed, selectivity and reliability. As dis-cussed, rapid detection of a critical system disturb-ance and immediate initiation of power decay is es-sential to derive the maximum benefit from the tempo-rary fast valving scheme. However, while valvingshould be fast, unnecessary initiation should beavoided to minimize wear on the valves and transient

150

effects on the boiler and turbine. Consequently, therelay and control system must be selective. Finally,reliability is important to insure that valving willbe properly initiated and completed for those systemfault conditions which could otherwise cause unittripping.

At the Rockport Plant, speed is achieved throughthe use of electronic controls in the plant coupledwith locating the control mechanism physically closeto the valves to minimize hydraulic fluid transportdelays. Further, all system fault conditions will bedetected using 1-cycle line relays at the Rockportterminal.

In order to achieve selectivity, permissivechecks will be included in the control logic. Thesechecks will be of two types. First, a plant loadingcheck will be made so that valving will only be ini-tiated when the units are loaded beyond a pre-definedcritical value. The second permissive signal, whichwill be required for double contingency cases, willbe turned on and off by the plant operator for prede-fined line outages on command from the System ControlCenter. In addition, the fast valving control willbe blocked for approximately ten minutes followingeach operation to permit the boiler to stabilize.

Application of fast valving improves stabilitymargins and thereby reduces the possibility of unittrips due to system disturbances. Fast valving ac-tion does result in boiler transients. The stressesaccompanying these transients are, however, judged tobe acceptable compared to the system and other bene-fits described earlier.

D. SYNOPSIS OF BALTIMORE GAS & ELECTRIC COMPANY'SEXPERIENCE WITH EARLY VALVING AT CALVERT CLIFFS(W. G. Thompson, Baltimore Gas & Electric Co.,Baltimore, Maryland)

In the late 1960's the Baltimore Gas and Elec-tric Company (BG&E) committed to two nominal 850 MWnuclear units at Calvert Cliffs in Calvert County,Maryland. Unit 1 was completed in 1973 and Unit 2was completed a year later in 1974. At that time theBG&E bulk transmission system was as depicted in Fig-ure D.I. It will be noted that the Calvert Cliffssite is external to the BG&E franchise area and istied into the BG&E 230-kV and 115-kV systems at WaughChapel Substation by two 500-kV lines 76 km (47 mi.)in length. The so-called Baltimore-Washington 500-kVloop that was to tie Calvert Cliffs to PEPCo's ChalkPoint Generation Station and Waugh Chapel to theConastone-Doubs 500-kV line was in the planningstage, but impediments to its completion were alreadyin evidence. Because of the very real possibilitythat the transmission system associated with theCalvert Cliffs Nuclear Plant might be limited to two76 km long 500-kV lines for a period of time, concernarose regarding possible generator stability problemsdue to transmission contingencies.

In 1969 Bechtel Corporation was commissioned todo an exhaustive transient stability study of theCalvert Cliffs plant to determine: (1) if any prob-lem existed and (2), if so, to recommend solutions.The study and its findings were reported in an IEEEpaper [7]. The study showed that, while the CalvertCliffs units were stable for a three-phase fault onone of the 500-kV lines at Calvert Cliffs with normalclearing times of 3-cycle at Calvert Cliffs and 4.5cycles at Waugh Chapel (Figure D.2), they would beunstable for the same fault with a stuck breaker and7.5 cycle back-up clearing (Figure D.3). The simu-lated effects of various stability aids, such as fastvalving, independent pole tripping, and extremely

fast excitation were tested. The one aid that wasmost effective was fast valving, and the study recom-mended the use of fast valving for Units 1 and 2.

TO MANOR \A

TO (

PENNSYLVANIA

TO DOUBS

--- 230 KV. CIRCUIT- 500 KV. CIRCUIT

CHALK POINT

Fig. D.1 Baltimore Gas and Electric Companygeographic location of 230-kV and500-kV circuits.

To quote Bechtel: "The stability benefits from fastvalving are so great for so small a cost that advant-age should be taken of them. Fast valving can obvi-ate emergency shutdowns for severe faults that wouldotherwise cause instability."

Theoretical benefits of fast valving can bereadily shown by comparing Figures D.3 and D.4. Fig-ure D.4 depicts the Calvert Cliffs unit swing curvesfor the same multiphase fault with 7.5-cycle backupclearing time as depicted in Figure D.3 but with fastvalving applied. From this curve, it is readily ap-parent why this new tool looked so inviting. And thecost of implementing fast valving into the final unitdesign was to be less than $10,000.

Using the initial information on the EVA systemthat was furnished by General Electric and Westing-house, a plot of input/output data was obtained forthe case illustrated in Figure D.5. The mechanicalpower input starts to reduce at 0.12 second andreaches about one-half the generator rating at about0.7 second. By reducing the accelerating power inputshortly after the fault has occurred, the angularadvance of the generator is decreased. The generatoroutput reaches about 1.35 per unit of rating at about0.95 second as the unit starts to pull back in angu-lar displacement toward the system. Figure D.6 showsthe turbine power response during fast valving. Thiscurve was provided by G.E. for Unit 1. In this case,the intermediate and low pressure turbines produce 70percent of the total turbine power and closing theintercept valves reduced the machine output to 30

151

percent of the initial output. This remaining outputwas supplied by the high pressure turbine. FigureD.6 also indicates time delays TI and T2 before valveclosure starts. T2 is fixed by the time for the sol-enoid to act which was stated by the manufacturer tobe 0.1 second. Tl depends upon the time for thepower unbalance relay to act, estimated by the manu-facturers to range between 1 and 3 cycles. The studyshowed that changing the total time delay (Tl plusT2) from 0.12 to 0.15 second increased the angularexcursions of the units by 8 to 10 degrees with allother conditions remaining the same.

U,u

:Ul

-I

w0

z

ulv

uJz

uiU.uiccU.

3L

0 .2 . 1 .6TIME IN SECONDS

.8 1.0 1.2

Fig. D.4 Swing Curves for same fault conditionsas Fig. D.3, except with aid of fastvalving.

( w`oP>

G.!-

'fl0

4(t0

-J ;

Lucc.s . 0

TIME IN gECNO5

Fig. D.2 Swing Curve--three-phase fault at CalvertCliffs 500-kV (3-Cycle clearing, nostability aids).

t160

80 XTLU /

C2 _

~~IM INSCOD

4

0

Z/ UNIT 2w 40W

U.

U.

0 .2 .4 .6 .8

TIME IN SECONDS

Fig. D.3 Swing Curves for three-phase fault atCalvert Cliffs on 500-kV line clearedin backup time of 7.5 cycles withoutstability aids.

0 0.2 0.4 0.6 0.8 t.O 1.2TIME IN SECONDS

Fig. D.5 Response of Unit 1 field voltage withconventional excitation responseration of 1.0, mechanical power input,generator output and terminal voltagefor same case as Fig. D.3 showing ef-fect of fast valving.

As the benefits of fast valving looked mostpromising, BG&E pursued the matter further with theplant equipment suppliers, namely General Electric,the turbine-generator supplier for Unit 1; Westing-house, the turbine-generator supplier for Unit 2; andCombustion Engineering, the nuclear steam system sup-

plier for both units. Initially, none of the sup-

pliers indicated any potential problems. Later, andwith respect to the turbine-generator only, GeneralElectric expressed some concern over the increasedduty placed upon their plug type intercept valves,since these valves must open against full pressure

differential. It should be noted that these valves

4 ED

2

a

A

152120 _ _

I

L.i MFAST REOPENING80E

_ _ _NOTE60 _ T1 TIMEDELAYTOSENSE

z I \ / SYSTEM DISTURBANCE

co

40~ T2 TIME DELAY OF SOLENOID

~~~~~~~~~ANDVALVE

20

0 1 2 3 4 5 6 7TIME AFTER INTERCEPT VALVE THROTTLING BEGINS (SEC.)

Fig. D.6 Power Pesponse to fast (early) valving forCal vert Cl i ffs Unit 1, intercept valveapplication, only. Curve for Unit 2 issimilar.

are designed to operate in this fashion to relievethe pressure bottled up in the reheat system suchthat loss of load does not result in an overspeedtrip. However, in that case the intercept valves areonly called upon to throttle a relatively low steam.flow, whereas they must pass full steam flow in amatter of seconds in case of fast valving. Westing-house, with their butterfly type intercept valves,did not see any problems.

Potential probleems were foreseen in the steamsupply system. Fven though Combustion Fngineeringwas of the opinion that the magnitude and duration ofthe pressure excursion experienced by the steam gen-erator on fast valving would not create any operatinaproblems with their equipment, Westinghouse expressedsome concern. Westinghouse thought that the severetransients caused by fast valving could adverselyaffect the systems within the steam supply portion ofthe plant. Systems such as the feedwater heater ex-traction and drain systems, moisture separator re-heater drains from both the shell side and the re-heater tube side, and the steam generator controlswere of concern. Westinghouse's experience up tothat time was that these systems could be marginalfor even normal transients, such as those associatedwith start-up.

Additionally, Westinghouse felt that the effectof water flashing into vapor due to rapid pressuredecays, and the effect of bubble collapse or implo-sion upstream of the valves due to pressure riseswould be difficult to evaluate. The low pressureextraction system can be subject to swell associatedwith flashing which is then collapsed by the pressurerise when the valves reopen. There was some thoughtthat on nuclear units the bubble collapse upstream ofthe valves may induce a significant shock in thesystem.

Finally, Westinghouse expressed concern withfalse trips. Up to that time, several false tripshad been experienced as a result of the fast valvingsystem. Considering that fast valving is employed tooperate only for low probability fault conditions,false operation as well as its inability to distin-auish among the type of fault, raised questions as tothe application of this technique. In light of theforegoing, potential disadvantages of fast valvingwere felt to far outweigh its potential advantagesand it was not employed at Calvert Cliffs.

Regarding BG&E's solution to the Calvert Cliffstability problems, the actual electrical impedancesof the unit transformers at Calvert Cliffs and thestep-down transformers at Waugh Chapel were consider-ably less than the predicted values used in thestudy. The actual system performance was thereforemore stable than initially predicted. The combina-tion of this with independent pole tripping on the

500-ky breakers and an excitation response ratio of 1was enough to eliminate the need for fast valving fora three-phase stuck breaker fault on the 500-kV lineat Calvert Cliffs. Completion of the aforementionedBaltimore-Washington 500-kV loop will further in-crease the stability margin.

SUMMARY OF PAtIEL DISCUSSION

Depending on system characteristics, either mo-mentary or sustained fast valving can be an effectivemethod of stability improvement. Peliable methods offast valving that do not compromise steam supply sys-tem and turbine availability would usually be pre-ferred over the unit tripping alternatives.

There are methods to reduce the frequency of fastvalving operations. Single-pole switching, independ-ent pole operation, and insertion of neutral resis-tors can eliminate need for fast valving for the morecommon single-phase faults.

Implementation of fast valving is, however, rela-tively complex involving all system components--thesteam supply system, the turbine-generator, and theelectric network. Close coordination among electri-cal and mechanical engineers of the vendors and theutility are necessary. We hope that this paper,along with discussions, will contribute to an under-standing of the benefits and limitations of the var-ious methods of fast valving.

PFFF PFHICES

1. IFFE Committee Report, "A Description of DiscreteSupplementary Controls for Stability," TEFFTransactions on Power Apparatus and Systems,Vol. PAS-97, No. 1, January/February 1978,pp. 149-165.

2. L. Edwards, R. J. Thomas, D. C. HoQue,P. Hughes, W. Movak, G. Weiss, J. E. Welsh, "Sus-tained Fast Valving at TVA's Cumberland SteamPlant: Background and Test Results," Proceedingsof American Power Conference, Vol. 3, 1pp. 142-152.

3. C. W. Durant, "Boiler Pesponse to Partial LoadPejections Pesulting from System Upsets," IFEFTransactions on Power Apparatus and Systems,Vol. PAS-101, No. 8, August 198Z, pp. 2630-2639.

4. H. F. Martin, D. W. Tapper, T. M. Alston, "Sus-tained Fast Valving Applied to Tennessee ValleyAuthority's Watts Bar Nuclear Units," ASMEJournal of Fngineering for Power, Vol. 99,series A, No. 3, July 1977, pp 399-405.

5. F. P. deMello, D. N. Ewart, M. Temoshok,M. A. Fggenberger, "Turbine Energy Control Aid inPower System Performance," Proceedings of.American Power Conference, Vol. 28, 1966.

6. J. Vandegrift, J. P. Woodall, J. T. Beckham, Jr.,"Fast Intercept Valving Aids Unit Stability,"Electrical World, July 13, 1970, pp. 56.

7. W. A. Morgan, P. P. Peck, D. R. Holland,F. A. Cullen, J. B. Puzek, "Modern Stability Aidsfor Calvert Cliffs Units," IEFF Transactions onPower Apparatus and Systems,Vol. PAS-PO, No. 1, January/February 1971,pp. 1-10.

153

BIBLIOGRAPHY

1. N. J. Balu, "Fast Valving and Independent PoleTripping Breaker Applications for Plant Stabil-ity," IEEE Transactions on Power Apparatus andSystems, Vol. PAS-99, pp. 1330-1342, July/Aug1980.

2. EPRI CS-3717, RP 1879-1, Assessment of FossilSteam Bypass Systems, final report, October 1984.

3. R. H. Park, "Improved Reliability of Bulk PowerSupply by Fast Load Control," Proceedings of theAmerican Power Conference, Vol. 30, 1968, pp.1128-1141.

4. E. W. Cushing, G. E. Drechsler, W. P. Killgoar,H. G. Marshall, H. R. Stewart, "Fast Valving asan Aid to Power System Transient Stability andPrompt Resynchronization and Rapid Reload AfterFull Load Rejection," IEEE Transactions on PowerApparatus and Systems, Vol. 91, pp. 1624-1636,July/August 1972.

5. R. H. Park, "Fast Turbine Valving,: IEEE Trans-actions on Power Apparatus and Systems, Vol.PAS-92, pp. 1065-1073, May/June 1973.

6. IEEE Committee Report, "Bibliography of Liter-ature on Steam Turbine-Generator ControlSystems," IEEE Transactions on Power Apparatusand Systems, Vol. PAS-102, pp. 2959-2970,September 1983. (Includes 12 references on fastvalving with annotations.)