ML20206E850
| ML20206E850 | |
| Person / Time | |
|---|---|
| Site: | South Texas  |
| Issue date: | 11/30/1988 |
| From: | HOUSTON LIGHTING & POWER CO. |
| To: | |
| Shared Package | |
| ML20206E839 | List: |
| References | |
| NUDOCS 8811180218 | |
| Download: ML20206E850 (311) | |
Text
r i ,
Steam Generator tema!!=rass STATION Feed Pump- Turbine l Overspeed Failure /
Recovery Report _
4
's ~2 - .T,. h _c 1
- ggWg
- a;7-i
!{ldir fL te
- f g$@%jjp-25' The Light company
, i _ _ u t.... , & ,.~ ,
e November 1988 ra " m !: s w F' s c,
{
STEAM GENERATOR FEED PUMP TURBINE OVERSPEED FAILURE / RECOVERY REPORT PREPARED BY HOUSTON LIGHTING AND POWER COMPANY SOUTH TEXAS PROJECT ELECTRIC GENERATING STA1!ON NOVEMBER 1988 G-0647
a TABLE -OF CONTENTS Page Executive Summary S.1 Introduction 11 1.0 PREPARATION FOR THE LOSS OF 0FFSITE Pok'ER TEST 11 1.1 Plant Status 11 1.2 Discussion of LOOP Test and Test Preparation 11 8.0 THE MAIN FEED PUMP TURBINE OVERSPEED EVENT 21 2.1 Eyewitness Reports 21 2.2 Discussion of Computer Data obtained During the Event 23 2.3 Computer Data Relevant to the Sequence of Events 24 2.4 Sequence of Events Developed From Computer Data and Equipment 25 Examination 3.0 POST EVENT PicANT STATUS 31 3.1 General Summary of Turbine and Pump Damage 31 3.2 Detailed Damage Description 31 1
4.0 ROOT CAUSE ANALYSIS 41 1
4.1 Ceneral Discussion on Root cause Investigation 41 4.2 Analysis Validating Root Cause Conclusion 43 S3/TGANA/a
Pan 5.0 EVALUAhoN OF THE DESIGN, TESTING, AND MAINTENANCE BARRIERS VHICH 51 SHOULD HAVE PREVENTED THIS EVENT 5.1 Steam Generator Feed Pump Turbine (SCFPT) Design 51 5.2 Description and Evaluation of Turbine Testing 56 5.3 SCFFT Maintenance 5 11 6.0 REVIEW OF CENERAL PIANT SYSTEMS AND EQUIPMENT FOR SIMILAR PROBLEPS 61 6.1 Review for Potential Impact Due to 80P. TSC or Lighting Diesel 61 l
l Operation l
6.2 Adequacy of other Turbine Stop Valves 62 6.3 Review of Seal Water Design for Other Pumps 64 7.0 PLANT MODIFICATIONS AND CORRECTIVE ACTIONS 71 7.1 Short Tern 71 7.2 Long Tera 73 7.3 Transient Information Retrieval 7 10 8.0 TURRINE AND PUMP RESTORATION 81 9.0 LESSONS LEARNF.D 91 10.0 RIFERENCES 10 1
- 1. Compiled Eyewitness Reports 10 1
- 2. Compiled Telecons 10 1 S3/TCANA/a i
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 1
?.*At
4 Stop Valve Test Procedure 10 1
- 5. Detailed List of ECNs, DCNs, and Modifications Incorporated 10 1 During the Restoration Effort
- 6. White Paper on Improving Sequence of Events Records, 10 1 Collection, and Archiving
- 7. Computer Data 10 1
- 8. Expert Subcommittee and Joint Company Report from Orlando 10 1 Conference entitled, "South Texas Project 5/25/88 SCFAT
' Incident Investigative Team Report'"
- 9. Telephone Conference Notes with Other Utilities, etc. 10 1
- 10. Onsite Investigation Team Reports 10 1
- 11. Evaluation Report by Stone & Webster Engineering Corp. (SWEC) 10 1
- 12. Coupling Vendor Repwat 10 1 18.0 APPENDICES A. Main Feedwater and Turbine Steam Sources Systems and A1 Component Nacriptions
- 8. Turbine and Pump As Found Conditton with Photographic Record 31 C. Kalsi Engineering Report, "Westinghouse Feedwater Turbine Stop C.1 and Throttle Valve Analysis Unit l' Document #1566C, dated August 2, 1988 S3/TCANA/a
ft Parc D. Turbine Shaft Fragment Missile Analysis D1 h
E. Southwest Research Report, "Metallurgical Evaluation of E1 Fractured Components from Steam Generator Feed Pump Drive Turbine No. 11" F. k'estinghouse Customer Advisory Letter 88 01 F.1 v
=
= 1 S 3/TC/NA/a L
INTRODUCTION On May 25, 1986, the all Steam Generator Feed Pump Turbine at the South l Texas Project Nuclear Power Plant, Unit 1, was destroyed as the result of i
extreme overspeed. This report documents the investigation, restaration, root causes and corrective actions purausnt to this event.
l Because of the volume of the information presented, readers may wish to consider the following:
o Upper Level Utility panagement Road Executive Summary, review the photographs of damage in Section 3.0, and read Section 4.0 on the root cause.
o Entineers Responsible for Steam Turbine Driven Pu.eps and Individuals Desiring a Thorough Knowledte of thts occurrence.
Read Executive Summary Appendices A and B, and then Sections 1 9.
o All Others Read Executive Summary and review Table of Contents to select sections of interest.
l 1
$3/ICANA/a 11 l
l
EXECUTIVE
SUMMARY
STEAM GENERATOR TEED PUMP TURBINE OVERSPEED FAILURE / RECOVERY REPORT During Power Ascension Testing en May 25, 1988, a catastrophic failure of a Vestinghouse Model EKM32A1 NP feed pump turbine occurred at South Texas Project, Unit 1, a 1250 megawatt nuclear power plant. This report contains information on plant conditions at the time of tho feed pump turbine failure, a description of the damage, a discussion of the root causes of the failure, a l summary of the actions necessary to effect recovery from the resulting damage, I and a review of the corrective actions found to be necessary to preclude recurrence.
l INTRODUCTION:
1 l
On May 25, 1988, at 1530 hours0.0177 days <br />0.425 hours <br />0.00253 weeks <br />5.82165e-4 months <br />, the final preparations were taken for the Loss of Offsite Power (LOOP) Test. Personne'. in the Control Room and throughout the plant were gathered to witness the test. The planned initiation of the loss of of fsite power, coincident with a turbine trip, was one of the last major milestones prior to plant power ascension to 100 percent and commercial operatior.. To start the test, the Control Room operator opened the switchyard breakers. The ensuing main turbine and reactor shutdown was normal. The onset of natural circulation in the primary system was as expected. All three (3)
(mergency diesel generators started successfully and sequenced their essential loads.
Approximately 30 seconds after initiation of the test, reports from the turbine deck were received in the Control Room, that parts of the feed pump turbine hal .
l been ejected from the machine. Attempts to stop the turbine from the local control panel were not successful. The Control Room operators actuated the 6 main steam isolation system and shut off steam to the feed pump turbine.
'/TCANA/a S1
An cxtensivo inv3stigoticn cnd rostoraticn effort cn tho d3stroyod turbine was initiated immediately. The investigation revealed that the pretension on the high pressure stop valve springs for all feed pump turbines was far below the value needed to prevent this occurrence. The investigation would also reveal that no instructions had been supplied by the manufacturer to 0,idicate that the springs required adjustment. Finally, it became apparent that a critical parameter, the springs' pretension, was inadequately addressed during the original fabrication and subsequent Westinghouse directed refurbishment work required to validate an extended warranty.
SEQUENCE OF EVENTS SYNOPSIS:
The Steam Generator Feed Pump Turbine (sCFPT) received a trip signal milliseconds after the switchyard breakers were opened; but, the turbine did not stop. The horsepower being delisered by the turbine exceeded that required by the pump. This caused the turbine to slowly accelerate.
At the time the switchyard breakers were opened, the main feedwater booster pump lost power. This affected the Steam Generator Feed Pump (SCFP) not positive suction head. Approximately five seconds after the mechanical overspeed trip device actuated, the pump speed had increasca 500 rps above its trip setpoint. As the booster pump coasted down, the hot fluid flowing from the deserator eventually flashed in the SCFP. When the fluid in the SCFP l
flashed, the turbine rapidly accelerated until it reached a speed over twice its design speed. The turbine threw two low pressure (1.P) blades resulting in rotor imbalance and severe vibration. The continued overspeed operation and severe vibration resulted in a loss of all 1.P stage blades, rotor shaft failure, and coupling separation. A 215 pound shaf t segment was ejected across the turbine deck. SCFPT blades, lubricating oil, and other debris were expelled into the turbine exhaust duct. Several pieces of turbine blades reached the condenser, damaging 22 condenser tubes. Control Room operators closed the main steam isolation valves, two minutes after the LOOP was initiated, to stop steam flow to the SCFFT and terminate the event.
S3/TCANA/a S.2
$UMMARY:
During a Loss of Offsite Power Test, the all SCFPT high pressure stop valve did not close, resulting in turbine overspeed and subsequent failure. This cecurred due to insufficient valve closure spring preload. Testing revealed that a spring preload greater than seven (7) times the preload present at the time of the event was required to ensure valve closure.
- Normally, the governor valve is available for speed control; but, loss of offsite power disables the control system for the governor valve and prevents its closure. In addi' tion, vibration, bearing temperature, and other normal trips became inoperable because of the loss of power. Therefore, corplete reliance was placed on the mechanical overspeed trip device to provide a trip signal for closure of the HP stop valve. Although this mechanical overspeed
(
trip device functioned properly and provided the necessary trip signal, the stop valve failed to close.
FUTURE FREVENTIVE MEASURES:
To prevent such failures in the future, the plant's corrective actions include:
- 1. The Steam Generator Teed Pump Turbine Manual has been revised to ,
I incorporate detailed instructions, developed by Vestinghouse after the event, for setting the high pressure stop valve spring tension.
Additional actions to enhance plant protection taken include:
1 t
- 1. The power source for the SCFFT Electrohydraulic C-ntrol System has been changed to a vital and non vital source, rather than redundant non vital sostees. This is to ensure that the governor valve control circuit will have AC power available. With power to these controls, the governor valve vill control turbine speed during a LOOP. A vital power source was also provided (or the Turbine Vibration Monitoring Systen so thre this instrumentation remains functional under LOOP conditions.
S3/TCANA/a S3
- 2. A turbine trip signal from low SCFP not positive cuction head has been added. This will protect the pump from a loss of SGTP suction pressure.
- 3. An electrical overspeed trip has been added to the SCFPTs.
- 4. The local manual lever, which can be used to activate the mechanical overspeed trip mechanism, has been relocated to improve accessibility.
- 5. The auxiliary feedvater and main turbine stop v61ves have been reviewed with respect to adequacy of spring force and adjustment instructiona.
This is to ensure their ability to close with high steam pressure and low steam flow conditions.
1 l
$3/tG.ANA/a S4
l 1.0 PREPARATION FOR THE LOSS OF OFFSITE POWER TEST 1.1 Plant Status In preparation for the Loss of Offsite Power (LOOP) Test, scheduled for 1530 hours0.0177 days <br />0.425 hours <br />0.00253 weeks <br />5.82165e-4 months <br /> Vednesd.y, May 25, 1988, the South Texas Project Unit i reactor and turbine were on line supplying power to the grid. The initial plant conditions were as follows:
o Reactor power was 24 percent, o Main turbine load was 19 porcent.
o The all Steam Generator Teed Pump Turbine (SCFPT) v.s supplying all four of the Westinghouse Steam Generators.
o The steam supply to the SCFP was from high pressure (HP) main steam.
o The Balance of Plant (BOP) diesel was in service.
1.2 Discussion of LOOP Test and Test Preparations Regulatory Guide 1.68 Revision 2, entitled "Initial Test Prngrams for Vater Cooled Nuclear Power Plants", provides a list of tests to be conducted during power ascension. Included, is a test to demonstrate the dynamic response of the Nuclear Steam Supply System to a LOOP and turbine trip.
Although not a planned part of the LOOP Test, the BOP diesel had been in operation since the previous day. This was the result of efforts to correct a problem with its output breaker. The LOOP Test procedure calls for the BOP diesel to be in standby but not in operation. It was decided to allow continued BOP diesel operation at the start of the LOOP Test.
l S3/TCANA/a 11 I
Vith the BOP diesel in service, power was immediately available to s the SCFPT AC backup lube oil pump. This could allow the AC backup pump to automatically start and prevent the expected low lube oil pressure SCFP turbine trip. However, a trip signal was generated milliseconds after the LOOP was initiated, probably due to transient low lube oil pressure which could have been present before the backup pump reached rated flow and pressure. Therefore, the continued operations of the BOP diesel had no effect on the turbine failure.
I 1
1 53/TCANA/a 12
2.0 THE MAIN FEED PUMP TURBINE OVERSPEED EVENT 2.1 Eyewitness Reports Eyewitness Reports were obtained from all observers immediately following the event. Key observations are summarized in this Section. The actual reports are available in Reference 10.1.
2.1.1 Observations by the Eng.ineer Stationed at the Steam Oenerator Feed Pump Turbine (SCTPT)
Summary:
This Engineer was the Startup Test Engineer for the SGTPT, and has over the past several years conducted numerous trips and tests on these machines.
Report Extracts:
1535 Loss of Offsite Power (LOOP) was initiated.
1535 Vhile observing bearing oil pressure, noted audible increasing rpm.
1536 Observed turbine rpm increasing rapidly through 6,000*
rpm.
1536 High pressure (HP) stop valve indicated shut; repeatedly pushed turbine trip.**
1537 Audible indication and turbine tachometer pegged out, indicated extremely high rpm.
1537 Noted excessive vibration and smoke; moved away from the area.
1537 Observed the "coupling" cross the turbine deck at a high rate of speed to the northeast.
- As observed from the local console tachometer
- Electrical Trip Button on loca). console S3/TCANA/a 21
Post Incident Observations:
- 1. HP stop valve is full shut with limit switch closed.
- 2. Deluge in full operation.
2.1.2 Observation by the Engineering Supervisor on the Turbine Deck S u.ma a ry :
The second eyewitness was also on the turbine deck at the northeast corner. This Engineer was in the general path of the turbine shaft missile upon its ejection. He had a clear view of the turbine and panp prior to the event. Notable, was his observations of the con:inued rumbling and rolling of the turbine, even after the missile ejection. This shows the turbine had not seized and was still being driven by some force. In addition, his observation of the spinning of the missile shaft fragment, after it hit the ground, indicated that the turbine was spinning at a high rate of speed when this piece was ejected.
Report Extracts:
1530 Positioned myself along east rail of TCB EL 83' to observe eli SCFPT, deaerator structure / piping movement, SCFFT problems.
1535 LOOP initiated . heard notaal shutdown sounds.
1537 Noted loud rumbling emitting from turbine - heard snap 6 witnessed "coupling" missile being ejected; missile <
stayed airborne across turbine deck as it traveled northeast; missile struck nothing on the deck; and fell toward Secondary Water Makeup Tank (SKT) through a pipe barricade, breaking oif two bars and a piece of PVC pipe. Missile bounced on ground and then struck the SKT causing significant spark and dent. Missile rebounded off the SMT and landed about 20 feet away, spun in place causing dirt cloud. No personnel were etruck.
1538 Noted continuing rumbling emitting from ell SCTPT -
S3/TCANA/a 22
heard turbine beginning to slow; no other missiles noted.
2.1.3 Control Room Report Summa ry:
The final eyewitness report is a compilation of key observations from the Control Room operators.
- The operators became aware of the turbine overspeed condition via a local observer.
Report Extract:
The following observations were noted about all SCTPT overspeed:
- a. Initially, turbine 6 ed decreased to zero as noted on Control Room rps indicator. All control board indications and lights were deenergized.
- b. Local reports at all SCTPT vere that the turbine was flying apart,
- c. Control Room speed indications were erratic, pegged high and returning to zero,
- d. Main steam isolation signal given to isolate steam to oli SCFl'T .
2.2 Discussion of Computer Data obtained During the Event The primary method of gathering data from the secondary plant is through the Proteus Plant Computer System. This system also includes the plant Sequence of Events (SOE) recording. SOE is the primary method of determining the exact cause and effect of plant transients and is highly accurate. The computer processes SOE data events as "interrupta" ensuring a highly accurate time resolution for data points.
l S 3,'TCANA/a 23 l l L
In addition to SOE, the Proteus Pim.t Computer System prints predefined alarm reports on the Alarm Typer. The time resolution for Alarm Typer data is not as accurate as the SOE data (normally, within two seconds versus the two millisecond accuracy of the SOE). This is because the time indicated is when the computer scans the input for a change of state.
A third method of gathering data is through the Point Status Summary.
Point Status Summary is the retrieval of archived data. The Proteus Plant Computer System vill display graphical representations of systems at the request of the terminal operator. Data that is used for these graphical representations, has the lowest priority for scanning. The time resolution of this Point Status Summary data is not very accurate, since it depends on scan rate and computer system loading. An analysis of this recorded data indicates that the events recorded by Point Status Summary, under heavy computer loading, may appear as much as thirteen seconds after they occurred.
After the lh0P Test, personnel gathered the various hard copy l printouts from the Proteus, SOE, and Alaru Typer. In addition, SGPPT Point Summary Data were obtained. Analog data from the Proteus computer for feedwater system and the SCF?T vere also collected.
2.3 Con.puter Data Relevant to the Sequence of Events f The SCPPT received a trip signal 328 milliseconds after the start of the IDOP Te s t . At three second'.. the turbine speed was recorded by the computer as 5,320 rpa, or 74 rps faster than the initial speed.
All four main feedvater isolation valves were closed at five seconds.
At six seconds, turbine speed was at 5,700 rpm. Turbine speed was off scale high, greater than 6,350 rpm at thirteen seconds. The DC lube oil pump was operating at fourteen seconds.
S3/TGANA/a 24
- - - /
Turbine speed was still off scale high at twenty three seconds. A low lube oil pressure alarm occurred with both the standby and DC lube oil pumps operating at thirty three seconds. At thirty five seconds, the overspeed signal cleared. Approximately fifty seconds into the event, bearing temperature alarm signals were received.
In addition, SCTPT overspeed alarms intermittently alarmed and cleared from one minute and 30 seconds on, until the panels were deenergized. At two minutes and 14 seconds, Control Room action isolated the main steam lines.
2.4 Sequence of Events Developed from Computer Data and Equipment Examination The 1DOP interrupted power to the inservice feedwater booster pump, the SCFFT Electrohydraulic Control System, and the SCFPT Vibration Monitoring System.
At 328 milliseconds into the 1DOP Test, the SCFPT received a trip signal. A review of pump discharge pressure and turbine rp. computer data shows that the SCFFT did not coast down following this trip.
The SGTP speed gradu lly increased as the booster pump coasted down.
The isolation of the feedvater lines reduced pump loading. The turbine speed began to increase due to excess turbine horsepower. A high speed alarm was received at six seconds. Turbine speed continued to increase because steam flow had not been interrupted, despite the turbine trip signal. Pump load continued to decrease as the fluid trapped in the pump approached saturation pressure. This was dua to the booster pump coast down and the closed discharge flow paths. At 13 seconds, the punp fluid flashed to steam. Catastrophic damage te the turbine occurred when the turbine went to extreme
$3/TCANA/a 25
overspeed. At this point, ti.e turbine rotor shaft failure l occurred. A section ef the rotor shaft, with part of the coupling, vas thrown clear. The turbine lost all of the low pressure rotating blades.
l At two minutes, the Control Room operator closed the main steam lin, j isolativa valves to isolate the steam supply to the SGTPT. The severely damaged rotor remained in the case. The rotor finally stopped rotating after several minutes when the steaa line depressurized.
$3/TCANA/a 26
I k
3 g ! r"i : lI '
5 lllllfl.
- qe ll8
!n k
,e!!! i il i!;
S ll l
+i ! , Ed18 8 :i
~
"0 m
o 0l i l!!I
<m E i ill3!l8 i;Ih s = 2 s E l55 ]
O l
- ,il r
H ! -
5 s
,lei,ilpll!lla!t l
s .
W-W
= ,
a
,lilglllb i
I lui il1 i- k.w.o l 1 1 **
F-- l- ll .
I@ !$- !
116.*
- r m DDIgl*El;ll ini l lag 8
!g- .
!!!!!IWBEll .
,ah xxc1111iWl..**
yli 1. '
a
3.0 POST EVENT PiR'T STATUS 3.1 Ceneral Summar, of Turbine and Pump Damage The Steam Generator Teed Pump Turbine (SCTP7 experienced a catastrophic failure. This occurred at a speed approximately twice the design speed. The destruction resulted in the loss of all blades on the fourth, fifth, and sixth low pressure (LP) stages of the turbine. The destruction culminated in three separate fractures of the turbine rotor shaft. These break points are depicted on Tigure 3 1. One segment , 22 inches in length by 11 inches in diameter, weighing 215 pounds, was ejected at approximately 230 miles per hour from the turbine operating deck. The ejection of this turbine shaft missile bent the 6 inch diameter solid pump shaf t by 5 inches. The pump inboard shaft bearing housing was broken into fragments.
3.2 Detailed Damage Description 3.2.1 Steam cenerator reed Punp Turbine Appendix B provides additional photographs of the t urbine and pump damage. These photographs, in conjunction with Figures in Appendix B, reflect the as.found positions of majcr pieces of the turbir.e.
The photographs in this Section are arranged to follev the narrative.
The turbine damage vill be described starting at the south end of the turbine. Refer to Figures 9, 10, 11 and 14 of Appendix A for the general arrangement of the turbine and the overspeed trip device pttor to reading this Section.
The upper half of the shaft split case bearing housings (south end of turbine) was found approximately 2 feet to S3/TCA.g/a 31
a- , ,,
![dhd .h:i, .,
, c
.,,1/'
i the east of tha shaft center line. The bearing housing bolts on the west side had been stretched and bent prior to 6 failure. The bolts on the east side were sheared off level
'with the housing. The lower bearing housing and pedestal f -
were in place (Refer to Photographs 1 through 5).
The end of the turbine shaft failed at the location of the
. speed pickup wheel, where a diameter. change of the shaft occurs. This piece, approximately 4 inches in diameter and 6 inches long, contained the mechanical overspeed trip mechanism. The segment dropped down the rear bearing oil
, return line and came to rest in the oil sump.
(Refer to Photographs 6 through 8.)
4 4
=
1 i
1 S3/TCANA/a 32 1
i
, _ _ _ . _ , - _ . _ . - _ . - - , _ . - - - _ _ _ . . _ ,- .. .._-_-_,m y- ,c___r__- _-___.,_. -__ - --,, .,y -,
l l l 1
q D 3 E
E .
E PP R SIR TG U G RT E i SNI G I
N V UR F R RA A
E O
aH.
a HE TB 1
+ L B
A N
u M i
[ E N
R .
I L
U O TK J FA
\- l/
N G .
_ NI E R
_ I A TL S F S
i l
- AK 7l E
T HA VE L
_ 2 SE N + L A R
- O N T B P R 1 .
_ - - C E U - T R O J
_ U r1E s
- 1 L 1
N P
_ T F
_ P T F
_ G F A
S H S
_ G E N N
_ I
_ R T B A CN R E
_ U B _ HO T
L A
hr s S P T
N N R E U
O J
B
- t
_ A
_ P E W S M
U P
_._ CN G
I R S
A C
G .
N I
_ R
_ A E
B .
M ym- M E
N R
U O
E [
B T
_ J S U
R H
T
r ; j Y,
3
, "W-W'I ' $.e. 9
+0 f, j j
'f ,p / ' .
f -
, \
l.
mi L ,'
Ja s % a l ,
rheto I.
The upper bearing half (part F) from the south end of the turbine is shown in the as found position. (View is looking north)
) t; r i
?
- QL l
s' \\ , .
,'mg*< m,' p , [
\ f ?, ;
t l 'A's s n
,7 '. \.C -
t
- ,p~
6a ..
photo 2.
A piece of the Turbine shaft (part G) can be seen in the lower right of the i.
l photo. (Part G is not in the as-found position)
I
P.
g ) ,
- i. u, e.
.q ,-
l
.7
, 4 .
.j i j i
s.
" . .y photo 3. !
The lower bearing half reveals stretched and bent stud bolts. t (View is of the west side looking north) !
/
,7
., '. Y, y . . , ;
'" ' ? ,r,
- i j
d f
(- .
- 3. ,
jc ,; ,
$ t
/ :
q._ w g. 'r i %, 3 ..
b .
photo 4.
l The lower half of the turbine bearing can be seen in place.The STP inspection
) ! cam partially removed the cast side stud bolts. '
i i
. . _ . ._ - _ _ _ _ _ . - _ _ - - . . . _ . _ __ .~.. _ _=. - _
~
- l,5 3. .* k -
s ,,
r
\. .
.i . ,
~
w rv rT .
eY
/
~ f,.'yllf f j f -
W l lf' \ 3 '$G
. . &. Al *
-- . v. . ,
photo 5.
South end of turbine looking west. I.ower bearing skid anchor bolts arc visible in the upper right and are intact.
- n. -
. ~.
b
{ .
ua _ k d
L.-
photo 6.
shaf t tical o the f led turb ne
l
' e
- g. '&
j,s -
6 0 -
photo 7.
'-/
- i. - ?
. % -q'. '
AF View of the south end of the turbine 1
- looking down into the oil sump through l I4" ' '
{ the bearing oil return line. Also '
J
- .~ 'i'* visible is the trip valve (part AF). l
.w. ~,
i e .
j~ -
[. , s ,
, 4 ,
b._. $ l i !
m ,_
- W%C%. ~.[;.y',eqg) _
,%La
- r. .
I
'Zj;;q^ x. !-
~*
y -
' Y *[..
' ,\ ,,,
4 a.. ,i' 't
=
(t.
s l<-
U , *( k ,
j l ,
Y T' % - ey .
Qg.' I l
l
< ' c,\
'l . '
M ~p -
/
i ,
[
/ ,
4 1
. ' ~ $,
f photo 8.
I I Shaf t piccc Al;' containing the overspeed trip weight is visible !
through the inspection hatch (view is south cnd of turbine). ;
i
Tha mech::nical evorspacd trip velva v:s fcund slightly dislodged, but in place. However, the overspeed trip lever was missing from the valve. Inspection of the valve showed that it would still actuate with thumb pressure. Valve stem travel was smooth. It can be concluded that this valve was operational at the time of the event.
The overspeed trip mechanism was retrieved from the oil sump.
- The overspeed trip weight was found lodged in place due to mechanical distortion of the overspeed trip weight housing where the shaft failure occurred. The end of the overspeed trip weight was extended approximately 1/8 inch past the l diameter of the shaft. The overspeed trip lever clearance setting is 0.06 inch. This indicates an extreme overspeed I occurred. See Figure 14 of Appendix A for details of the overspeed trip mechanism.
The pivot pin which mounts the overspeed trip lever on the trip valve had separated. The lever was found in the oil l sump. It was broken in two at mid span. The surface of the i
lever, where contact by the overspeed trip weight is made, i
was smoothly hammered over 90 percent of this area. The l hammer marks were circular peening indications. This is l evidence of actuation of the mechanical overspeed trip mechanism.
The other side of the trip lever had heavy denting at the l point of contact with the pilot valve stem. These circular
! dents matched the diameter of the pilot valve stem. There was also heavy denting of a smooth circular nature at the l point of contact of the lever with the stop pin. These indications confirm that an extreme overspeed condition was in progress prior to failure of the turbine. From this j evidence, it is obvious that the mechanical overspeed tr.'p l
mechanism functioned as designed and was destroyed trying to trip the turbine.
53/TGANA/a 33 1
l l
l l
I Th3 rosat cnd manual. trip plungar reds for tha mech:nical overspeed trip mechanism were separated from the rods entering the mechanism. The link pins connecting the rod j extension pieces had sheared. The manual trip plunger rod '
I was bowed where it penetrates the valve body. Review of maintenance records indicates the manaal trip rod was bent prior to the turbine f!r.ilure.
(Refer to Photographs 9 through 26.)
4 i
i I
S3/TGANA/a 34 g9
/ i i /
,_m a ,'# '&
4 photo 9.
The normal position cf the Trip valve and the Actuator arm in the tripped condition are visible on a undamaged turbine shaf t. (view is of the south end of the turbine,looking north).
,' 2 I
l
,&e
- m. s l j"
l'
'/ gg '{
aq
- t ,
1 I
photo 10.
The Trip Actuator arm is shown held in the normal (untripped) position but the trip valve rod is in the tripped position. (View of the south end of the undamaged turbine)
- ~ ~ ' ~. .,
4.
p:
'.. K ..
' *'r.
ys
~j x '% y,
%s
[
- +
F
'9a. x g i .
p ) .? Jr
.a : ~.
>~
.\ /
[ I 6
s , . , ,
1
.'T . ;3 -
fa
- s. , ..r . .c' .
7; , - 0 {:'{ '
- y photo 11.
The at-rest position of the trip weight actuator button is shown on a undamaged turbine shaf t (view is of the south end of turbine, looking up and north).
~
i t
l
\ ,
, l
~
&l
.p d 4
- ^ ,
/
5--' })
< N ,
l photo 13.
The broken shaf t end, containing the overspeed trip weight (part AF'),
was found in the oil sump. It is shown undergoing inspection.
}
i
_ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _____..___m ____m._____ . _ _m_ m_ _ _ -__ ,_ _ _ _ _ _ _ _ _ _ , _ _ . _ . _ _ _ _ _ , ,
I
? )
YMW M -
% W< -
l I
"% \
photo 1s.
l The damaged Overspeed Trip valve was still capable of hand actuation and !
smooth movement of the valve stem, !
l I
i
!1
,1 I!
5 t1 a :
.* 3 f I %
~
i i photo 14. (
A close up view of the Overspeed Trip valve (item AP) showing the location of the pivot pin (note the Trip Actuator Arm is missing). l I
_ _t
l
. . N ,
,g ' V . ~ j ER STOP PIN r, ... !
- 3 . ., ( ,
- p. ; dv' y',
s - ., , y;;;.,,, . . #[ ~ ~ 5;
- ,p.e , ..
g
< l i ' p .o , '
v., . ,
' ~
j
.Q:. !
I p
- r :
)
i 4
i
, tr l
! photo 15.
An end view of the Trip valve reveals flattening of the Trip Actuator Lever i pivot stop.
1 I
l
. l A :
i !
' l 1
i l
4 + % .
- 4. ,
i PlVOT PIN e wf we ,,
- '* *
- Muo ,
- l [yty y, ' _ T 'f V" j
I .
..<..'A:..4~.,..,so
.. ., a u'
f l
J photo 16.
Two pieces of the Trip Actuator Lever, shown here, were recovered from the !
l oil sump. the pivot pin was sheared off and can be seen lodged in place. t l
.l !
I ,
i t
i i
~. ~
es mg -
M- _ ._
photo 17.
Shop examination of the overspeed trip d- see,in the turbine shaft piece, reveals the extended (jammed) position of the trip weight button.
,4
\
photo 18.
Shop examinatiot of turbine shaft piece allows viewing of the overspeed trip weight, visible near the point of shaft failure.
i l
s.
.o N
V 5 I ...
! c\
sL llyL ' s-';; J
! . =J l ; --
' 1
_m.. . }%,
1 ti t' l l
.d i l 4 .
photo 19, 1
Shop inspection of the overspeed trip weight adjustment, close up, 4
I& ,
i t.
j '
1 l
4 i
i
^
f j
j _ _ _ .. .
j photo 20. L l
A close-up view of the overspeed trip weight illustrating the extended (Jammed) l i '
l position of the trip weight button.
I s
s
l l
l 1
l vi] , M n. , ,, .: l
- n. c' :- ~ .';* i w
- '?
- ? e
- . qo;
- i y(gLp:.}ii
, i' .
t, . .
.a -
. '- m ,. S ,>.a[p;i :e
... t
,;. . ;;; y:1g
~ -
, ,, .t :
l ;, l: : 1. -
i .
)" . ,. u...
l rg . > < ,g '
t 1
f i' . i
.i L
% i: ' ,
eg.p; ,
lLk_ lll ' .
l l photo 21. !
l Overspeed Trip Actuation lever showing pcening caused by the counter l weight button. i
.h? -.
) l
~w . l0;41 ?
}
, 8t 1 l l
ll I
t I
i l
I l
1 I l
.AN .
j photo St.
1 l The Overspeed Trip Actuation Arm shows signs of huvy contact from the trip l
weight button.
r
if ONTACT V POINT OF C MVOT STOP i 3
\ .
/
- r. _
yry- .
CONTACT POINT OF VALVE STE d
9 4*
photo s3.
The backside of the Mechanical Overspeed Trip Actuator reveals signs of the arm pressing into the trip valve stem and pivot stop pin.
4 SHEAREDp LINK PIN .,
photo s4.
he e nne tion p nts of th et o rods.
(
'[ ,,
T . .
I ,
l k ,' l i .
- 1 e' $
~ .
{
< '?
.+
l l
i i
I i !
1 I l l photo 25.
A slight bend of the manual trip rod can be seen on the Overspeed Trip valve.
- i 1 j 1 l f i l l
1
,f i
i i
,f I
, 3 ..
c . . ., .
j 1 1 -
! photo 26.
I
! The minual overspeed trip rod was found in the tripped position .
(View is of the west side of the turbine looking east.)
f,
Th3 turbins rotor shaft bearing shoes had been ejected and were scattered around the south end of the turbine enclosure.
A piece of the south end of the turbine shaft, 20 inches-long by 6 inches in diameter, was found next to the upper split case' bearing housing. This shaft segment was produced by failure of the shaft where the shaft penetrates the high pressure (HP) casing and at the failure at the overspeed trip l mechanism. This shaf t section came free of the bearing housing. It was not traveling at a high rate of linear or l rotational speed at the time. This was evident from the
! minor damage to electrical conduit in the area where the l '
l shaft piece spun to rest. This is also reflected in the shaft segments close proximity to its point of origin. The shaft piece shows a slight 4 or 5 degree bend at the break
, next to the fiP casing. All inlet besring temperature and l
l vibration monitors were broken and found scattered about the l south and of the machine.
(Refer to Photographs 27 through 30.)
i i
I l
i I
l l
l S3/TGANA/a 3*5 l
l l
b
mwe em.
- 1 t' '.
N $,',. ,,,.,
~
,,<,e .-
/h '"
2' h
- y '
<- / .
, / I
/ -
/ -
,(j'
,e
%) \
[ .
>.~.
.{ .
, , c ..
4 f '
.,,b y
,(V \
,, p-
. .. 'L*4.
photo 17.
A close up view of bearing shces found ejected near the south end of the turbine (looking south). >
- 4
' ~ '
- ' .. ens M. g - -
~
~ . - - - . .
~ -
--__.7..,_._.
1- N ~,.
_ ]
il \ \ l p3 ./;v ,/ ,(( g )'
/ \ < <
g A, ' , .-
J ' l,' % , - /
- 4 A , . *- '
% <- / ,- ,
l 1( '. .
'I* 8 N ,(t 1
' l 1
. c:- .
t ,
g ..- ; ; h
/ [ i%w%+ 9 f '.
k l'i t e f- } (
N , * -
, , L !
photo 18.
The final restinr. point of the 20'long broken shaft end was in the south end of the turbine (looking down and to the cast). Spin Marks are visible near the the broken shaf t.
____. _ _ , _ . . . _ . . _ _ _ _ _ . . _ _ . ~ . ___ _ . _ _ _ _ _ _ _ _ _ _ . . _ _ . _ _ _ _ _ . _ _ _ . _ . _ . _ _ _ _
i y ._~
l
~%
e t'
N !
- .s s,. , r. .. m
+
ff l a .- ;
I photo 29. .
A close-up view of the turbine shaft (part G). Minor damage to the conduit can be seen in the background (view of south end of turbine looking cast). !
i l
l p' ye- ./ .
y -
\
- .. s
?* i
.- - ~~ ,
! .- .sN, \[ l
/'
. ', f" . 1 4 !
- e
.)
( ,,
.{ . .A\ f 7 i
.%. \'
/
i .
k ., r ,
c@r g 'l ?$* , %'
+
! C . ..s . l photo 30 j
A close up view of the 20' shaft piece (part G) reveals a slight bend angle (view is of the south end of the turbine looking down and to the southeast). i I l
, l I l l
Tho n2rth end turbina damag2 will now ba d:scrib3d. The turbine shaft failed 22 inches from the north end of the machine. Photographs of this missile can be seen in Appendix D, Photographs 24 through 30. The upper section of the outboard bearing housing broke loose. It was found several feet from the lower bearing housing. The bearing housing bolts on the west side were sheared and broken off. The bolts on the east side had been pulled through and stripped out of the lower housing. It should be noted that this is opposite the load and failure mode that was seen at the south end of the turbine.
The north bearin5 support pedestal bolts ware broken off at j the p.destal connection to the skid base. The lower bearing pedestal had shifted forward and to the west. This displaced the shim packs around the pedestal base. The bearing cil piping was intact and had restrained the lower bearing housing. The turning gear, turning gear drive motor, and gear assembly ware found inside the turbine enclos are I northwest of the machine centerline. All bearin5 shoes had been ejected and were found around the north end of the turbine enclosure.
I (Refer to Photographs 31 thru 38.)
i l
l l
l i
S3/TCANA/a 36
(
,. - - - ~ . . -
. _ _ . . , - . _ . _ . _ , _ = , .
q;s' '\
pp
~
- A T'
}
- + . - - l
.lp p
- 4
/ y i. '. ' p
- s i . 3 phc.to 31 The bearing housing is shown in the as found position in the north end of the turbine looking west.
-4 .
f ,
{ , . ap . s. , . s - As k e
'h .
-d .
~ , ,
N ---. .
5 .
V p ', y, :
,' L '.*
the to St A clok up of the turbine main shaf t and bearing in the as found condition (siew of the north end of the turbine looking southwest)
I
p m- .
, ,, 3: MP END s%
- s. r'"~- L
'j,'
'E. -,J ~
- i l . . . . ,
- , g., -, -
1 l
1- 3
[ l c ,
- c. ' j' '
h A Q. ,
s<
\ l' " .
i;. .
O' .
1
(: q' ~~ C'; *g
. # "e ,
v -
i
,c ; ;
r.t r .
, .'.. , :* ,y s-v , ,;e. ;. ;; . _
_t .,
p 1 3_', '
i 4,. ,.
i j - N ,
photo 33.
Looking down on the lower bearing housing, after removal from the turbine, f i
the west bolts (on left) are shown broken off in the lower housing. l 1
I l 1
i ,
+
i I
i' ac, 6
yml i
photo 34.
j A close up of the cast side of the lower bearing housing shws stripped out bolt holes (view of the north end of the turbine looking down and to the west).
]
I 1
i __
.__-- - - - . _ - - . . . . = . . - - - - . . - . . . - .--- - - - - - . _ - _ - . _ _ - _ - - . -
p 3 , . * '
} .,
(, 1
. *Y,
- I '{
e 3
) BROKEN h
BOLTS
~[' '
f%
l
- y
- . e j
[ -
- l 4
I photo ss. l l Broken bolts were also found on the bearing pedestal (view of the north end of !
l the turbine looking west). (
1 r A
i -
i l !
k l l ,
! _ . .c. . L l phonese.
-l f Movement of the bearing housing displaced the lower bearing pedestal shims (view of the north end of the turbine looking down and to the southwest). !
! I
- _ _ _ _ _ _ _ ___ _ _ _ _______J
- .A 4
' ~
'( 7
(. t ., 5 s
b .
~ kh
. .;, s
}..,
~=
, [.
, [ i?" '
t 4'
M ,
,S s.
s ,. , ~
fg
- A- - ,e ff
. , l is '
l
% as photo 37.
The as found positions of the turbine turning gear (part D) and assembly l (part E) located in the north end of the turbine (looking down and to the west). [
I
' j
- r '1 , l, .f
'} !
' W L
- n 1 y ;[*, -- /
't i
ji .,' $ j >
f't '. '
'. i -
- - p photo St.
North end of turbine, view from the east looking down at the lower bearing t housing oil pipe which restrained the lower housing. }
I f
Moving to the east side of the turbine, the pump to turbine hollow coupling spacer was found on the bearing oil pipes. A triangular piece (4"x4"x4"x1/2") from the pump end of the hollow coupling spacer was broken out of the spacer and was never 1.ocated.
The LP stop valve was found fully closed. The closed limit switch was triggered and the compression closure spring :
i extended. The spring was visually checked and preload ;
existed. No damage to the valve or spring closure assembly was.found. The hydraulic piston, used to open the LP stop ,
l valve, was not connected to the valve body. The four bolts I which secure this actuator had come out. Examination of the bolts showed no damage to the threads ever the 1/2 inch of evident. thread engager.,i.? Vibration during the failure appears to have shake.s th + ' bolts loose. Examination of the
, valve seat, stem, a-; p'og showed no abnormal indications.
I The plug to seat contact line was even and clearly visible.
This indicated positive closure forces were acting at the ,
full closed position upon seating of the plug.
(Refer to Pnotographs 39 through 44.)
$3/TCANA/a 37 l
. .-. _ -. - - _ - _ . - . - _ . _ - ~ _ _ _ - _ . - _ - . . _ _ . . _ . _ - - _ _ . . . - . _ _ _ - . - _ _ _ .
t l
i 1 l
't i
photo 39.
The hollow coupling with a 4"x4*x4" piece missing from the top can be seen l l at the north end of the l i turbine (view is looking northwest)
I l
l 1
- \.l;> .p-4b ,,
. >, ul .1 l
j
- l l
) {
I !
l 1
l j~ A ,- ,; i ', I l j p
^ -
'i y? ,* l' l l J
l+m- ,
l t
2}[ s t l
l
]
photo 40. -
- The low pressure stop j '
f-1 valves on the east side r I
I of the turbine looking @"
j northwest. . ;
l !
i % i t
! e" !
.' [
l 1 - _ _ __ - _ - _ .__ - - - - -
l
i l l
y -
p ~ '
l
/'
3
. . s l
ua .
i c
, l
~
% l
.I I_ -
, = - > . . , ,
d 0 - I I
3.5 n l l g
'h* '
'g s i phots(1.
A close up view of the L. P. stop valve limit switches in the as found closed position .
k'
> t' photo 42.
g., A close up of the I..P. stop valve hydraulic operator shows the bolts connecting the actuator to the push block are not engaged.
\
).
- (
k
I l
c.
8 -
! L I
'/ t !
l i
I I y ,,
i 1
. . .,_f.
1 1 ,t
,s' t
'f {, f f
.y ;, y >
l '
i
'; s
,- . 4 ',
, }t 4 ,
'. ~
I. :
' ~~
s .l l t 6 photo 43. ;
A detailed close-up of photo 42.
d
! 6 p, -- gy.7- ' - ~ m,3 1 l
- f. gy . (C 4 j 3 .
Z w. ,$.
r !
. efp g t ' :,
' ; q',' :
~-;
~
W 4
. .g _ !
l
'b.!
- *}$. ,
5 !; }4\
l I ~ .T y
. . ,: h, -.? . ," , I i ,9 g .
w
(- p, .
3 .,,,g- 4 Q * ,. ,
.. ,, # ? 4 a , ,
w 4' ai ..
9/ -
affffF h.
i 9' ,, ,
, .f., , .
e
- 1 a
, , .s
- - + -dE
. .b photo 44, 1
- The stop valve actustor to push block connecting bolt had .5* thread i engagement.
I i
.1
- g. W"P
i Moving to thi west oide of ths SCFFT,'ths HP stsp cnd -
governor valves were examined in place. The governor valve stem was extended from the body indicating a full open ;
position. Upon governor valve bonnet removal no stem, seat, '
- or plug abnormalities were seen. t j
i (See Photographs 45 and 46.) {
t
.i ,
t I
I I
r 1
1 l
P
'f
) t 4
I I
a j :
?
l !
r I
i i
t I
l
?
i S3/TCANA/a 38 i i
?
. r
._ . _ . .__ _ ,~., ._ __ .. _ . . - _ _ _ , _ , _ . . . . _ _ . _ _ . _ - , . . . . , _ _ , . _ , . _ . . _ , _ _ _ _ . . _ _ . _ _ _ _ . _ _ _ , _
, I I
ll l
- y. 31:'
~* .
I e
t . .-. r ,
cf 4
}
'M. -
,- '$ k
+. l 1 l 1
l 1
j photo 45.
, l
! The ll.P. governor valve limit switches and valve stem in their as found !
j position (view is of the west side of the turbine looking cast). ;
I !
! I l
b { t l j '
.Y l
l ' .&
. l
( + *
- n. -
j ,r,' ,
,. j* .
i . ,
4 l ____
r s
, s ;
8, P "*
w :- ;
f 1 1 -r 4
- l
% 1 j e ** '
(, f ,
- ; s d,.
,.2h 0$ .
L 6 2:3 j photo 48. l f The ll.P. governor valve stem in the as found position . j l
2
Continuing the examination of the west side of the turbine, the HP stop valve was found closed. The HP stop valve
- closure springs were examined and found to be relatively loose. The spring could be easily rotated by hand about the t
,, attachment pin on the pivot arm. (See Section 6. Figure 4 of Appendix A, and Photographs 1 through 3 of Section 4.2.2.3 for a description of the valve linkage.) The spring could be rocked on the lever are spring pin with one hand. This indicated little spring load existed in the full closed position. The HP stop valve springs of the undamaged turbines exhibited similar low spring tension in the full closed position. The coils separation was less than 1/16 ,
inch. Approximately four ster threads were noted to be exposed from the stem clavis. As discussed in Section 4.2.3 on root cause, this represents an important aspect in the '
turbine failure.
As part of the inspection of the turbine, the HP stop valve actuation lever arm was freed from its west fixed pivot point. The lever center pin was removed to free the link l
arms at the lever arm. Rust was noted on the lever at the i link center pivot point upon separation of the lever from the stem links. Although rusted, evidence of pivoting was noted i
on the lever center where painted surfaces had been disturbed by rotation of the south end of the links. (See Photograph [
52.) The valve bonnet, along with the plug, stem, elevis.snd i
links, was removed as one assembly from the valve body of the I HP r:op valve. Unsuccessful attempts were made to pivot the links on the stem elevis. Corrosion was found to be binding r the link arms to the clevis over the links' large contact ;
area.
Valve stem galling over a 120 degree arc for 1-3/4 inches of the valve stroke was noted when the stem was retracted. '
Although galled, the stem moved freely in and out when pushed ;
S3/TCANA/a 39 r
by hand. Subecqu:nt testing d3scribsd in Section 4.2.2.1 found this corrosion to be inconsequential.
The fact that the stem was galled but not binding led to the possibility of stem side loading during valve closure.
Due to the corrosion found on the links, the potential stem side loading and the slack spring condition, Kalsi Engineering Inc. was contracted to perform an independent evaluation of the overall valve closure forces. These conditions are addressed in the Kelsi Engineering Report in Appendix C.
The HP stop valve, plug, and seat showed evidence of uneven seating. Low seating forces were evident by the light contact indications. A detectable plug seat contact line of only a 30 degree are at the lower seat edge was visible.
(Refer to Photographs 47 through 54.)
S3/TCANA/a 3-10 l
> -6
,, ,. + ,.
.e !
e } \.% '"
' ');f-'
~
j
$ s .y e a 'q 'f,
.r_ 4,- a
- h. l ,/ $
/ 'h-g, #f ,.
l .: ~
,/ .
- x. ..-
photo 47. ,
The ll.P. stop valve linkage for the damaged turbine is shown in the as found (full closed) position (view from the west side of the turbine looking northeast).
l 1
, . }" ?" - .
- [ s
-s l ,
- }t; F , . .
pn 3
- s- ,.
(g "# ** * ..
's
[ '
e
,. h k /
photo 48.
The II.P. stop valve linkage for the damaged turbine in the as found position.
The springs (lef t center of the photo) pivot on pin indicating low positive spring closure force. ;
_ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ . . . . _ _ _ _ _ _ _ _ . - . __.__..______.__.___.__.m. _ . _ _ _ _ _ _ . _
l
. ,3
. .g.t. a i a Y
i 3
, f <. ~ .
i C/
( s .
% g ., \ %
- 1.
- I F 4
r
) t he-to 4)
The ll.I'. stop valve closure springs for the damaged turbine in the as found (f ull elosure) position, af ter manual isolation vals e closure. Visible is s cry slight coil separation indicatise of low spring load.
'~
- ~
\
J
~ w 3.- .
k' h g \
l i p ,
~ a o .,
i I
[ l l
}
)
w . _ , . =, g,., y
] rh u so ll.I'. stop satse linkage shown disassembled during inspection of damaged turbine. Rust is sisible at the center les er arm pis it point hiew of the w est side looking eastt
I i
i
\
, f, ;r",s'tY _..:e N g . + *
\
n J.T ; } 'at .
' \
- ~ . .nM ]:vf.C . . y p. >
. ' I . *gh, f.i's " 'sr ;>' ;,h ,c ..
.T.^ . ~
1
.~s- ~ , .
, *y .*
. . ' I.n.(
' ~ w
'* - .....e..
,',.4., y- '
I' ;
e
=
- v'
'n, ,
j[. ,'
h = l *
- e' t. , 3' o s
, ; l6 ,;.'- ^.{'J ' . .x. .
, gI. '
\
\
a; 4. . .* ,f
- gs: , .
- g \
' .s ,i .
- t. : .y. ., ,
g
- n ' i:-T'** <..
[I
.;.a y -
, w
'..p.,se
! i,f. r, i' .f i
' . *. ... } ,c. * -g*
,,e, ; .
t '* , .
, ? .'se p . e*
t' '
i ' 5 ,'hg ' n[,2.? .- k.
-? . .l '
L;,, s- ~,* fo r, *. .
k
' ".3,, ,- l ,
[
. i
.g f
. , f s, . is
, a -
2
. 1~ _;
- i, a ,%
- ., # 'g
> Js .g .
'y,QT.; js' y
t i .,, J *s ' *?
r *
\ 1 .,'.. , .
- . }
\ .
. /. ) .g. t e \
. pho40 $\'
1
\CN\' gigkS ghttC thcy C\0S*** cott0550" yigib\C 0 thC\cNCI gg \0 SI $ i wCtC (to'tCO to thC CgcNiS- \
I l
\ ., * \
, \
e
\
\
\
\
\,
\.. \
\
{
\
1 .
\
. \.
e
\
i
\ ..
e
\
\
\
t \
\
i
- goso %I rst,c\ose-09'
\
Tbe gy. s\09 ys\NC ginkSSC \cNCt STIO gggtCI iv01 t
I i
l
n~,. ..g .. . ,. , .-
.g ,, a w
, , ,m ,
s ; e e
.# , e e*
f . y ,
3 y y W.
~g *
, .7 y , # -
,, s + ,
- 6
- p. .
s t
s 4' sa
. re>7. ( .X . 7-
. +
Q. i j+,s.g ,
s
- i e s 4 ' . , i c
+ .
't
. ) l
? ,/
,gp ~
l
~. .. ,
L j?s y
- L s . ,k %. .t . , .
[h'to 53 Galling can be seen on the ll.P. stop valve stem for the damaged turbine.
y
' l
.. j
. l
. j e r 4
/ .
t
)
~
e
~
r
[. .
% . rp 1
y -
$4 !
5 ': l u:, '
[
g heta $4 A close up of the salse seat in the ll.I'. stop valve from the damaged turbine, ,
l Prict to sepcration of tho casings es p:rt of ths inspcetien, L l
- the LP turbine upper and lower casing were noted to be I l j dented outward in the area of the sixth stage LP blades. '
! As photographs 55 through 59 demonstrate, the potential for [
serious personnel injury and extensive damage to the l adjoining deserator and main turbine generator existed. This was prevented by the LP casing which stopped all thrown blades. -
l l
l Also noted was an indentation and cut on the upper east side I
l of th6 LP casings. This apparently was caused by the l segment of the turbine shaf t which separated. .
1 >
l The gland seal had broken off the face of the LP case. The f upper bolts were broken off in the upper LP case. The l lower bolts were pulled out.
{
l (Refer to Photographs 55 through 62.)
l l
I
! t l
I i l t I I h
i
! i
+
t I
t l
S3/TCANA/a 3-11 b
+ i v s -,
l.. p I
..>%.C ,q g- ,-j.
b p .g.
.'I p; a
\ . n ;
l E '
photo $5.
, w .. >
, , , ' C Lost paint and dimples are i j - . _ , evidence of L.P. blade impact. '
d
~
.. on the casing of the damaged f .'d turbine,in this photo looking cast.
JL l I i l
i
.i t
l l
1 l l i
! I 1
,.,c-.- -
1
. . ~ ~ ~ ~ ~
~~ ~'
%,4- ..
j I
J a
I j %* ' . . .; 6
-f .
. 's ' 'O , l s s >
k.' .
l l
l 6
photo 56.
A close-up of blade impacts on the LP. casing near the top rupture disk.
I
l l
l '
1 4 .
. 4 N _
I . ,
~
l i
.. i
! .. (- ,
5 !
if i
,o l .c '
y.-
1 4
photo 57.
. A close-up of blade impacts on the L.P. casing near the split case center line.
i 4
4
) .
\ .
}-
$ ,,n, ,
's pi
, 4 [~ l
.l
~~
\ [< !
\
$j
[ .s ,
t .
3
\. -
.. ,a s
, W. -M, d
photo $8.
i A close-up of blade impacts on the L.P. casing near the split case center line 1
(view from the east looking down and to the west).
1 C _._
O
.- -- y.= , , , l j 1-
., g, .
. r t
e ,
e .
, e s-l, i
1 photo 59.
A close-up of blade impacts on the L.P. casing ocar the top rt pture disk (view from the cast looking up and to the west). l 5,%.
'6 9$%
a
. i '
?I
,, , f'.
- . ?
, '9
..,a.. .
. .. ?'...... 5 i. I photo 60.
Initial missile contact on the 3. casing of the damaged turbine (north end of turbine looking .p and to tne east).
l V -
- l
~:
1
\
l
- V l t- , -
N i
. . N l
[ *
[
l
+.
i I
i i I f
) photo 61.
l The turbine gland seal mounting bolts broke off in the L.P. casing upper l half. The lower half bolts were stripped out (view of the north end of turbine looking south). ;
[
l i
p :
I w e.,
i
~
i l photo 63 f The initial removat of the damaged low pressure turbine casing. l (view looking southwest) j i
I
(
Up:n r:nov21 of thh LP uppsr ecso, no lesso debris, metal shavings or filings were found inside the LP turbine. All broken pieces of the fourth, fifth and sixth stage blades had been blown down the exhaust duct toward the condenser. All the LP blades except two were broken off several inches from their christmas tree attachment to the rotor. One of the two remaining blades, blade 53 was broken halfway in the '
blade root. Photographs of blade 53 can be seen on Fage 7 of the southwest Research Inc. Report which is Appendix E. The l other blade stub was found in the exhaust duct. On thfs i blade, the christmas tree had been stretched and plastiaally ;
deformed, but the christmas tree had not fractured.
HP/LP diaphragm was in two pieces. One piece was found [
between the fourth and fifth stage of the LP blades. The l t
other half was found lodged in the turbine exhaust duct sose twelve feet below the turbine deck. [
I Most of the LP sixth stage blade stubs had moved out of the l
rotor approximately 1/2 inch axially in a northern direction i toward the pump. Some blade stubs had moved out several I t
inches. The LP blade stubs remaining attached to the rotor i i
were bent and worn. All were freo of oil or metal shavings. .
This provides evidence of steam flow through the turbine throughout the event in that a steam cleaning effect had l
occurred during and following the damage. l r
The sheared blade retaining pins showed evidence of a violent i movement of the turbine rotor and shaf t toward the pump.
[
I (Refer to Photographs 63 thru 68.) '
l t
l I l r l l l
S3/TCANA/a 3 12 I I
I
- - , . _ , _ . . - . _ , , _ . _ , ,._-.,,,m_ . - . . . - . , , . , . , . ,,....-_,,m,, , . , . _ _ . , . . _ , _ . . , , _ ,
d il..,)
- ,y y m e m n. ,g_
photo 6s.
After removal of the upper casing of the damaged low pressure turbine, two blade stubs were noticeably missing as was li.'If of the ll.P/ L.P.
l
- ..~
V .-
- s _
1 1
l *
~* . ,
.~,
- .M m ,
P .
An
'g \
'~
Ry . . .
k T, ,m photo 64.
Closer inspection of the (Iow pressure turbine) sixth stage blade stubs reveals the blades have shifted out of the root connection, toward the pump (view is looking south).
_ __ _ _ _ _ i
1
% 9 '
1 l
k ,
1
% *m
- A 1
t -
2, v e
i i \
t' l
photo f>s. l Detail view of the sixth, fifth, and fourth stage low pressure turbine ,
blades looking south east from the west side, t
A- ~
i 1 g h ' ' ',
g,, .-
ah ;
~ -
o.,
! < j i: ; -
y
-_ \ ,
- ',,3.isu.-
- t (b!O
,- - O * * * ;O _g je.:9 _
.m o m, t heto f4 The sisth stage blade stubs had loosened where many could be removed by hand.
I i
l
, <~ # # m 's
.g f
- l w /
l :
1 I
t n
h t
/ - s G. % h;;
dp i
- /' }
t
) I photo 67.
! A close-up of the fourth stage low pressure (lower left) and third stage l I high pressure turbine blades. The third stage appeared undamaged. The {
blades did shift, up to a quarter inch, out of the root connection, toward j the pump. ,
n ,- .
.s T
- ( (,lf.l( i I
[
I ,
l J I j -
+
] ,
O
- g. .
f l
i?s. .,
v Yll ' : . & '
' \
photo 68.
. '_1 #
llalf of the II.P./LP. diaphragm between the internal supports of the damaged turbine's exhaust duct, i
1
1 i Tho b1cde ends from ths fcurth. fifth end sixth secgs LP turbine were found strewn along the sixty foot turbine horizontal exhaust duct to the condenser. Some pieces were found in the condenser. Twenty two condenser tubes were plugged due to damage.
The south side of the first stage HP blades showed evidence of rubbing. Half the first stage HP blades had shifted 1/4 inch toward the pump.
(Refer to Thotographs 69 thru 73.)
S3/TCANA/a 3 13 L .
l 1
l l
1 l
l
. l l \
l l
i t
! i i
i
! l i i l
l I i I l
1 I
)
t I ,
e< !
I \
')
~
! ; ,1 *)?
l i 1- l 4 ; i
' I j <\ ;
/ ( \ ,1 (
4 ,
3 ,
- i .
l
- 5
. I I i
! (
j h, . y):
, , .4,- e 1
}
\,?. ,
l phote 69 l
1 Pieces of the LP. turbine blades were found in the horizontal run of the [
exhaust duct, which leads to the condenser. ;
l '
i I
I t B
t t
,I F .O
'h
% r sl g -
. ,K
- 8 4eeg
'L . s f5
%. ~k . / '
I. "
x4. .
, ^ 'W
/
p p
) - ~
.a ,
,A ,
~
R' .
l &'-
m ., , .
g_
's,. - .
,;$ ~.Q 4 -
- ~
, . a ,
photo 70.
Close up view of low pressure blade fragments in the horizontal run of the exhaust duct, photo f1
. . ,~ Cr .U-yw.t; gtC '.,
. g !.lEf.
- 's f +,phan; 99 '.
.g 4
, , . f:
6p,
~ Q, . ' , 4, j h'f s . , 4
. cs
.L
- bG
l l
.) ,
' ,. , # ' s
' or" ,,
,s,o - .,
l i
i e
\
photo 72, ;
Slippage of the first stage turbine blades from the root connection is visible on the north side, r (
A . .-
I i
,,- RUBBING LINE
]
i l ..
- A
~/l+
P*- ~
\
photo Ts. l f '
! Signs of rubbing on the first stage of the high pressure turbine rotor are visible.
1
l The intorier of ths LP uppsr ccso show:d ovidenco of b1cde pieces impacting the case. Several blades penetrated the upper and lower exhaust hood. However, the outer case I stopped these blades. Blade rubbing and cutting of the hood was also evident. Rubbing has plastica 11y deformed the metal I
of the 1.P fixed nozzles. De second and third HP stage '
blades were relatively undamaged, nose turbine blades and
! shrouds were intact, he seal rings machined in the north end of the shaft rotor !
were intact, he rings were rubbed off the rotor on the l
south end, he HP fixed blades were blued and the metal '
plastica 11y deformed. These blades had provided support of I the rotor, as it rotated, after the shaft failed at the bearing. '
s Blade impact was evident on the vertical section of the i turbine discharge duct. A one inch diameter hole was l'ound i i
in the middle of the turbine to exhaust duct bellows.
I (Refer to Photographs 74 through 37.) !
i i
I I
I l
(
I i
i i
l 53/TGA.A/a N 3 14 l
l r
l
a h 1
l
- h .
- s -
l
% 7.r .
l s g h...f ;
4
( !
, s
%, I t
4 j photo 74.
j The low pressure lower case exhaust hood was penetrated by the sixth stage 4 turbine blades. l l
l l ) t* , 1" I' -g
'y \ .
. ? 'a ,' l\
j i i' ! tr -
" L' ... _ 'h!'ll -
n t j4g ; pa . _ ,
g ' .
%.w y g.w f j
-. ,L l P
1 l
I ' e/ )
l l
1 ,
'{,, ted ;
!- i j j ' '
. g , ,,
i t l ?
W' l l t photo fl.
The low pressure hood showing blade induced damage. j
I i
Y l t.
.s f, .y ',
l 'I '
, # s,.
i
)( i ,%
( el "
s .
'g, N
jj i
' q. ,
+ '
9%
.J -
y j I, *.' .
g ' 4 .Q i
- ..e*.
photo 76.
5 Low pressure turbine upper casing damage from turbine blade missiles. ,
i W
P s
}.
~
(
I
\ :
photo 77.
Additional low pressure turbine upper casing damage from turbine blade l missiles. j
l *S .., s .
Y ;;. .
! *^+-
I l e l !
f 7" w A - i s ,
p .
/ , .-
. _ f*
"J g l photo 78.
The flattening of metal on the II.P. fixed blades is visible where shaft support occurred. (View is the lower turbine case with L.P. fixed nozzles on the lef t.)
9' .
k l
r 6-
~
Phote 79,
! Close-up of fixed norales of fif th and sixth low pressure blades. The fifth stage norzies are on the bottom.
l
e
= ===a se 97
[
, /A l
,F=i- 1
,'g ,
1
- t Q q
~
~(
_.?
I' ;
. I_
v.e,: .
N 1
~ ~- % -
f { ~- ,t .. !
/ a ,,,,. r i
-5% - ".7 i gM j
_ _ - - k' e M h
~~
, s . l;
&g.
)*' k ,
t
~ ,"
4% ,
ik+
photo 60.
l The LP. fixed nozzles are visible at the top of the upper turbine case,
([:'
j
?
i . ,- ,
' - " - ~
r.n .
E'V a
~
i1 -
7' -
/
4 ,
. 8 l k l
- t. ' l
~, ~
- s I
a s& .
.)
.....w l A side view of the upper turbine case. ;
I L_ _ _ i
d 1 7 3
-4 .
~, a
:' 1~ ,'.
)NUJ,h.
phooJ1 Close up of the first stage high pressure fixed blades.
, q ..py . .
- 1. g ,c , - -9 e
r* ,
. . <n , ,s . 6 - s**.~<
, ;t - S -
~
$U ,g
'r ' . ...
'e b.. '
. .;. y _
O, photo 83 t
Close-up of the third (top) and second stage (bottom) fixed high pressure blades.
g at i
~
- . j i
~
'i .
! ), $i
.o..
+.
- e. .
' p; oto 84.
The south cnd of the turbine rotor shaft gland seal rings. l
. M i ,
t ,
, i4T' i ,
} ,
1 l
r* ,.
7 >' '
I' . at photo 85.
I The north end of the turbine rotor shaft gland seal rings.
._ . - - - _ . - _ - . -. - . - _~- . -_- - - . . . - . - - . . _ . _ ~ . . _ - - . . .
i l
m ~ m 1
t
- y,I e
+ t ,
i }.
l
.. I i
- }.
y ! .'
t i photo 86.
'4 .
S, Darrage to the bellows and impact
~ ~
' ^^
'r.;; points on the vertical exhaust duct
[5 of the damaged turbine are shown. !
=.. -
i ._ _
i~
\ i
]
i 5%: \:3 i i 1 .....J ,
- u. .<. x '
n.,4 ,
l i
1 l
, 4 l'
- n. -
- 1
, .m t
d i
go epJ l
. y. ,
- m j m N k
i I t
j 7,*, p hot, 47. ;
) -
- =- , em A hole was found in the exhaust !
. , duct bellows l
, ifne > l
~
$ fyrim* .
{
! rye ',4 l 5
J L 5
J. 4 L Iw.. .1, .w . *
. . , .m i I i
- l 1 )
1 1 j c< w .
j
_ q ~ ~ ~ ~~~ .~. .- . .p....__.. , ;
1 %u.: w i !
l .
3.2.2 Steam Generator Feed Pump The solid steel 6 inch diameter pump shaft was bent 5 inches in a westwardly direction. This is opposite the turbine shaft missile path, as would be expected. The pump drive
-flange was bent, but remained intact.
The shaft extension piece nut was gouged at the corners of the nut. These gouges matched cuts on the interior surface of the coupling liollow spacer. As can be seen in Photos 92 through 97, this evidence establishes the original coupling spacer orientation to the pump drive flange. This indicates y, , the turbine shaf t was thruct toward the pump. This is i consistent with the axial direction of movement of the turbine blades.
The coupling spacer at the turbine end spline connections had a 4 inch area of mangled splines. This was used to match the coupling spacer to the shaft missile as can be seen in Photographs 96 and 97.
(Refer to Photographs 88 through 97.) !
S3/TCANA/a 3-15 i
) s A_
f
\ "
's . -* !i. '
'y : $
..Yd[
- IP 8
.x \
.. g
-I'
,4
.i,
~
\
i i
) '....s., % $\
' Y $ .
's 4, ,
44, ; '
l; m
i
. s s,
- $h 4 ,
4 s,t
,, ha ;
, , b i e
'. $4. 4. .
photo 88.
Damage to the turbine enclosure was due to the rotation of the bent pump shaf t.
View is looking west between the pump (on the right) and the turbine.
~
i l
t i
+-.
P .,
l k '
l photo 89. [
. .~
, The main feedwater pump shaft
\ i was bent at the p u m p. The damaged shaf t extends into the '
r: turbine enclosure. View from (
4
. ' ) outside the enclosure looking cast.
- l
.e >;
j' s f -,
ee .g.
, .i i ,1 .
)
4 l
1 .- -- -,.. ... . : . 2.=..
I l
?
,jW :
7, "'6 f (
l,, -
- q. ~ .
o , ,
. .,s
- h1. x.. .
l
~.
4.
A-
~'
g
~
, $j-
/.
4,n.
p-K.:: ,
l
?. *
. 4 (tu { 7r:
s A _. N ':.8<d
(
photo 90. [
The pump drive flange was bent and the inboard deflector ring was split. !
I i View of the pump shaf t looking northeast.
I e
+ , ,; ( b t .
1 h v) .
s,
- \ .,
1 i
4 f *b y '
i a '* Ij
.;> r; -
./
, .>.- . .L'N . .
-, I i
~
photo 91.
4
- llent pump shaft af ter removal of the turbine enclosure. The denting of the I 4 pump drive flange coupling nut, where the coupling nut was jammed north, is ,
j visible at the co ncts. '
4 l l
r % .
[ ,
'.e
&,t,.w -
- .; 7;,, , - - -. . x. 3
,c.
- ^
' f ,,
q
. 4 .,
1 f '
l photo 93.
Close up of the pump drive flange retaining nut with gouge cut', on the corners of the nut.
e _
>. g ,
4, h ,
N .
%'- 8
/
% < g .
'e . ..
, 4. .
.. e .. % }
A' d Q .
. S f
g.'
h.
photo 9s.
The nut gouges match up to the hollow coupling tubc. This established the orientation of the hollow coupling tube to the pump flange and turbine missile.
33 .
,a l .
\
'1 e .
i F
'a
. , j g' f '
W-J.' $ _
4 A ._ . . . _ .
4 l
photo 94.
A close-up of the hollow coupling tube (the pump would be to the right).
a 4
4
~
.. l
] &. .
i
! i
~
-3 e i
, l
~
- .. ,. i;, a e .
i
- J." .
n g
1 p
' l s
- \ .
\ .
~'
< ) .- . . ..
e .
I 1 - -
'- () ,, .- ,
j photo 95.
A close-up of purnp-end of the coupling tube shows the nut gouge match marks.
i 1
i
... - l l
' ~
l h
j
' (
O J .
g.
photo 96. ,
! The turbine shaft missile oriented to the hollow coupling tube. l l
.[ I l .
r~
l l ,
. A b a .
p >4
/. 't . , .
! . 1 !
s/
i'
' I I
4
,g M Wan photo 97. I I
The hollow coupling tube out-d-roundness damage resulted from radical ,
' separation from the turbine drive flange splinas.
1
.. ~
Th3 pum} inbacrd d3ficetor brcss ring w s aplit, but was found still on the shaft. The radial bearing housing had broken off the pump shaft and shattered. The pieces came to rest between the pump and turbine.
The 1 inch diameter bolts, which secure the bearing housing 1
to the inboard pump case cover, were broken off in the cover.
i The bolt heads were found scattered around the north turbine l enclosure. Several bolt fragments were embedded in the l enclosure wire screen where they had impacted with I
1 considerable force.
The rotstion of the bent pump shaft had damaged the immediate area of the turbine enclosure south face. The tee steel
( enclosure framing and outer steel skin had been battered by l the bent shaft rotation.
l The pump drive end throttle bushing had cracked along its horizontal length. A piece of the throttle bushing had peeled off where it extends out of the south pump case cover.
l t
(Refer to Photographs 98 through 101.)
1 l'
l I
i l
l l
l S3/TGANA/a 3 16 1
I
'~
f _w 4 () '
N
- p. k '
/
~
, g
-: \
, .9 s l s ,
- ' % j l
1 5
4 i f .
d-photo 98.
Close-up of pump cover bearing housing.The retaining bolts broke allowing the bearing housing to separate from the pump.
r~' ~ r !
f t
, I e .
. ; ,+
4 -g - '\ !
. )4 +- >g '
/
.a .
ye ,
i
~x i p
'A,,, /, 4 f .
l -
A,.
i , s. '
l - y , .
- , ,/ ,
l l g (' g .. , . . ;
. t ;< . . I
- e . . l t
- 2
I
- , i f l
~<-
l e e photo 99.
The pump bearing housing was found next to the turbine turning gear motor between the pump and turbine.
I !
f l
t . . . . . . , _ m. ._ , , . . . ,
\
- - l
'[
'p. 4...34
,\ -
~
l
.\
., i
~
Iy e[g
~
7 .
. . l.:
I / ## q .
g 4., 4 - .
/
s .
4
- i'
'l 1
.
- Q, RACK . I- @
?
i l'.
(!E
., f S) ..
j '.. i
! .' S .
~
~ Mi 6 l
6: - * . 5 ,- -
g, l7 l ~ -
w
\ .- .
- 48' ,
I
- A> J~
c.- !
' s .
I - -
- f. -
photo 100.
l j The inboard throttle bushing with a full length horizontal crack.
l I
( l l
l l
I a
i
' \ ,i l h>~'*':> y .,?l _>
k
,-.--,Q.,, -
- la \
i !
i
,, "1 '; i
's .p l l
~ed 4 i t ,l l T.,; j l l
g pg'}' M photo 101.
2M Close up of the pump impeller with the outboard throttle bushing to the lef t. [
The effect of heating is evident on both the inner and outer impeller wear rings. j i
m,. . . - - - . . . . _ _
The imp 311e r sh:w d na signs of ccvitetien pitting. Th3 pump wear rings showed no signs of rubbing. The wear rings, i however, snowed signs of exposure to high temperature as surface ctacking and scale formation were present. The south end of the pump shaft was discolored due to heat 1.1put at the throttle bushing area. The north end of the shalt did not exhibit signs of distress, nor did the north bearing.
The pump flange coupling bolts were broken. The coupling flange had separated from the pump shaft and had impacted on the enclosure roof. Its bolting pattern was imprinted on the roof screening. This indicates the flange was not spinning when thrown into the screen.
The turbine hollow coupling spacer, pump shaft section, and the impeller vere matched up in the warehouse.
Minor damage was sustained by the 90 degree elbow of the main feedwater discharge line. This elbow was hit by the shaft missile.
(Refer to Photographs 102 through 111.)
i t
S3/TCANA/a 3 17 i
l l
, v i
k;
' N ,.
O
'^
[-
%. " 4 ' /, ,
4 f
-s% g r
- ' s., ~x
. J'.7.,c. M,, ,
Y pT,;> > , ,. . ]
,x M' --
+rp ..
k,.g Y.M d '...,_; y % ,.
1 7su'# *-- _ ;
<- .kQ)Tg
{ '
f
[ ~",;46
(' , .,
4 I
, photo 102, t l Close-up of the pump casing wear rings, i f I
p- -
t, .- s, ! * -
i j w/
i'
~
/
~'A 4-e su
,JW
' lc
.s
,, A . . , I e ,f ., .4
/ -(
i . , ,
T
., ,h .
.t
, ' O Tf ~s 't .
s
\
~
- N photo 103. 6 j Pump impeller south face. Shaft was cut to allow the impeller to be removed.
j i
,i i P
~-----ww_,ea,- eww w.
)
1 _.
l / !
! /' )
4 l' 1
! i
.. 5, .
. // -
t
(. l V -+
V' A ,
'% I I
. -v.
- 'g ..
i 4 , ,
/ )
~
i
{
I ,
- 2. . , .,
- .m photo 104, i
! North face of the pump impeller. The glossy area (top part of impeller) was polished and liquid penetrant was applied, i
.~ -
l l I'r;. . _
i y
?'
t .
l
! l S I
. - e
- l l
. C pt *o tos.
lh '
j North pump bearing (looking south). The bearings showed no distress and were reused.
i
e
^..; ep ' . .._ {. ;f
+.% -
. r :f ._. _ * '..: ; . . .'{qi *
- . f. , .
_ ' . . . ' . . ' ' n.
~.
~. M
..e.
- ~
n *\ -
- !,:s .
- . _ i:
=
. . - g- " . a_.# 4
',, . ;.g , 'h ' %
g 4- ,.
j F. ' *-
,.g- . , , . . ,
1 ..
i.
~
~;- ' ' '
- g. .
~-
s a F '* I
, ' h t T
h, sta.Ay i o -- N
.[
l photo 106.
North pump bearing, view looking cast.
,f * ..w' ~ N-y )' / :" .
.'ff, $I ' . ~ ,
r
\
x x
NN k
iN N
\ .
photo 107.
The imprint of the pump coupling flange is vit!ble on the ceiling of the tuibine en.:lasure nest to the knec brace.
r
. .. is s
, ' 'r q - . . . < . \
! i !
--t-, ', s*ni . ; .y
.f \.
i 1
l 'l
\ ,
-u Wl i.' t' ' j .
" g;> v;g+a- q q[: -
.>. .{
[ ' ' ^ ~* - * * ' y ,
AW ,s
. lk' ;
A:' '
$1 f
f I)' '( .
i .*- '
77 -
l, . r
,g;, ;'*? ..
l- 1C} c .
f
- l q ,7
.c, * ,y t
i
.. ..)
' r photo 108.
j Impact point of pump coupilos flange from inside the enclosure (looking west). l I
I
+
sn * > k 1 /
r 4 .
A
't.
., . 4
, : 3
- g,t y
- i. o*
,' a t. qf7';tr l*,
- ;; . ~
- s'
,.y,.,..1
- ,
- y; q g V,,
- t. $'
- g;b'~
,f. 3,; ,, _ ,
c , ,,w ,
.? e r;si ' >
{if 1
).
t.,
,},
uu a wr wts. .s
..? ' a, ;^ :'
photo 109. l Close up of pump coupling flange imprint (looking south). l t
I i ,
i
A .
[ .+.
~~
y <
W .
. pl - ' '
~
- y [ P-ef'%.. - ~
PbM U, 7 pf -
4 '
- z. e.
W e.;.
f .MISS
, , .y , . . ., .,
' ' , ' ',; ^j -z -
~ ,' 'f ,
l,ff OLLOW
.s LING SPA .
- ".s..,
,;7x,- - 4
- i BEARI -
.'.. HOUSI .
i .) * ,
- y. .
. ~,,
_ ' '*.y_,.
,r ,
. [*
- g T ' '
yQgr , fr . y, y,W-nu ,'. *f . X;,
- i s*-[," .
, 1 S$5EC&$h,D ll. % ; , ,. n , _l
photo 110 1,ayout of turbine shaf t missile (on the lef t), hollow coupling spaccr pump coupling flange,(cut) bent section of pump shaf t, bearing housing, and the pump impeller (on the right, pointing north).
I
~
Q
{5kN#Y s- *
'W ~' 9 .
s ,r ;g s ,v <.u7 j
.w
- c. . a y1, e E ' k[ i [,
+,
. .- 'y3 '
- g. . ,
'47 .' . % i 6,
! f,' _- _
f f
tn c.
, Dh y%
' '7s 5'h ' 3 '
b2' y[kj[
Close-up of pump shaf t af ter
% flame cutting of the bent section g
to remos e impeller.
I W, % ../,.
^ " '
f j
-:,fh%ewy' .
<h.~. 7/*
' pang 4 kl 'o Y ?!
- 1.t M*,">~*,, < j ,. . J. _
,; ,' , t' .A y 4
/ 1, i
f)* # ' yr**
- y
.}.,. {
- x,/
._ I
.. _ _ i.
4.0 ROOT CAUSE ANALYSIS 4.1 General Discussion on Root Cause Investigation Following the incident on the Steam Generator Feed Pump Turbine (SGFPT) on May 25, 1988, written statements were taken from approximately thirty engineers assi5 ned as test observers. These statements were to identify any abnormal occurrences in their assigned areas. Key observations are summarized in Section 2.1.
Computer data collection and retrieval from the Proteus computer occupied most of the Engineering personnel that evening. Formal investigation teams were established on May 26. The investigation teams and their scope are outlined below:
- 1. The Inspection Team The Inspection Team was charged with the contro11ad collection and shop reassembly of broken turbine components. This included inspections during disassembly of the SGFP and SCFPT. The performance of a missile trajectory analysis was also in this team's scope.
- 2. The Possible Scenarios Review Team - The Possible Scenarios Review Team developed the expected Loss of Offsite Power (LOOP) Test sequence of events. They were also responsible for developing the possible causcs of turbine overspeed.
- 3. The Data Reduction Team - The Data Reduction Team was assigned to the collection and evaluation of computer data.
They also established the incident sequence of events.
- 4. The Restoration Team - The Restoration Team was accountable for identifying isolation and restoration plans for the removal and replacement of the damaged turbine.
Details of the above teams are discussed in Reference 10.10.
53/TGANA/a 41
4
" Upon completion of these tasks, this data was used to formulate a list of possible root causes. As these efforts progressed, the need for experts in specific areas was identified. These needs were met by the addition of pump, turbine, valve, and meta 11 orgy experts from Bechtel, Westinghouse, Stone and Webster, Southwest Research, and i Kalsi Engineering.
A conference was held in the Westinghouse Design Offices on June 6, 1988, to reach a consensus on the root cause. Over a three day period, five subcommittees independently evaluated all the data collected. The five subcommittees were the Turbine, Pump, Systems, Valves, and Metallurgy Subcommittees. Each subcommittee was asked to review the facts concerning the incident and to identify the most likely root cause of the problem. They were then to make recommendations on any equipment, system, operation, or maintenance practices to minimize the likelihood of a reoccurrence of this type of event. The recommendations were classified by each subcommittee as those recommended to be implemented prior to restart of the SGFPTs and those that could be made at a later date. The subcommittee's recommendations can be found in Section 7.0 entitled "Plant Modifications and Corrective Actions". The Subcommittee's Reports and Main Committee Report are on file as Reference 10.8, It was the consensus of the subcommittees that the high pressure (HP) [
stop valve failed to close fully, stopping at a point some 1/8 to 1/4 inch off the seat; thus, allowing the turbine to overspeed.
t For the LOOP Test being performed, the SGFPT governor valve did not receive a closure signal. This was the result of the loss of power !
to the Electrohydraulic Control (EHC) System. Thus, the governor ;
valve, was not available for turbine speed control.
f l
l k
l S3/TGANA/a 42 l l
r_
I t
4.2 Analysis Validating Root Cause Conclusion 4.2.1 Pump Hydraulic Analysis Rotational speed of the turbine and pump was calculated as a function of tice for this event. An assumption of a constant horsepower output from the turbine was used. Results for the first 14 seconds of the event are shown in Figure 4-1 at the end of this subsection. The pump horsepower matched the turbine horsepower until the feedwater isolation valves closed at 4 seconds. At this point, the machine accelerated because the required pump horsepower dropped sharply.
The rate of speed increase is affected by the position of the pump minimum recirculation isolation valve. The initial position of this valve and the time of closure are not recorded or available. The valve received a closure signal when the SGFPT received the initial trip signal at .328 milliseconds.
Two benchmarks were established from computer data for comparison to the calculated turbine speed. The first benchmark was a direct computer data point of turbine speed at 6 seconds. The second benchmark was derived from computer data points of pump suction and discharge pressure at 13 seconds. These two data points were used in conjunction with the pump discharge head curve to back calculate the pump and turbine speed. Both benchmarks correlate well with the calculated speed versus time under a constant turbine horsepower assumption for the closed minimum flow valve condition. These benchmarks are reflected on Figure 4 1.
The SGFP suction pressure dropped as the booster feed pump coasted down. The SGFP was capable of pumping at highly degraded not positive suction head conditions. This was S3/TGANA/a 43
c .
shown by tests conducted in the vendor's test facility during manufacturing of the pump.
The suction condition recorded at 13 seconds is slightly worse, but comparable to, the extreme of the conditions ,
tested.
Very shortly after 13 seconds, the combination of SGFP speed increase and the booster feed pump coast down resulted in fluid flashing to ste am in the SGFP. At this point, the hydraulic horsepower requirement of the pump dropped to '
essentially zero. The turbine was under a "no load" condition and started to rapidly accelerate.
The turbine speed versus time following the loss of hydraulic
' load was calculated after 14 seconds assuming constant turbine horsepower output, and complete loss of pump hydraulic horsepower requirements. The pump mechanical losses were assumed to be 2 percent of the full load torque at rated speed. These results are shown in Figure 4 2.
Under these assumptions, a speed of 13,000 rpm would be reached at 38 seconds. The 38 second total elapsed time from initiation of the LOOP to the approximate speed at turbine failure agreed with the eyewitness accounts as described in Section 2.1.
In summary, the predicted turbine speed, as a function of l time after the LOOP, under the assumption of constant turbine horsepower output, is consistent with the available data and eyewitness accounts.
i e
S3/TGANA/a 44 ,
i b
SGFPT SPEED VS.T E MAX ACCELERATION 0 to 14 see -
WITH LOSS OF SUCTION '
l E 480 RPM /SEC f 6500 APPROACHES 6750 RPM TACHOMETER PEGS e 6350 RPM A
\ SPEED VS. TIME MINIFLOW VALVE CLOSED l
CiS TRIP SETPOINT e 6000 RPM <-
. 6000
$ APPROACHE
@ 6050 RPM
. E m O/S ALARM e 5700 RPM .
$ \ SPEED VS TIME m
MINIFLOW VALVE OPEN 1
5500 CALCULATION BENCHMARKS- FROM COMPUTER DATA FW ISOLATION 1.) @ O/S ALARM e 6 SEC.
1 VALVES FULLY ~ 2.) A FROM PUMP DISCHARGE & SUCTION j CLOSED PRESSURE e 13 SEC.
l 5000 O 2 4 6 8 10 12 -
14 16
}
! ELAPSED TIME (SEC)
! FIGURE 4-)
5,
4 SGFPT SPEED VS. TIME 0 to 40 see 13 12 11 ,
T 8
e 10 i
S 9
8 7
< PONT OF FLASits 0F PULP FLUD 6
= START OF TUBT TO PULP HORSEPOWER EWATCH 5
o 5 10 15 20 25 30 35 ELAPSED TIME (SEC)
FIGURE 4-2 L__.
y 4.2.2 High Pressure (HP) Stop Valve See Photographs 1,'2 and 3 at the end of this subsection for ,
a depiction of the stop valve and its linkage.
t 4.2.2.1 Evaluation of HP Stop Valve Capability as Designed From the results of Kalsi Engineering's Independent Valve Design Evaluation, provided as Appendix C, the original preload settings of the spring were not sufficient to cause full closure of HP stop valves.
l The original West..ighouse design was intended to provide sufficien'. opring force to overcome the stem i
blowout forces and linkage drag forces. Stem blowout force calculations were based on an auxiliary steam source with a pressure of 1,060 psig. This is the turbine rating point. Linkage l drag forces were assumed to be 10 percent of the l
l stem blowout forces or 130 lbs. A 25 percent margin above this combined requirement was included for a design margin.
l l However, during plant operation in the range of 0 to 50 percent power, the estimated steam header pressures range from 1,100 to 1,140 psig, respect ively, as seen in the Westinghouse Thermal Performance Data for the south Texas Project. The l
actual steam header pressure, when the incident l occurred, was 1,160 psig, l
1
! Subsequent testing, also showed the HP stop valve i linkage drag and valve internal frictional forces to be on the order of 367 lbs. A header pressure of S3/TCANA/a 45 l
l l
L m
?
1,160 psig existed when subsequent HP stop valve testing was performed. During these tests, the HP stop valves on both undamaged turbines failed to close with a spring preload of 1.5 inch.
The combination of higher stem blowout forces due to higher header pressures, and larger than assumed frictional forces, were too far from the design assumptions to be overcome by the design margin provided. Thus, for the condition at the ';ine of the accident, HP stop valve closure would not be expected if the valves had been' set as designed.
Although Westinghouse representatives have stated orally that the design preload was nominally 1" 1/4+1/2 inch, documentation verification of this value is yet to be provided. For a complete explanation of stem blowout forc5s, see Appendix C.
4.2.2.2 IIP Stop Valve Closure Spring Adjustment History During March, 1986, all SCFPT was disassembled and inspected by Vestinghouse as a precondition to the extension of the warranty. These inspections, along with similar disassembly and inspections of the Unit 1 #12 end e13 SCFPTs, were performed under the technical direction of a Vestinghouse Field Service Engineer.
The disassembly of the HP stop valve involved removal of the closure springs, the actuator linkage assembly, and the valve bonnet. During this inspection, the all SCFFT HP stop valve was found to be in satisfactory condition with minor external rust. The valve actuator linkages were found to be corroded to the extent of restricting linkage S3/TCANA/a 46
freedom of motion. The rust was removed, bare metal primed and a preservative applied to the linkage.
The valve was reassembled, and the component recertification document was signed and accepted by Westinghouse on October 29, 1986, to validate the extended warranty for this equipment.
The inspec.lon documentation indicates that the stop valves were disassembled, but does not indicate that the closure springs were preloaded upon reassembly.
No mention of a preload adjustment was made. From this record, it appears that no preload of the closure springs was implemented. In addition, upon review of the prewarranty inspection documentation for the other two Unit 1 SGTPTs, no mention of spring preload was found.
4.2.2.3 Post Accident As Found HP Stop Valve Spring Preload The HP stop valve of the #11 SCFPT was found fully closed after the incident was terminated by remote Control Room action to close the main steam isolation valves. Visual inspections and ability to swivel the HP closure springs of the #11, #12 and
- 13 turbines showed that little preload force existed on these springs. (See Photograph 49 in Section 3.2.1). Both the upper and lower springs of all. *12 and #13 HP stop valves could easily be rotated by hand around the Jpring attachment pin on the lever arm. Had the spring been prestretched to provide a positive closure force for the valve closed position, the spring would not have been free to rotate.
53/TCANA/a 47
A maintenance work request for the #12 turbine HP stop valve had been written on April 19, 1988, because the valve would not close fully upon a turbine trip. The HP stop valve would only close after the manual HP isolation valve was closed.
This maintenance work request on the #12 SCFPT HP stop valve was still open when the all SCFPT overspeed failure occurred. The difficulties with the #12 SGFPT HP stop valve are discussed later in Section 5.3.
The springs from the oli HP stop valve were removed and tested to develop force versus displacement curves. The design value of 400 lbs/in agreed with the test data for both the upper and lower closure springs.
Although, spring prestretch measurements of valve
- 11 and #12 were not made, a correlation can be made with actual data taken of the spring loads on valve
- 13. Actual measurements of the spring prestretch on both springs of valve #13 show that the upper spring prestretch was 0.278 inch and the lower spring prestretch was 0.144 inch, giving an average of 0.21 inch for both springs. The magnitudes of the spring loads on valves #11 and #12 can be estimated to be the same as for valve #13 because rotation of the springs on their attache , points could be done by hand on all the valves (#11, #12 and s13) with the same ease. This leads to the conclusion that the same relative amount of load was present on all the valve springs.
As part of the investigation, the SCFFT HP stop valves of Unit 2 were examined. Unit 2 yielded
$3/TGANA/a 48
r prestretch values in the range of 0.75 to 1.0 inches preload.
It should be noted that field tests, after the incident with refurbished valves and linkage, showed that valve closure would not occur with the preaccident steam header pressure vhen spring .
preloads of 1.5 inches ure espit,yed.
(Refer to Photoghraphs 1 thrvot,h 3.)
$3/TCANA/a 49
l l
t
- ,,,c
,e- A apf.
L y $.:. W
%,! u .,
\a
< ' ' ' .,4 I
! -}p - .)$.
f THREADE CLEVIS F
, ,, g m ,, ,, .,
' ~
- g, /. ',43 i
\;*\ f
- t [A '
, LINK ,-
i (h.;;.,
,~ -
ARMS
/ '
- je gSN,e. i l $
n A
- \ -'* b l
1 I
i
- photo 1. (
t General orientation of the high pressure stop valve actuation linkage and !
j springs and discussion of operation. The hydraulic cylinder (out of j l view, behind springs) extends outward, to pivot the lever arm about the !
j fixed pivot point (fulcrum), opening the ll.P. stop valve. The link arms :
i pivot to prevent jamming of the valve r. tem by the are motion of the lever i a r m. Depressurizing the hydrauw., cylinder allows the valve spring to l Quickly close the valve. .
During the recovery, it was discoured the clevis could be unscrewed from !
the threaded valve stem to increase the tension of the springs to provide l l an adequate force for valve closure, flowever, no procedures, !
instructions, or information had been provided indicating the valve was {
adjustable prior to the accident.
J l
! l i
i l
1 l l :
I f 1 l i !
4 I
i i
i .
I i !
l i
\
l 1 !
i !
i
! i l
i f
4 i.I? ; * *,'.)u, il,,
t j ;
u . ;. . , , ;w
.s
' I 1
TEM ,a :
l s -
r r
EV1 'I l l c Qj) k l
K:w l
' i I
phute 3.
I The valve was installed with the clevis threaded onto the valve stem I with minimal preload spring tension. .
push ABM ,
- u. - l j ,_
4 I l i !
1' ...
4 I, ' \
p Qf - .
~
~
a l
4 i
' \
) 's ' , a
\
t
'g, .
~ :
l HYDRAULIC f, ._ _
LEVER [
3
] CYUNDER / Q "
ARM i I
i W
- 4d i photo 3.
'r si !
i l
Typical as found spring preload condition The visible coit ;
j i'
sepseation is the as inanuf actured no load condition. Th: full closed spring pre stretch is 0.2 inch. ;
) t l
' _ _ _ - _ . . _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _. _l
4.2.3 Evaluation of the SCFPT Damage Against the Root Cause Hypothesis The turbine and pump damage observed was evaluated against the root cause hypothesis. The questions needing to be answered were: 1) Vould turbine overspeed lead to the type of damage incurred in the blades? 2) Is the rotor shaft failure preoictable upon blade failure? 3) Do , results of these evaluations support the turbine shaft missile behavior? The following sections address these questions.
4.2.3.1 Analysis of SGFPT Overspeed Nec6ssary La Cause Turbine Blade Failure An evaluation of the speed that would be required to cause the SCFPT damage was performed. In order to support this e4nalysis, it was necessary to make a number of simplifying assumptions. These assuaptions reflect the fact that there are cany variables involved and the dynamic relationship of all of these va/iables, during the incident, are not fully known. How this vibration level varied with speed and the load it imposed was also inde te rminate . The rate of vibration loading on the turbine due to pump hydraulic changes occurring during the incident are also not known.
The fact that the first three rows of HP blades of the turbine had not failed in tension was used to establish an estimate of the maximum speed reached by the rotor. (See Photographs 82 and 83 of Section 3.2.1.) This indicates that the speed of the SGTPT must not have exceeded 14,840 rpm. This value was obtained by assuming that all loads were acting as concentrated loads on these HP blades and that the ultimate allowable of the materials of the HP blades was not exceeded. i S3/TGANA/a 4 10
To determine the minimum SCFPT overspeed which occurred, the LP blade #53 overload failure was employed. This failure mode was established by Southwest Research, as discussed in Section 3.3 of Appendix E. Again, making some simplifying i assumptions (for example, bending overload was not involved in the blade failure), a shaft speed of at lear 12,792 rpm occurred. This is the speed c a" lated to cause tensile overload failure of blade #53.
In summary, the rotor reached a speed between 13,000 and 15,000 rpm. This calculated speed is in agreement with that obtained from the Hydraulic Analysis of Section 4.2.1 and that of the Missile Analysis of Section 4.2.3.3.
4.2.3.2 Analysis of the Force Required to Break the Turbine Rotor Shaft The turbine rotor failed, due to bending overload, at locations as shown on Figure 3 1. This failure mode was established by Southwest Research Inc., as described in Section 3.3 of Appendix E.
The two main shaft failure locations were as follows:
- 1. Just outboard of the gland region on ths pump or exhaust side of the turbine; and,
- 2. Just outboard of the gland region on the inlet side.
S3/TGANA/a 4 11
Making the types of simplifying assumptions as noted in Section 4.2.3.1, the following approximate values of shaft loaoing were ,
calculated: '
- 1) The bending loads required to produce these failures were: -
- 6,442,000 in lbs for the pump end break
. 3,728,000 in lbs for the inlet side break
'2) The load required to break the pump end of the shaft, incorporating the fact that the '
pump end bearing failed, (as seen in Photographs #35, #36 and #38 of Section 3.2.1) is 339,000 lbs.
- 3) With the pump and of the rotor now free, the force required to break the inlet end shaft is 57,910 lbs. As the unbalanced force, due to a missin5 last row blade, greatly exceeds this value, the inlet side rotor break can be explained. This sequence is in agreement with the observed lower speed of the south shaft segment discussed in Section 3.2.1.
In regards to the pump er.d break, it is believed that the vibration induced by the i
blade failure created sufficient bearing
- I loads to cause bearing housing failure at the pump end. Once the top bearing cap failed, the shaft then failed in bending about the :
t lower bearing housing. '
4.2.3.3 Shaft Missile Analysis Summary An independent analysis of the turbine shaft missile was performed by Bechtel Power Corp. and is provided S3/TCANA/a 4 12
a in Appendix D. This analysis established the range of turbine speeds necessary to eject the turbine shaf t segment and cause the turbine deck damage as seen in Photogre' 9 of Appendix D. The energy and required missila velocity to create the hole in the
, the Turbine Building deck was calculated. This velocity was related to the turbine shaft rotational speed. A turbine shaf t bend angle of 10 to 15 degrees was used. This bend can be seen in Appendix D. Photographs 26 and 30. The results yielded a turbine rotational speed of 12,700 rps for an assumed bend angle of 15 degrees. A speed of 19,000
, rpm was calculated for a 10 degree bend angle.
These results compare well with the preceding sections in light of the accuracy with which the angle can be measured.
l
.e l
S3/TCANA/a 4-13
5.0 EVALUATION OF THE DESIGN, TESTING, AND MAINTENANCE BARRIERS VHICH SHOULD HAVE PREVENTED THIS EVENT This Section probes the design, testing, and maintenance espects of the Steam Generator Feed Pump Turbine (SCFPT) as they relate to the SCFFT failure. Each of these aspects is first discussed and then evaluated. A graphical presentation of barriers to failure which were penetrated is provide.! on Figures 5-3 and 5 4. Figure 5 3 represents the aspects of design, testing, and maintenance which were not successful in preventing this event. Figure 5 4 reflects the aspects of programmatic control of training, supervision, and human factors which acted as barriers to this event and how those barriers were penetrated.
5.1 SGFPT Design 5.1.1 High Pressure (HP) Stop Valve Design 5.1.1.1 Original Design Basis Discussion The design of the HP stop valve incorporates a hydraalic opening cylinder working against two tension springs. The tension springs were designed to provide the closure torces.
The required spring force was based on the assumption of 1,060 psig steam pressure (at the turbine rating point) acting on the stem area of 1.23 square inches, and a total frictional force estimated as 10 percent of the stem blowout force.
To this value, a force e.ergin of 25 percent above the total of the stem blowout and frictional forces was added. Sliding parts of the plug and stem are 400 series stainless steel and are of the same material hardness. This material has a very low threshold of galling. The HP stop valve design also includes a 1/4 inch plug bypass port for pressure S3/TCANA/a 5-1
t application to the top of the plug.
5.1.1.2 Design Evaluation The 1/4 inch bypass port channel drilled in the stop valve closure was evaluated by Kalsi Engineering, Inc. The evaluation found that this 1/4 inch port was undersized in flow capacity for service as an assist in closing the valve. The intent of the bypass was to create-a positive differential-pressure from the tay of the stop valve to the downstream side to aide in closing the valve when under fluid flow conditions. The analysis showed that the resistance through the long narrow passage was much higher than across the stop valve to the downstream side and therefore virtually no pressure buildup could occur in the valve cavity above the stop valve. It must be noted, though, that this bypass port should not be counted on to provide the minimum required force to close the valve, it should only be considered as an aide in increasing the closing force margin.
As mentioned in Section 5.1.1 1 a main steam pressure value of 1,060 psig at the turbine rating point was used to calculate the closing torces required rather than the higher pressures the valve must close against. However, this approach was in error. The force requirement to ensure valve cloture against the mair, steam header pressure should be based on the first code safety valve set point plus 10 percent accumulation (i.e., 1,285 psig
+ 104). As discussed in Appendix A, at the 100 percent power rating, the steam supply to the turbine is from extraction steam through the lov S3/TCANA/a 52
pressure (LP) stop valve exclusively. It is only at lower operating power (below 50 percent) that steam is entering the turbine through the HP stop valves.
At these power levels, the pressure of the main steam is 10 to 15 percent higher than the 1,060 psig value used in the HP stop valve spring design.
The second item affecting the closing force requirement was the magnitude of friction. Based on measurements of the actual stem equilibrium position during controlled testing in which the valve did not fully close due to insufficient spring preload, a more accurate estimate of the actual friction in the system was made (See Appendix C). The spring preload required to account for total system friction was found to be approximately 300 pounds higher than the value originally assumed by Westinghouse. To eliminate this spring under-sizing on a short term basis, the spring preload has been adjusted as described in the Westinghouse Customer Advisory Letter 88 01 (See Appendix F).
Finally, the design use of stainless on stainless sliding surfaces can affect closing force margin and has resulted in replacement of numerous valve internals. First, the use of 400 series stainless steel without any hard surface treatment makes these parts very prone to galling. Secondly, the valve linkage design imposes some degree of side loading
<n these parts. The linkage assembly weight is supported by the stem. This would promote stem galling. Finally, the design makes no provision for adjustments for fine alignment to prevent side forces on the stem. This shortcoming will be resolved by nitriding of internals as discussed in Section 7.2.1.2.
S3/TGANA/a 53
v '"
j 5.1.2 Desian of Steam Generator Feed Pump Turbine Trip Systems t 6
5'.1.2.1 Trip Sinnal Discussions The SCFFT is protected by a variety of trip signals..
- The two mechanical trip devices are the mechanical overspeed trip mechanism and a lever which can >
. perform this same function manually.
l The manually actuated electrical trips ar6 from '
local and remoto (Control Roon) push buttons.
Automatic electrical tries occur on low turbine !
c exhaust vacuum, low turbine operating oil pressure, i safety injection initiated from Ht Hi steam )
generator level or low compensated steam line I pressure, excessive thrust bearing wear and high i turbine vibration. The excessive thrust bearing. f i
wear and high vibration trips are powered via ;
_ redundant 120 volt non vital power supplies.
'I
', Additional SCFPT protection would be provided by the
?
throttling action of the governor valves. These !
! valves wonld start to close on a speed missatch [
signal. The governor valves would be fully closed {
by any turbine trip signal. (See Figure 5 1 for the i
SCFFt Control 011 Diagram and Figure 5 2 for the I SCFFT Trip Function Logic Diagram.) f I
L i
4 r
I l
i
! S3/TCANA/a 54 !
i i
- I
5.1.2.2 Evaluation of Trip Signals and Power sources The LOOP Test resulted in loss of both non vital electrical sources.. As a result, the bearing wear and high vibration trios became inoperable. The th0P Test also rendered the governor valve controls inoperable. * ,
Design changes to improve the reliability of the [
trip circuitry have been made. These included i providing inverter backed vital power to the i vibration monitoring panel and the SGTPT control panel. This assures operability of the bearing wear and vibration trips. This modification also i
provides vital power to the governor valve control system.
SGFPT trips were addod for electrical overspeed.
This will provide redundancy to the mechanical e overspeed trip. ,
A low net suction pressure trip was also added.
This trip was added to protect the feed pump from t
extended operation without adequate suction pressure. 14w suction pressure would exist on loss of condensate pumps or SGTP booster pumps.
L All of the above trips are preemptive to the
! mechsnical overspeed counterweight trip mecha.nism.
The mechanical overspeed trip mechanism should not routinely be challenged during operation, but should l l
only act as the final source of turbine protection. ;
i i l
l i 1
I l .
l S3/TGANA/a 5-5 l :
1
5.2 Description and Evaluation of Tirbine Testing 5.2.1 Description of Testing Performed (SCFPT) 5.2.1.1 Description of Startup Testing As part of the Startup Test Program, conducted just prior to and during Hot Functional Testing, the following Ceneric Prerequisite Tests were performed.
Those aspects of the tests that relate to HP stop valve performance, are described below:
- 1. Prior to admitting steam to the turbine, the following operations were erified by test:
- a. All protective trip devices operated properly,
- b. Hangers for HP throttle stop valve were properly adjusted,
- c. HP stop valves tripped properly,
- d. HP stop valve indicating lights operated properly,
- e. A check was performed to show that unit did not latch up until the stop valve controller and governor valve positioner controller were run to full decrease position,
- f. HP stop valves indicating lights are operational for the closed, full open, and l 3/4 open positions.
l l g. Trip turbine with solenoid trip observe the following: (a) Turbine trip indication is given on Control Room indicator, (b) Stop i valve controller and governor valve 1
positioner run to the full decrease posi tion.
l h. Trip turbine with manual trip and observe all valves close.
l S3/TCANA/a 56 l
l
L 1
- 2. Prior to rolling the SGFPT with steam, the following check was made. The trip solenoid was ,
checked for proper function. both electrically '
and mechanically, as evidenced by tripping.
- 3. During initial uncoupled SGTPT operation, the following checks were performed;
- a. The auto stop trip was operated by remote control and manually at the pedestal. All valves closed satisfactorily,
- b. The overspeed trip was checked with oil pressure.
- c. All protective trips were operating properly in accordance with the control setting data sheets.
l d. The overspeed trip was checked by t l overspeeding. Tha hydraulic overspeed test I trip was rechecked and recorded as 5,905 rpm, as fourd, and 6,000 rpm as left.
1
- 4. Following the completion of Hot Functional i Testing, the Main Feedwater System Acceptance Test was performed. During this no steam applied
! test, the following actions were performed: [
i
- a. The turbine trips were confirmed by 1
depressing the turbine trip button on the
?
turbine console, the manual trip plunger on i the turbine, and the turbine trip button on !
the control Roon panel. '
- b. A simulated low bearing oil pressure trip was l performed.
f
[
i S3/TGANA/a 5-7 l
I
upon depressing the turbine trip push button l on the turbine console, was performed.
Immediate closure of the valves was confirmed.
- d. The LP stop valve test push button at the Control Room panel and the turbine control panel was depressed and a verification of 25 percent closure was performed.
f e. The HP stop valve test push button at the Control Roon panel was depres6ed and full closure of the HP valve was verified.
5.2.1.2 Description of operation Testina The objective of this test, as related to the SCFPT.
was to demonstrate that the turbine trips on a safety injection signal from the Solid State Protection System. The acceptance criteria was ,
verification that all SCFPTs tripped upon receipt of a safety injection signal.
The Generic Prerequisite Test was conducted during Hot Functional Testing. This test consisted of checking all the SCFFT protective devices; the stop and governor valves, solenoid trip function and overspeed trip prior to admitting staan, during operation after admitting stena, and with the pump coupled to the turbine.
The objectives of the acceptance tests for the SCFPT were:
- 1. Demonstrate SCFPT interlocks (non safety) operate as designed.
S3/TCANA/a 58
- 2. Demonstrate Lube 011 Systems function as designed.
- 3. Demonstrate seal water systems associated with the pumps function as designed.
- 4. Demonstrate that the associated annunciator windows and computer points function as designed.
The acceptance criteria for SGTPT were:
- 1. SGTPT controls, interlocks and trip / resets operate as designed.
- 2. SGTPT associated components operate within design and vendor specifications.
The Start up Acceptance Tests for HP stop valves consisted of valve closure timing. These tests were performed without steam in the system (main steam isolation valves were shut). A stop watch was used to time the closure by observing the physical valve position from start of the trip event to valve full clocure.
5.7.2 Evaluation of the Testing _of the scrPT Perfermed Two basic important HP stop valve tests can be conducted on these machines. The first is to trip the HP stop valves and verify full close each time the turbine is placed into operation. This is conducted prior to admission of steam.
The second is the closure of the HP stop valves during turbine operation once the plant power level is above 50 percent. At this point, LP extraction steam is used to drive the turbine.
S3/TGANA/a 59
i 4.
The preoperational testing was performed. The in operation test, which could have shown that the valve was nct closing fully, had not been performeJ. Push buttons for the second test are available on the local SCFFT control panel. This test,-however, could not have been performed prior to the
{
LOOP Test. The plant had not been operating above 50 percent '
power as the LOOP Test was a restraint to operation at this level. This test function will be incorporated into the plant operating procedures for periodic testing of the HP stop valves during higher power operation.
The initial test also did not include determining the effects of a loss of EHC control power. Ehc pump response time and EHC system pressure. transient times were not measured. These aspects were tested during the restoration of the turbines.
The SCFFT trips were checked during,.Feedwater System Acceptance testing by push button actuation from the turning gear speed of 2 rps. The HP and LP stop valves were verified to close "immediately", but since no main steam or extraction steam from the Moisture Separator Reheaters (MSRs) were present upstream of the stop valves, proper operation of the valves under actual conditions was not verified. This testing with steam was not possible, at that time, due to plant status (i.e., no steam was available to the SCFFTS as power operation had not started).
The prerequisite tests of the all SCFFT stop and governor valve open and closed limit switches verified the functional performance of these limit switches. Verification consisted of opening or closing the valve ard verifying that the open and closed indicating lights functioned. The testing did not verify a quantitative correlation between the limit switch actuation, indicating lights and actual valve position. For HP stop valves, this aspect is critical. A small error in S3/TCANA/a 5 10
closure can, in effect, be a full open condition with respect en steam flow.
A review of the characteristics of the limit switches on these valves, the valve stroke, and the amount of limit switch movement prior to contact closure has been performed.
Due to the long valve travel (3 inches), it is not possible to obtain the accuracy needed to verify the valve plug is completely against the seat with limit switches. Because reliance cannot be placed on the limit switch accuracy, it is essential to assure proper spring adjustment and positive spring closure forces are employed to ensure valve closure.
Identification of the Vestinghouse supplied Balance of Plant (BOP) components which are critical to personnel and equipment protection, technical design basis, and the testing required to assure that they are properly maintained, is being pursued with Westinghouse as discussed in Section 7.2.3.1.
5.3 Steam Generator Feed Pump Turbine (SGTPT) Maintenance 5.3.1 Discussion of Maintenance History 5.3.1.1 varranty Extension Inspection The SCFPTs had undergone an interim storage period during construction of the plant. To ensure preper operation, a program of disassembly was undertaken.
The turbines were disassembled, inspected, and any problems found, corrected. For Unit I this included disassembly of the HP stop valve spring linkage.
The conditions found were not deemed significant enough, by Westinghouse, to warrant linkage disassembly of Unit 2.
S3/TGANA/a 5 11
A Westinghouse Field Service Engineer provided technical direction for thic effort.
From a review of vendor supplied information, no mention regardin,g a spring closure adjustment or a spring "setting" was found; nor were any warnings or cautions to contact Westinghouse regarding these matcers found. This review included the Westinghouse Design Drawings and Maintenance Instruction Book for the SCFPT, Vestinghouse has now developed spring adjustment ins: ructions for these turbine HP stop valves.
Those instructions were conveyed formally to owners of these Vestinghouse Turbinee by Vestinghouse Customer Advisory Letter 88 01, provided as At.tachment F.
5.3.1.2 HP Stop Valve Problems with the s12 Turbine Starting on April 11, 1988, maintenance work requests were written due to problems with the #12 turbine HP atop valva. This valve would not fully close and stop the turbine. Vestinghouse Field Setvice Engineers directed a slight adjustment of the spring closure force by a one turn adjustment of the stea clevis. There is no indication that in giving this direction the Westinghouse Field Service Engineers verified that the preload on the springs was in accordance with the original factory setting, This investigation has determined that the valve was 8 to 12 turns off its original design point. The day after the problem was "fixed" and the first work order closed, the #12 HP stop valve again failed to close. A second work order defining the fact that S3/TGANA/a 5-12
the valve plug actually stopped short of reaching the seat was issued. This work order stated'that the valve would only-go completely closed after an upstream manual isolation valve was closed and the steam header pressure bled down. The implications of this inadequate spring closure force situation on the safe operation of the turbine was not l recognized. This was an open work item at the time of the accident on the oli turbine.
5.3.2 Evaluation of Maintenance History From the above description, it is obvious that the importance of the turbine stop valves as equipment protective devices j ~ was not fully appreciated.
l f similarly, had there been warnings or cautions in the
- operations and maintenance manuals or had the equipment I protective function of HP stop valve closure been otherwise 1
more fully understood, the disassembly of the linkage and springs for the warranty inspection would have prompted questions on proper reassembly spring preload.
l l
l l
l l
l l
S3/TCANA/a 5 13 l
l
A anmaan vue m
.*_C =
=
____ _ _ _ q vs 9,t ---
_ ~ ~ ~ i
== = = -
.. j= g I
_ . _ - {= %
x = . _
=
j I
=
n , i
$w = [_ l
%"I tI ^ -a.".. l -- I
.. j =
.F- - 1 I r I
- . 2- -==- l x =
--- = == i
~
i l
m =
5_ _
- l
_ =
.W
_ g ________ ____, i
_ & o- ______ _ _ .
._ a l
$ r Dd - 5 1 11
"*f" l _ _ _I g 4
- ! =
p ""
_ jXy ,I u u
_ y
~
Y
~~
STEAM GENERATOR FEED PUMP TURBINE bIL CONTROL DIAGRAM FIGURE 5-1
ll 2_
I J ! , ' I I 5 E - _ .
E
,- ."' =
-M O
S-S -
mI -
t e
s n.
R W
- I
_ _Lew- n. F I_ .- , 1_ . "*" v. - I w_.l t
a v . . e y
_I I
J_Js ,w -
V .
= O r .
' _G _ _ ar .. _ s i
e-m,-
_s
_ Er e- "A" - _
r- -
w_e.. _
s .
s .. _ een sw s
ru i
_ ci
_ na l
, l -
u _% _7 5 - _
' t
,..n.
a ta sm T_ w=
e
_s w_s rw u-r a
__e
. t E _s a
_ w_
N .w . .
s . r I . ,. _
- w. , t ,I B I g
m w
n.
RM .
. g I l I I I UA ..
- ., .. I , I I TR . .
_,l .
G .
PA MI r.r_
mu ao_
_,w*_
e s
n _
a .
UD . er ne su...
P M .
ao_
s C
E D
A ic n.s_f_,."
,w w" - v m. _
,w o .
DG I
wr t
e ,rS s
rs_ o . r _
EO m_ ,w=""'
w,0 t
9-ur s - .. .
E F
L e,9
., 0 .
ca
.t_
- s ..
e.
_ w= _
e _
n4 n .
N I
re r
es l
i_ ,_ ' ,.
R O si * , ,
OI w, D o ,0
. s t,
TT .os
- s. r AC RN 0
w ,3 C.
- a s
t r
EU U F o e
NF s u
E a xe GPI s R .
r w
e w
o MT A ,
m n
s c
r o
E T t I
. I I
=w I w
m r
s e
a t
o S - m e T
s t r
t .s T
M 3
s.
- ~ .
s a.
i 4
s s s
t e 7
b n t
,U s c u r
a
. .. . ' i
.w r. .
=. -
v . .
I , I I .
=
s .
r:
t g I l
FIGURE 5-3 (PAGE 10F 2)
EVENT AND CAUSAL FACTOR CHART
SUMMARY
FOR LOOP WITH
- 11 SGFP DESTRUCTION (05-25-88)
,. m o.Cxs PDFDf.ED TO VetFY PRDPet CDietTEDN OF 1NILVE MA38 N AcmATOR MECHAIEDS ErmDt)
SY TESTBdB Oft DISPECTEDeL/ O CAUSAL FACTOR f PResuerm CAUBAL FACTOR SFPl.XD BY y OP9t4 TEN 4L PRDSLDS OFDMENGD ,__,
I WTH SEPPT IF STDP VALVE 3. ' N l l IMAdBGERS ATTBFTED.DeeDft I f ELECTRICAL Oft f LDRS OF an== mens M=c.==tm i
- )I PtMER Ac
% TRAfe5ENT , /
i ~_,_-
v 4 - y v SWPT DP STOP .
S VE m BdtTIATED. MABI M DelrECT10B6 peTe up( COBENTHet TUWDE & - - 30057Dt > PAK 2 I Pown=ED ,,RoPei oETDuonnTrs REACTOR DUP PU r TRIP l
H M W- M 3EIGhE7+ Min 3Bd7+
35Ght7
FIGURE 5-3 (PAGE 2 0F 2 )
EVENT AND CAUSAL FACTOR CHART
SUMMARY
FOR LOOP WITH #11 SGFP DESTRUCTION ( 05-25-88 )
~
EMC CONTROL \
P0utR & Pu es g
( POWERED VITAL AC PouER /p BY NON-
% / . -
'7' % sestu EVENT r,,0, ,; s AC -R ,.OT x O cA 5- --
( mit=E DuRi A LOOP Cj -
( ,,,,3
=Su-rive Causal rACTOR T-m - -I-5 -
- 1~
a + r5ET x ,,, , ,
aPPLv s155 Twagg FEATU4E 00E5 8807 h 300 STER Pur DE5i = cE55a , CL=E c=R==
CLosm E roncE VALVE j CD,,,, ,,0,,,
g I
T' [
+ + +
SOFPT st1 teP SCFPT #11 MP SCFPT #11 STEAal VOIDINC #tt SCFP SitP VALM GOVEfestyt VALVE ACCELERATES OCCURS IN OVER5PEEDS SEtenINS (PEN HEt0&lN5 (PEN ett SCFP TD
"" 1 > > > > !aPELLER > DESTAUCTION 15:35sti+ 15:3D1T+ 15:35:35 15:35:22f i
FEEDWATER 15m.AT10N ACTUATItus SOFP et t 8tTORALIC Loa 0 DECREASES
FIGUf1E 5-4 EVENT BARRIER ANALYSIS
SUMMARY
FOR LOOP WITH
- 11 SGFP DESTRUCTION (05-25-88)
PREVENTION DETECTION DerE TBus BWErfmf F SWPT tr F SEPPT y DEL
- -r-
- -e -
Greubf3 pes sever penstf.me v W SWPT EMutn7mm senser. s Pet CBGRE Ptst SEPPT <
N s e gyVmt W MT F ret N ES8* MIL CSusteK F WWPT IF TEDestat F 1DE STW 15.W 3FugenTes
- serfs
> mestnu me iusine M EEVEW W w sePr nome serPT pueIKTWI DF STF StBEBC DeLW
> s c
-- - -- g_ - - -
l l
l I 6.0 REVIEW OF CENERAL PLANT SYSTEMS AND EQUIPMENT FOR SIMILAR PROBLEMS A continuing question, which must be addressed, is: "How could this problem apply to other plant equipment?" The aspects of this event that l
were identified as having potential cross system implications were analyzed. The rasults are addressed in this Section. The topics identified are:
- 1. The potential effects of the operation of various diesel generators on squipment protection.
- 2. Is the stop valve spring preload on other turbines on site sufficient to protect those machines.
- 3. Since the Loss of Offsite Power (LDOP) Test had affected the seal water supply to the SCPP, this was one scenario initially on the list of possible causes of the pump damage, thus, other attuations of continued pump operation after loss of seal water were addressed.
6.1 Review of Potential Impact Due to Balance of Plant (BOP). Technical Support Center (TSC) or Lighting Diesel Operation A review was conducted to assess if potential equipment damage could result from diesel power being supplied to equipnent but not supplied to auxiliary components or systems associated with this equipment.
The supply of vital power to pumps without powering booster pumps, or powering equipment without power to its controls or interlocks necessary to preclude damage was reviewed.
This review was undertaken as a result of the BOP diesel supplying power to the Steam Generator Peed Pump Turbine (SCPPT) standby AC main feedwater turbine lube oil pumps without supplying power to the Electrohydraulic Controls (EHC) or to the main feedwater booster pump. Such a sitvation, which existed because the BOP diesel was in operation prior to the start of the LOOP. might have prevented the SJ/TCANA/a 61
1 i I SCTPT lov lube oil pressure trip signal from being generated.
The results of this review are provided in the Tables in Reference 10.3 antitled "TSC, LOP, Li.hting t Diesel Running Analysis". Two !
areas identified as requiring correction were found and a brief summary of those areas follows: ,
o The TSC computer battery room and radwaste monitor room heaters are backed by TSC diesel power. The heater is I intended to trip on safety injection situations to prevent overheating of equipacnt. However, the heater shunt trip j does not have diesel or battery backed power. This shunt trip device requires power in order to achieve the heater shutoff. To correct this situatien, plant operating i procedures are being revised to rack out the breaker to these heaters during summer months, when heater operation would be detrimental.
1 o The SGFPT e2 standby lube oil pump is BOP diesel backed. '
Diesel backed power is now being added to the turbine controls end a turbine trip vill now occur if a booster pump is not running through the low not suction pressure trip.
For further information, see Section 7.0, {
r 6.2 Adequacy of Other Turbine Stop Valves ;
6.2.1 Auxiliary Feedwater Pump Turbine Trip and Throttle Valves Review of Design:
The design of the trip and throttle (T end T) valves of the Terry Turbine Driver for the auxiliary feedwater pump was reviewed to verify sufficient spring force exists to ensure positive valve closure against maximum header pressure conditions. It was found that the Terry Turbine manufacturer had notified HL&P (in 1984) of a problem with the Gimpel S3/TCANA/a 62
manufactured T and T valve springs. The springs were not strong enough to close the T and T valve under lov steam flow and high steam supply pressures. These T and T springs for both Unit 1 and 2 auxiliary feedvater pump turbints were replaced in 1986. The replacement springs provide sufficient force to overcome steam unbalanced forces with full ratd pressure at all flow conditions.
6.2.2 Main Turbine Trip and stop Valves The design of the main turbine trip and stop valves has been reviewed by Westinghouse in the past. Much of the review is centered around the low pressure turbine missile generation probability question. Probability statistics have been determined based on various valve and trip system designs and the possibility of the resulting overspeed conditions that could exist under each trip / valve system, h addition.
Vestinghouse is reviewing the operating hisiory of the main stop and governor valves.
5.2.2.1 Acceptability of the Springs on the Main Turbine Stop Valves l The main turbine stop valve springs are sized to l
l ensure closure when operating at a nominal pressure of 1.200 psi. Margin has been provided above this nominal value to ensure adequate spri.ig closure force for anticipated abnorisal conditions.
Hl.AP is evaluating Vestinghouse Availability Improvement Bulletins (A1B's) 87 14 and 87 15 that address some concerns on the operation of these valves separate from spring closure questions.
Vestinghouse has recommended the impleuratattor. of the valve upgrades contained in the AIB's.
53/TCANA/a 63
i l 6.3 Review of Seal Vater Design for Other Pumps Evaluation:
l l
Seal water is provided to the feedwater booster pumps and low pressure heater drip pumps by the condensate pumps during normal and i hot standby operation. The condensate makeup pumps provide seal l
water during startup. Loss of electrical power to the condensate and j condensate makeup pumps at the same time will result in reduction of l seal water to these putps, i
l l
l As the condensate pumps coast down, seal water pressure vill drop and l
the local low pressure alars will sound. The condensate pumps will l continue to previde seal water during pump coast down. This is I sufficient to prevent pump shaft damage to the pumps rece:ving seal I water.
l l
l The SCFP coast down is longer than the condensate pump coast down by approximately 8 minutes. Vestinghouse has confirmed that, by this
! time, the SCFP speed will be below 600 rpm. Coast down, with seal
! vater pressure declining, is acceptable and no damage to the pump 1
shaft vill occur, l
i l
l !
i 1
$3/TCANA/a 64 l
l l !
l '
l
Y 7.0 PLANT MODIF: CATIONS AND CORRECTIVF. ACTIONS As discussed in Section 4, the Five Subcommittees on turbines, pumps, systems, valves, and metallurgy made recommendations for short and long term plant modifications. Additional concerns were also identified during tha course of the investigation.
This Section categorizes the recommendations or additional concerns for short or long teru implementation. Within this categorization, the recommendatinns has e been grouped as to the type of concern they address.
7.1 Short Term To ensure that the high pressure (HP) stop valve performs its intended function by closing completely, the short term recommendations discussed below were added prior to returning any of the Steam Generator Fead Pumps (SGFPs) to service.
7.1.1 HP Stop Valve Operability The most essential short term modification was the proper adjustment of the HP stop valve closure springs on all Steam Generator Feed Pump Turbines (SCFPTs) as recommended by Westinghouse Customer Advisory Letter 88-01. From an engineering evaluation of the opposing forces at work within the valve and site testing, Westinghouse established that sufficient closure force could tw obtained from the original springs if they are adjusted to a new prestretch value of 2 inches. Emphasis has now been placed on written procedures for the adjustment of the HP stop valve springs.
To address the valve linkage binding, the linkages were cleaned and new spacers (washers) ware added to reduce contact areas at the binding points.
S3/TGANA/a 71
8 4
j Improved deces'J1bility'of a local manual lever for the overspeed trip rechanism was achieved by lever relocation.
t The Valve Subcommittee evaluated tr.e HP valve balancing port l
design. Eight 1/4-inch additional holes to introduce HP steam above the plug and assist in valve closure were added to each'HP valve prior to operation.
7.1.2 'Minimizinz Challenges To The Mechanical Overspeed Trip Mechanism l
The System Subcommittee recommended that vital (uninterruptable) power be provided co the Electrohydraulic (EHC) Controller.. vibration monitor and thrust bearing went trips. Also recommended was a low suction pressure trip and j electronic speed trip with power from a vital (vainterruptable) power source. With vital power available, these trips would continue to function un' der Loss of Offsite Power (LOOP) conditions. These changes were made. This will-reduce the potential of future challenges to the mechanical overspeed trip mechanism, l
j 7.1.3 Recommendations To Assure That Items Uhich Cgtid_Cause SCFP Failures are Averted No short term modifications were recommended by the Turbine Subcommittee. However, recommendations to check the coupling alignment, vibration records, and coupling teeth for misalignment were made. The Subcommittee also recommended that the exhaust bearing bracket bolting be checked for tightness. The bearing bolt check was made since these bolts were found loose at other instati.?tions.
These items were performed to ensura no abnermal conditions existed.
S3/TCANA/a 72
t
,.7.2 Long Tera As part of the event evaluation, four general categories of potential long term modifications were developed. These improvements were grouped into the categories of Reliability of Turbine Trips, Avoidance of Mechanical Overspeed Trip Mechanism Challenges, Avoidance of Equipment Degradation and Transient Information Retrieval.
7.2.1 Reliability of Turbin' Trips Four major items associated with the Reliability of the SGTPT Trip System were identified for further evaluation. These items are as follows:
7.2.1.1 Method and Frequency of Testing HP Step Valves The first of these was the method and frequency of testing of the un stap valves. To enhance the HP stop /e test.ng an overall St #PT stop valve testin program e has been added to the vendor manual and on line HP stop valve testing will be performed.
This testing will be incorporated into the Operations Testing Procedures. This test procedure is in Reference 10.4.
t 7.2.1.2 Elimination of Stop Valve Linkage Corrosion The second topic associated with reliability of the trip system was the occurrence of HP stop valve linkage corrosion. As stated in Section 7.1.1, non ferrous washers have been added to reduce the contact areas where this corrosion was
- S3/TGANA/a 73 i
occurring. Action was also taken to reduce the likelihood of this corrosion. A definite relationship exists between the stem galling of the
- 11 HP stop valve and the fact that corrosion has occurred on the stem clevis and lever arm to link interfaces of this one valve, No corrosion occurred on the lever fulcrum point of this same valve, nor has any corrosion occurred on the other HP stop valves.
Evaluation of this situation has led to the conclusion that the valve stem galling had induced steam leakage at the stem. This steam impingement on the clevis and links of the all va'.ve has caused corrosion at these locations. A review of the valve stem and plug material revealed that the two mater'als are both 400 series stainless steel. As discussed in the Kalsi Engineering Report in Appendix C, this stainless steel to stainless steel interface is s2scepetble to galling. Nitride hardened stem and plug replacement parts for the HP stop valves are being obtained to eliminate stem galling and subsequent steam leakaga.
l Grease fittings are also being added to the HP stop valve linkage assemblies. A high temperature grease and periodic inspections for stem leakage vill be implemented. This will ensure that steam leakage around the valve stem does not bake the grease to a hard abrasive consistency should leakage occur.
S3/TCANA/a 7-4
- 1. _
N r<4i 7.2.1.3 Replacement HP Stop Valve Closure Springs The third item considered for turbine trip reliability was the replacement of the HP stop valve' closure springs with stronger springs.
Although the existing springs adjusted to a prestretch of 2 inches achieves the required closure force, other aspects of the valve's original design should also be considered. In order'to achieve the 2 inch prestretch, the original design amount of stem thread engagement into the clevis had to b'e reduced, with Westinghouse's approval, to the minimum required.
The second compromise made to get the 2 inch prestretch was to shorten the valve stem stroke. As a result, the plug is no longer retracted fully up into the plug guide. Again, this is acceptable as a short term fix. Long term operation in this condition could result in steam erosion of the plug seating area. New stronger springs designed to give the required spring force without these disadvantages will be installed.
7.2.1.4 Dedicated Path For EHC Oil Dump The EHC Oil Dump System, is described in Appendix A, Figure 13.
A question on the need to provide a direct path for oil release to assure a rapid release of hydraulic fluid from the HP stop valve actuator hydraulic cylinder was raised. The design has the dump valves emergency trip fluid drained from both HP & LP stop valves to a header common with other drains. There is an orifice provided between the HP supply and the S3/TCANA/a 75
emergency dump trip drain. This orifice leads to the af interface valve, solenoid valve 20/00T and finally to the EH reservoir.
During a SCFPT trip, the dump valve trip fluid will drain in the first millisecond after either the interface valve or 20/00T solenoid opens and ensure HP & LP stop valve closure. Based on this EH design, no EHC fluid piping changes are needed to protect the turbine ~.
7.2.1.5 Change' in HP Stop Valve Plum Seating' Angle Consideration was given to a change of the HP stop 4
valve plug angle to provide more positive shutoff of steam flow as seat leakage was occurring. A change in seating angle of the HP stop valve from 45 degrees to 37-1/2 degrees was considered to be a design improvement that enhances the seating action of the plug. With Westinghouse's concurrence, this modification was perforned on one stop valve plug installed to resolve a galling problem. The valve seated properly. Two other HP stop valves performed satisfactorily with the original 45 degree seating angle. Westinghouse has reviewed the valve seat an31 e issues and has noted that the codification is not necessary on the remaining HP stop valves. This modification will, however, be further evaluated for the replacement HP stop valve nitrided plugs when they are procured.
7.2.2 Avoidance of Mechanical Overspeed Trip Mechanism Challenges The addition of a trip to the SCFPT on a feedwater isolation signal, as recommended by the Pump Subcommittee, is being S3/TCANA/a 76
evaluated. Trippir.g the feed pumps is not a standard Westin5h ouse design feature and it has some potential unit availability influence, o
- The second item was the addition of a trip to the SGFPT when no booster pump is running. This change was evaluated as not necessary to protect the SGFP or the SGFPT. The low net -
suction pressure trip is more than adequate to protect the SGFPT. Addition of this trip, without a time delay, would degrade plant availability at loads of 50 percent or less.
Only one booster pump is required for 50 percent operation; thus, a single falso signal would trip the plant.
Implementation of this trip with a time delay would add an additional failure mode and would not necessarily be any faster than the low net Suction pressure trip. For these reasons, this trip will not be implemented.
A third item considered was the change of the manual steam isolation valve to a motor operated valve. If implemented, this motor operated valve would have the same trips as the HP stop valve. This change wt1 evaluated as being too slow to provide an appreciablo increase in protection.
In summary, only one of these items, a SGFPT trip on feedwater isolation, is receiving further review.
7.2.3 Avoidance of Equipment Degradation Four items which could help avoid equipment degradation were identified. These are discussed in the following paragraphs.
s S3/TGANA/a 77 1
k i'y <
F" 7.2.3.1 ' Adequate Technical Information Adequate technical information from Westinghouse _ro minimize the' possibility of the recurrence of the SGFPT failure or-similar failure is_being pursued. '
m HL&P has formally requested that Westinghouse
'*- provide a recommendation which defines what would be required to minimize the possibility of a recurrence of this type of event, y
- 7.2.3.2 Replacement of Existing SGFP Driver Shaft Couplings.
Another item considered was the replacement of the existing SGFPT shaft coupling with a flexible low mass, non lubricated coupling. This evaluation was undertaken based on recommendations from the Stone
.and Webster Review (See Reference 10.11) and from the Coupling Vendor's Review (See Reference 10.12).
The existing rigid coupling may have a short operating life in this application and a long lead time for replacement should a coupling fail. It was recommended that a low mass, non lubricated coupling b4 considered as a near-term modification to all the feed pump drives. The present couplings between the
, Steam Generator Feed Pump and Turbine are of the grease lubricated gear type.
The choice of grease becomes very important, since grease tends to separate and lose its effectiveness at high rotational speed such as in the application for the SCFPT. Also, these couplings require frequent maintenance and should be filled with grease on a regular basis. The major advantage of switching to a dry type flexible disc / diaphragm S3/TGANA/a 78 L
i .
coupling is the fact that they are virtually maintenance free and require no lubrication. Cear m couplings are either continuously lubricated, or are filled with'a specific quantity of grease and sealed. The South Texas Project (STP) couplings are
, grease filled and sealed. Another advantage of the flexible disc / diaphragm coupling is that it transmits less axial and bending loads on the bearings and shafts. This is due to its lower weight and reduced stiffness as compared to the gear coupling presently being used at STP. Finally, although the initial cost of a flexible disc / diaphragm coupling is slightly more than that of a gear coupling, the lower maintenance requirements and longer expected life make them economically justifiable.
The modification for installing flexible disc / diaphragm coupling to replace the present SGFPT gear type couplings is being evaluated. This
^
evaluation will include the cost, capability, and lead time required. Also being considered is the procurement of spares of the existing design to minimize plant downtime should the operating life of the existing couplings prove to be short.
7.2.3.3 Addition of coupling Alignment Monitoring System The third item being evaluated is a system to measure the misalignment of the couplings between the SCFP and SGFPT under actual operating conditions. This can be accomplished by installing a hot alignment system in the spacer element between the couplings. Such systems consist of inductive proximity probes rotating with the coupling. These S3/TCANA/a 79
7_
.c probes measure the amplitudes of the variations in probe gap due to misalignment at each end. The misalignment signals can be displayed or, an
... indicator. Continuous analog outputs can also be
. obtained. This system has the advantage that the shifts in misalignment can be measured under all load and speed conditions during operation. Horse power can also be measured for making heat balance i calculations. Consideration will be given during the evaluation,to potentially integrating this system with the plant wide monitoring system. This would provide on line readout and capability for alarms.
7.2.3.4 Time Delay for SCFP Minimum Flow Valve Closure The fourth item being evaluated is the addition of a time delay on the closure of the SGFP. minimum flow isolation valve. This valve allows pump recirculation back to the deserator. This valve presently closes when the SGFPT receives a trip signal. Additional pump wear may occur if hot water is trapped in the pump and flashes during pump and turbine coast down.
7.3 Transient Information Retrieval i
The last category of long term plant improvements deals with the I
storage and retrieval capability of information about plant transient conditions.
7.3.1 Improve the Ability to Store and Retain Data for Transient Analysis In order to improve the ability to retain and analyze analog S3/TGANA/a 7 10
a data in its scanned frequency, modification requests have been initiated to examine.the feasibility to add optical storage devices to the plant computer. In addition, the issue of alarm slippage is being reviewed to determine if any improvements can be made in the alarm handling function of the system. A further discussion of this evaluation is provided in Reference 10.6.
7.3.2 Traininz Required During the investigation of the SCFPT incident, it was noted that improva-t .1 in data collection and interpretation could be made.
If this failure had occurred in a non test situation, less data would have been available. Because the Nuclear -
Engineering Department was manually collecting thermal performance and power test data at the time of the LOOP test, nore data was available for event reconstruction. The information from these activities helped resolve some of the initial apparent inconsistencies in data about the system conditions during the failure.
Training aspects for personnel to properly interpret computer data and eliminate this problem are listed below:
, 1. Basis for time averaged data.
- 2. Basis for time slippage of computer data output.
- 3. Basis for ensuring that needed data is collected in the event of future transients.
4 Identification of key contacts in Westinghouse, Reactor Operations, PED, and Support Engineering who understand the above, such that data can be collected as required.
S3/TGANA/a 7 11 i
i
8.0 TURBINE AND PUMP RESTORATION The all Steam Generator Feed Pump Turbine (SGFPT) casing, rotor and shaft were destroyed. An installed SGFPT casing with rotating element was transferred from Unit 2 to Unit 1 as a replacement. Also retrieved from Unit 2 were all of the SCFPT stop valve internais to replace galled Unit i valve internals. The basic Unit 1 turbine skid, including such refurbishable major components as the lube oil pumps, control console, high pressure stop and governor valves and the low pressure stop valves were reused.
A major effort to clean metal debris and water from the SGFP'4 oil sump was required along with complete retesting of the replacement turbine.
Metal debris and oil removal from the condenser and condensate system was also required. Since the main feedwater valves had closed prior to the turbine distruction, no debris or oil entered the steam generator.
The pump casing was found to be reusable with the replacement of the impeller, shaft, wear rings and the inboard bearing housing. The overall time frame from the event to restoration of the use of the all feedwater pump was approximately six weeks. Within this time frame: modifications and resdjustments of the other turbines were also performed.
4 a
i i
S3/TGANA/a 81
9.0 LESSONS LEARNED 9.1 The spring pretension design for the stop valves should be based on maximum steam supply pressure (not turbine rating pressure) and lov [
(or no) steam flow through the valve, for valves where differential ;
pressure from flow may benefit valve closure.
9.2 A concern with respect to ma'intaining the design configuration of the SGFP Turbine HP stop valves, following maintenance, exists since no criteria were provided by Westinghouse indicating that spring pretensioning was required on the stop valves.
9.3 There should be diverse, reliable pvver sources for the solenoid trip valves in the Electrohydraulic System used in the protection logic
. for main feedwater pump turbines and the main turbine. Loss of AC power from any single vital bus should not prevent automatic trip of these turbines.
, 9.4 When turbine stop valves are tested both the trip actuation and valve positive closure should be verified with high steam header pressure and with the governor valve closed (simulating low flow conditions).
9.5 Stop valves should be tested shortly before any planned Loss of Offsite Power Test. Vendor technical manuals should include detailed instructions for setting the stop valve actuation preload spring tension.
9.6 Low suction pressure trips should be implemented for all main
- feedwater pumps.
9.7 Preventive maintenance programs should include lubrication and cleanliness checks of stop valve actuators to ensure smooth actuation.
S3/TGANA/a 91
?? : ,
f 9.8 When changes are made to normal plant lineups for test purposes, the impact of these changes -on automatic equipment protection should be specifically addressed in the technical' review.
mb J
.Y
'4 4
i S3/TGANA/a 92
10.0' kEFERENCES
'4'
'The following references which are, in general, raw input data, are on file at STP to assure that they remain available for future review, if required.
- 1. Compiled Eyewitness Reports m
- 2. Compiled Telecons
- 5. Detailed List of ECN's, DCN's and Modifications Incorporated During the Restoration Effort
- 6. White P9per on Improving Sequence of Events, Records Collection, and Archiving
- 7. Computer Data
- 8. Expert Subcommittee and Joint Company Report from Orlando Conference entitled, "South Texas Project 5/25/88 SCFAT ' Incident Investigative Team Report'"
- 9. Telephone Conference Notes with Other Utilities, etc.
10.'Onsite Investigation Team Reports
- 11. Evaluation Report by Stone & Webster Engineering Corp. (SWEC)
- 12. Coupling Vendor Report S3/TGANA/a 10 1
i APPENDIX A MAIN FEEDWATER AND TURBINE STEAM SOURCES, SYSTEMS AND COMPONENT DESCRIPTIONS l
l
- 1. OVERVIEW OF SECONDARY PLANT The prime purpose of the Secondary Plant is to take steam from the Steam Generators, deliver it to the Turbines, turn the turbine blades (which turn the Generator to produce electricity), condense the steam, then treat, deserate, pre heat and feed the condensed water back to the Steam Generators.
For each of the Units at STP, there are 4 Steam Generators which feed a Common Header (See Figure 1). From the Header, the major portions of the steam is directed to the High Pressure Turbine and the rest goes to the Moisture Separator Reheater and other miscellaneous users (one of whie.t is ,
the Steam Generator Feedwater Pump Turbine (SGFPT)). From the High Pressure Turbine, steam goes to the Moisture Separator Reheater to be heated by the High Pressure Turbine extraction steam. Once heated, the i steam goes to three Lov Pressure Turbines; also, part of it goes to the SGFFT. Af ter passing through the Low Pressure Turbines, the steam is condensed in the respective Condensers. The condensate,is pumped by the Condensate Pumps through Feedwater Heaters (which use Low Pressure Turbine extraction steam for heating) and up to the Demerator. From the Deaerator, the water is pumped first by the Booster Feed Pumps then by the Steam Generator Feedwater Pumps (SCFP) (See Figure 2). The Feedwater is further heated by the High Pressure Heaters (which use high temperature drainage from Moisture Separator Reheater and High Pressure Turbine extraction steam for heating) then sent back to the Steam Generators.
- 2. THE FEEDWATER SYSTEM The function of the Feedwater System is to supply heated, deaerated, chemically treated water from the Demerator to each Steam Generator during all loads from 0 to 100 percent. The Feedvater System is shown on Figure
- 2. Feedwater from the Demerator (which is composed of one deaerating feedwater heater with two horizontal interconnected storage tanks) is pumped by two of-three 50 percent capacity electric motor driven Feedwater Booster Pumps (FWBPs); Then, pumped by three 40 percent capacity Turbine S3/TCANA/a SD.1
i o
, Driven Steam Generator Feedwater Pumps (SGFPs) through two 50 percent-capacity High Pressure Feedvater Heaters to the Steam Generators (SGs). A 20 percent-capacity motor driven Startup Feedwater Pump (S/U SGFP) is available if ens Turbine driven SGFP is not available. '
e.
- 3. STEAM GENERATOR FEEDWATER PUMP AND FEEDWATER PUMP TURBINE DESIGN BASIS Pump:
The total pump systems operating either individually or in parallel have the characteristics of a continually rising head from approximately 125 percent of design flow to zero delivery.
The design allows for either individual or parallel operation of the pumps over their entire load range without creating undue vibration or other deleterious effects.
The design of the pumps is such that no detrimental pressure pulsations are generated during any operating mode of the pumps or piping systor.
See Table 1 for additional detail.
Turbine:
Each of the three SGFPTs drives a SGFP, supplying feedvater to Westinghouse steam generators.
Each of the three SGFPTs is installed outdoors and exhaust downward into the main condenser.
Each of the three SCFPTs operates at variable speed from no load to full-load condition.
The SGFPTs are of the condensing type, normally supplied with reheat steam. Each SGFPT is directly connected to its steam generator feed pump.
There are no shaft critical speeds within the normal operation speed range.
See Taule 2 for additional detail.
S3/TCANA/a SD.2
- 4. STEAM GENERATOR FEEDWATER PUMP SGFP OPERATION The SCFPs take suction from the Feedwater Booster Pump (FWBP) Header.
Normally, all three 40 percent capacity variable-speed pumps are running.
Each pump is equipped with a recirculation line. The pump required minimum flow is maintained through the SGFPs by controlling the air operated recirculation valve to prevent pump cavitation and heatup.
These valves can be manually actuated or automatically controlled by sensing pump suction flow. Each pump discharges into a common header via an air assisted non return check valve and a motor operated discharge valve. If the feed pumps trip, the non return check valve and motor operated discharge valves trip and cannot be reopened until the SCFPTs are reset. This ensures that the flow will not reverse in the feed pumps and cause dmaage or lots of fead flow to the SGs. Each non return valve is replaced in the swing free check valve mode when the feed pump turbine is reset.
Similarly, the S/U SGFP takes suction from the FWBP discharge header.
During normal power operation, the S/U SCFP is in a standby condition to serve as backup in the event of an operating SGFP trip. The S/U SCFP discharge includes a recirculation line, stop check valve and motor operated discharge valve in a similar design to the SGFPs. In the event of a SCFP trip or failure, the two remaining SGFPs will increese to 40 percent fu11 flow and the S/U SGFP will provide the additional 20 percent flow to provide 100 percent valve wide open (VVO) feedwater flow rate to allow plant full power operation.
The Turbine Driven SGFP speed is automatically controlled between 15 and 100 percent power to reduce the pressure drop across the feedwater control valves and pump power consumption at partial loads. Feedwater and steam header pressure and steam flow measurements are utilized to control pump speed such that the pressure drop between the steam and feedwater headers varies linearly with steam flow.
53/TCANA/a SD.3
- 5. MAIN STEAM / EXTRACTION STEAM INTERFACE WITH STEAM GENERATOR FEED PUMP TURBINE During low loads, up to approximately 40 percent, High Pressure Main Steam is supplied to the Steam Generator Feedwater Pump Turbine (SGFPT),
supplementing the extraction steam syst.em. At higher loads, low pressure steam is supplied from the extraction steam system only.
Steam extracted from the high pressure turbine is sent to the moisture l separator reheater (MSR); then, from the MSR, a portion of the resulting low pressuro steam is sent to the SCFPT. This path constitutes the extraction steam path to the SCFPTs.
! 6. METHOD OF OPERATION OF THE HP AND LP STOP AND COVERNOR VALVES (Reference 4)
Minh Pressure Stop and Governor Valves:
Figure 4 shows the high pressure stop anti governor valve whose function is j to regulate the flow of high pressure steam to the Turbine. The valve consists essentially of two diffuser type plug valves (3) and (38) assembled within a single housing which is rigidly mounad on a separate support. The steam outlet at the governor valve is connected to the high pressure nozzle chamber in the Turbine inlet cylinder base. The piping connecting the governor valve and nozzle chamber is fabricated with ample flexibility to compensate for differential expansion between the turbine and valve assembly.
The stop and governor valves are sissilar in configuration. The stop valve is equipped with a needle valve (25) to admit steam to the chamber in back of the valve and into the valve itself. This relieves the pressure unbalance between the chamber in back of the valve and surrounding chamber. Balancing these pressures prevents sluggish valve operation and helps the valve close under steam flow conditions. The larger end of the stop valve stem acts as a pilot valvi to equalize pressures in front and in back of the valve prior to actual opening. First movement of the valve stem releases the pressure in back through the valve to the chamber in front. The pressure in front acts against the front valve surface to effectively balance the valve, greatly reducing the force required to move S3/TCANA/a SD 4
it. Similarly, the governor valve (3) has longitudinally drilled holes for chamber pressure balancing.
Each valve is independently operated by its own servo actuator through operating levers (12), (34) and linkage. The Stop Valve linkage is hinged
! such that outward servo-actuator piston movement opens the valve and inward movement closes the valve. The Covernor Valve linkage arrangement ,
provides the opposite valve action for the same servo-actuator piston movement. The Stop Valve servo-actuator hydraulically powers its valve only in the open direction. Dual springs provide the stop valve closing force. The Governor Valve servo actuator hydraulically powers its valve in both the open and close directions.
In operation, the Stop Valve (38) is opened fully and remains so until the unit is shut down. The Governor Valve (3) is positioned by the electronic controller to vary HP steam flow in response to changing system requirements. The normal mode of operat*on is for the HP Governor Valve to open following the opening of the last LP steam chest valve if steam "demand" exceeds the supply capability of the extraction steam systems.
Low Pressure Stop and Governor Valves:
- Figure 5 shows the manner in which the integral linkage of the LP Stop Valve is mated to the servo actuator to form the Stop Valve operating mechanism. Extending from the bottom of the stop valve is a stanchion to I which the servo actuator and servo actuator guide are assembled and secured by bolting. Threaded on the end of the servo actuator cylinder rod is a spacer which makes contact with the end of the stop valve pilot [
valve stem whenever the servo actuator functions to open the Stop Valve.
Upward cylinder rod movement unseats first the pilot valve then the utop valve, permitting steam flow. l S3/TCANA/a SD.5 l
Figuros 6 cnd 7 shtv o typical crrcng:Ecnt of occam chsst govornor valvas and linkage. The steam chest is cast with an integral nozzle chamber and is independently bolted and pinned to the inlet cylinder cover. The nozzle chamber is cast in a manner to admit steam in a circular pattern with each valve controlling a certain arc of admission. The total are of admission and the number of valves are dependent upon design steam conditions and rating of the unit. For each particular turbine there is a "Coverning Valve Setting Diagram" showing the number of valves, correct velve opening sequence and the setting dimensions for each valvs.
- 7. STEAM CENERATOR FEED PUMP TURBINE AND PUMP DESCRIPTION (Reference 3) !
Steam Generator Feed Pumps (See Figure 8 and Table 1):
The Steam Generator Feed Pumps (SGFPs), located on TCB Operating Deck elevation 83' are horizontal single stage centrifugal pumps coupled to their respective turbines. The pumps are variable capacity from 13,894 gpm and a differential head of 2,300 ft. at 5,075 rpm to 15,570 gpm and a differential head of 2,420 ft. at 5,420 rpm.
The SGFPs are provided with a seal injection system supplied from the condensate header to control leakage across the throttle bushing. The system employs the use of pneumatic pressure controllers ard valves to accomplish its purpose.
Steam Generator Feed Pump Turbine (See Figures 9, 10 & 11 and Table 2): ;
The Steam Generator Feed Pump Turbine (SGFPT) is single axial flow with six Rateau stages.* Steam from the main (high pressure) steam header (low power operation) or outlet of moisture separator (low pressure) reheaters (normal operations) enters the Turbine through a throttle valve steam chest assembly. The steam flows axially through the Rateau stages and exhausts to the condenser. '
- Rateau Stages describes an Impulse Turbine Design which has pressure compounding stages.
l S3/TCANA/a SD.6
Each feed pump turbine has an independent speed control unit operated from CP006, located in the control room. The turbine speed is varied to maintain a programmed differential prsssure across the steam generators (45 to 175 psig).
Steam Generator Feed Pump Turbine Cland Se711ng Steam and Leakoff (See Figure 12);
The Feed Pump Cland Sealing Steam System provides a neant to prevent air in leakage at points where the rotor penetrates the outer cylinders.
Each feed pump turbine is supplied from the Cland Sealing Steam System via a 4 psig pressure regulator (Pv7130). The pressure regulator is pneumatically controlled, fails open on loss of air, and can be bypassed if a failure of the controller occurs. The downstream side of the pressure regulator contains a 25 psig atmospheric relief valve and l strainer with blowdown connection.
Leakoff from the feed pump turbine glands (1ebyrinth type), high pressure l and low pressure stop valves are routed to the unit gland condenser.
l l
Steam Generator Feed Pump Turbine Low Pressure Oil System (See Figure 13);
The Feed Pump Turbine Low Pressure Oil System supplies oil for lubrication, the low pressure trip header, ano to hydraulically test the overspeed trip and the oil reservoir level switch.
Oil is supplied from the integral reservoir built into the turbine bedplate. A separate compartment beneath tha inlet end bearing bracket serves as a drain collector tank. The drained oil flows through a strainer in the collector tank, then through a baffled port into the reservoir to be recirculated.
S3/TCANA/a SD.7
4 The oil r:strvsir h:s two main AC motor driv:n cil pumps, cn eaerg:ncy DC motor driven oil pump, twin oil coolers, a duplex oil filcer, a high and low level alarm switch mechanism together with test solenoid valve, an electric oil immersion heater, and an oil vapor extractor mounted on the oil reservoir.
The collector tank section houses another electric oil immersion heater, the turbine trip solenoid valve, and the overspeed reset solenoid valve.
Under normal operating conditions, oil to the lubrication system will be supplied from one of the main oil pumps. The emergency DC pump will
- supply oil to the lubrication system during an emergency or when the 1
turbine is rolling on the turning gear.
The Low Pressure Trip 011 System serves as an~ interlocking arrangement between the turbine prote:tive devices and control devices that regulate the steam admission valves. The actuation of any one of the protective devices will dump trip oil pressure, which trips all turbine steam valves by virtue of the interface trip valve opening and allows the high pressure fluid system emergency trip header to drain.
SCFPT Electrc Hydraulic (CH) High Pressure Fluid System (See Figure 13):
The Feed Pump Turbins EH High Pressure Fluid System supplies synthetic oil from the Main Turbine EH System to provide the motive force for positioning of feed pump turbine steam valves in response to electrical commands from the electronic controller, acting through hydraulic servo actuators.
The major elements of the EH Fluid System are the HP accumulator, the drain accumulator, the steam valve servo actuators, interface and trip solenoid valves. The major components, except for the servo actuators, are located on the bedplate. The servo actuators are located near their respective valves.
S3/TCANA/a SD.8
SGFPT EH Gov:,rning Sy, ten:
The SGFPT EH Governing System provides precise control throughout the turbine npeed range plus immediate response to changing conditions. An operator's panel mounted in the control room provides all necessary controls, lights, and indicators for starting, operating, and stopping the Turbine. The SGrPT can also te operated locally on the Turbine deck.
The SGF1f EH Governing System controller assembly consists of DC power supplies, speed and servo amplifier cage, and a Relay and Potentiometer Panel.
- 8. OVERSPEED AND MANUAL TRIP MECHANISM (See Figure 14)
The trip mechanism automatically shuts down the Turbine when it overspeeds. Shutdown is accomplished when trip header oil pressure is allowed to drain through the overspeed trip chamber. This allows the spring loaded interface valve to open to dump EH trip header pressure and also actuates oil pressure switch 63/TT. Dropping the EH trip header pressures allows the dump valves to reposition the dump EH fluid from the HP and LP stop valve actuators through the normal drain line. Fressure switch 63/TT also actuates the electrical trip circuit. The mechanist. is composed of an 6ccentric trip weight, an oil release valve, plus manual trip and reset plungers in a common housing mounted on the inlet bearing bracket, k'henever the mechanism has been actuated it must be manually or remotely reset but only after turbine speed has decreased to allow the trip weight to retract.
Hanual reset is accomplished by pushing the reset plunger handle which allows the pilot valve and the oil release valve to reset. This permits l trip oil pressure to build up and hold the piS?t nalve and the oil l release valve on their seats.
S3/TCANA/a SD.9
l
^! i l
_ When the remote system is used, high pressure oil it admitted by an external three way valve to the back of the reset plunger, causing the mechanism to reset the same way as pushing the manual reset.
- 9. OTHER SGFFT TRIPS i t
o Law Basring 011 Temperature 63/LBOT o Low vacuum Trip 63/LV2 o Locally mounted (at instrument console) Turbine Trip Push Button o Remote mounted (at CP 006) Turbine Trip Push Button o Vibration Trip (ZLP014)
E.*W.-Tr ip s :
o Pump Suction Pressure Switch 63/NPSH (Per Mod Package 88098) o Electrical Overspeed Trip 83/SP (Per CCP 1 J+FST508 and ECN 88 E0088) r
- 10. REFERENCES
- 1. "Feedwater System Description", 5S130MD0120, Rev. 2
- 2. "Main Steam System Descriptica", SS100MD1026 Rev. 2
- 3. HL&P Training System Description, SD FV02 "Steam Generator Turbine Driven Feedwater Pumps", Rev. 0
- 4. Westinghouse Instruction Book "Steam Generator Feed Pump Drive Turbines" Vendor Draving Log No. 4010 01046 CWP
- 5. Specification for Steam Generator Feedwater Pumps, 8S139M50017, Rev. 0
- 6. Specification for Steam Generator Feed Pump Turbines, 85139MS0033, Rev. 0
- 7. Bingham Williasette "Installation, Operation, Maintenance Manual" Vendor Drawing Log No. 4006 1008 BBW S3/TCANA/a SD.10
- 8. P& ids
- 1. 65139F00061 #1, Rev. 11 - Feedwater
- 2. 55139F00062 #1, Rev. 9 - Feedwater
- 3. SS139F00063 #1, Rev. 11 - Feedwater
- 4. 65139F20009 *1, Rev. 12 - Feedwater
- 5. SS109F00016 *1, Rev. 10 - Nain Steam
- 6. 65109F00017 #1, Rev. 11 - Nain Steam
- 7. 65109F00018 si, Rev.13 - Main Steam L
S3/TCANA/a SD.11
1 1
" l STEAM CENERATOR FEED FUNF DATA A. Quantity Three per Unit i
B. Type- Horizontal, Centrifugal p
C. Manufacturer Bingiiam Willamette Co.
C D. Rating
- 1. Capacity (each), gpm 13,400
- 2. Total head, ft. 2,500
- 3. NPSH required, ft. 220
- 4. bhp @ rating (15,750 gpm) 9,965 E. Design Pressure, psig 1,750 F. Design Temperature, 'F 375 03/TGANA/a Table 1
N a ir. ,
STtAM GENERATOR FERDWATER PUMP TURRINE DATA g $
- A ^. Qaantity Three per Unit
,S. Type Condensing steam C. Manufmeturer Westinghouse
.D. Rating
- 1. hp 10,000
- 2. rpa (@ rated condition) 5 00 J
- 3. Rated steam flow @ 4 in. Hg abs .
exhaust pressure, Ib/hr 100,600
- 4. Turbine trip speed, rpa 5,940
- 5. Turbine critical speed, rpa 2,500 i
a 53/TGANA/a Table 2
a en n u.=
______ /
out r _r_
I i
i p_- __F g . -
. /
n m Po u, I :
"oB 8 Er [ u6 o I I g I
I
<___A
" C : m U I i I l I
E }s '
~
f o'@6 lll ll / \i ,, l p l I
I l
I
[mA I I
[ '
I I g
/
i I M I l
= T 168'F I >
y I 4 -_ -
p b---_ 11 Ig C9g U T 218 'F l
I /
[m IN IU 3
W 548,000 ___f 0 8
N
$ [ f t T-266'F l l
O Z
_ / W 563,000 __y I O 317'F
~~~~ '
Q 11,600,000 5 I ch @@
C $A I hSg w
I F U
g a
- - 9m -___-_
l W - - - - , p- >!C--- "
Js, -
3 (j9t_w.gy_lf= i n Sw o
mo x p--- __- -
l gg
$ xm gh ' 1 '
n m _
1 i
8
< ,-y2 Fj im l
u___.):::7 \j $2 I
[s I
) W 1.204,000 P 436 f H 8
b~#
\( \ Em i I
???
204 u,gg
/ 3 W 17,432,000 "E __g i 9
g' g P-1,071
.m 8 *
, msm$
D
'55
$4RO M #
.s
N I
e x
,ny w,a
, m, a ,
n a ,
' m a ,
f,---- ---- ---- ---- --- y --- ---- -.
-l 1 3' '
b& B-' W '
$ 3--k ( G-$ (de-$ Gyd
!.~',"] l n 61 prEl Im nilpII) si ni GEE > $
9 igtg!i3]
,r-- {4
, r--- N, r - -- N9 , r--- h0- '
8 m,
[
,gb [
, , ,@,k[ , ,@,h[
2 3 1- t 3 1- r 3 q- r 3 1- r L----- . L------- L-- ---- L.------
y, 2..,
y 2._,
V) 5 0 -
g 4" '
nW
~ ~ ~
h 5 Ego lo
=
o N "
1.I 'lj rt,gg A, eo SJ l-
=
I B
s i
V
e 1
C g { u og v
' I t s (N t b ti 'g
~
m!\"'
8 E
- a i 8
- e4- {$ ,
si t
, a ii mg W .
s,i W
i $ l*E\
t e
4 II 5
- - P{
a sm
/ \e e X l-4 e*
z an
$ C D a
W E
E t
a X
$ li 4-1I 3 J I
o O
\
( /
. j ll O
- f E
R
(
4 E
R U
G I
F
=o
=
f w?P E V
- L ggnt_-
x A_
V _. _
mQyx R
as1 O
_g !
1 m N I
2 n m g R 3
f ,
n,h E V
n-U 5~ 4 j __
O i1 _ E G j,
n,n 8
,?i:
- P
- s O i ['
' l
., j T
g
)y ,( .
.a S yi l -
f i
E q(,)Q*
L ; gt a' y. 5_ 1
')
\ i-R U
~
- f$ S S
4 3/ k_
~
- I ff[i/)#R c
, E
, R P
H G
_ t I H
-N,.
i .L i
. $) b j j q
)
E -
= :
E DC5cmfrt10N
, um O'l 1
i 11
(
? e 8 e
{ Ol ,
5 ptAtt.e gag De povtsterv 4 Tweef'?t8 M88W
. llAVll e owe
-/g, [
e emeone 3g g ? ttwo8W" N g . e svuman
/ / e m.us. m as
- / / -- ,e emsn g h /J
/
y4741._, #///
- y 4
. W(
/
/
\\ '
ie en.v is nasv
.,+.=
rapt (*(w se ee rL5+Er.S C
@ - ,e .
- A w
e 7 s& DIS C y / y; y yj- ,e e- e-
-.n gl y' en e*es j
. no It FLamW j
gy y t
,, . ~ , ,
w m,e- ; .. e_ ;
v hs .N w
l i'NNNNXv NN, j.y.
j a ,
N A
, N, g -
- e. m e-to c28 Scar
- l l
^
k' : ;
} '
9\ x\n ,,
3, pg,3rg pgi, hD *~' ; I l es se couw.
- 3. u.
l H- Si matca
, su pues wo.
8 83 C &# iC*tw e ,;
s se u>:a mapea 31 M, ,
j ,, ad u cs anew
= m ,. - e
- ) '
l y [
s .
] s se se covav=e art warv l l ** 'caz
.. m.
' " .: .; r. .o 4 , ,
- mW
-a ; I ll Th. "'
- . l
@-- / $1RN NC r "TTY i GTT- -<h4 I I
? l h '
l ///
N'///d .
1 LOW PRESSURE STOP VALVE FIGURE 5 (REF. 41
.- .. _ l Q [-. \<
/
1
-L. , ! 4
,% I*,.* + + +
- f * * * * +!* *[
. .L {p A- - --
4+- . . . . - . _ . - . -Q jt (*9 . - - +- - + '. . -* b
.,. u , ,
. , l
] q,...................
L_ 7
,. ; s== 0 *-
g .,_.,.
_..e._._._<._,..-._.-.- r .-+- li _
y
- - -- - - - : . . . 3 g~I 'l N'T h b
/
ah j
- eu..e tau.a
\-
4p.
si, .at -
.' 1 -m
~ ~ ~ _- -
_-a%
I-I.. l s y
/ =-2r. -
L _0 sw\,,
/ ,
aw/wswanannys v/w; - ,
'y v g ., -
$ l Mbkhkl[j-l h.. t .
.i /// , ed l y } )
~Y b'-
% \,. ,, .. 2 '- - -
n.:"., - ~. = ,.---.-, .J LOW PRESSURE GOVERNOR VALVE FIGURE 6 (per 41 v
l I
M r.
\ u i
i
,, ,N r -
n
'x N <
k
>.,1 i
--h: -
..,m..... .~ ,
$/ 1/ - pt,
~s %/'N'xg ,
(5 g\
v mi,- '
-I
, , ~ -
/ y..
. s). L_\. 7
~
. g,_ 3 -
\ - ,/ r i
's t- 9'**
W/ .
~
\k "'
j / _
-- M s - -
, ri .
,, - ._ ) :
m .. ,
p #
"' .., , c m.
(t%aaGt06 r
I LOW PRESSURE GOVERNOR VALVE FIGURE 7 (Rtr 4)
rri m
\ ILLA i ij .
,1- -r -
n s..
l . -
,C --- --- %,
. mrme ;- ----- - --g -
,_:I!Y*h M f,
-7 7 f c * -
v.h : 4 a
y~-1-7 i
g K'Q4&,6 ly 1 1 d.
k
( 9@::y%U$ 1. - Fr w i T k.-M..7 ,
T@
l e- EAST t"""*"
m 3 /
% . ,. -.h' y+A
,W '3-O t ""- t ,,
f-t _11 Lap -,
r, L
w!A m 4, d O;.1j r8 ....
4 7-yn }g .g;I: t ~, s
" 3,$ Uyj&g;4~ y:i }- ,
i , ;
+ ,p ,
1* I /
! + ' g. .j'
' )
-b> trisrf gu.
._ _ _, s _
r -
a :a. ... .u: ii.
u M;;;. 2., .:.m....ri. l g.. in e-er-: :22 m
_g y. -- .
l . - - s-1- ,
-.m--
i--J ,
.t.
MAIN FEEDWATEP PUMP PLAN 8 ELEVATIONS FIGURE 8uur.73
~ ~ ~ ~ ~"
- ~] . . - . -
- = . r"*I /1{7 1:i- - -
i l Fil/ ,
,, l_ -- -
I m l]~ } , _1_ ,
Ese
._i& , ,*[K g
'-.L -
, q -
n.c.str a sutti f i-g_ s { [--
g 9
,1 s
FN \l '- N-- . i bJ
'. [ , kg.IT'
--"-')j ;<ill. v_ L.
- t. ,
_ . , ['
._. E ~
ria- -
! 'o '
a
, -f.u 4i
- lgy
' f 34 ye j j q
_;_ y't=u=s^jL)aM 1;3 l,f 02 Q% _, # M L. b;
( a%.3..
i t....
,[, . L 4
__ , Ii g' sd w fLFDM*g J
~
I
=
M --
w_ - - -.i , - - -
=71__ _. z .
I STEAM GENERATOR FEED PUMP TURBINE FIGURE 9 mu 4:
EAST A u.e srtmas emotT A
' h
+ NORTH s --g 3 ,
~ _ .
.m' 1
'*=
L-
)
ser resv c is s ,
-;R f L
=
r,p i 8 _ _u .
e ;
p l r lD th- i u p'k. . r/ .
gm L--lJ .
~ $$'r$w'[nY'cSC". t) = r. srca,, L 2
=tst g g, Ys U .]-
's."E.'r.m"<'w"E EEEEo* *$Is**** J ,_
"f d Tf rue in.
^"
-h.-- ,
h-1_ l
[- N Rm e
/
M u .. p ,_ _l :_ - . s - - - - -* * --- rl 4
- g- ___
i,L I' -
-I- kJ
-j 1
,! '\ .
r r
, I -. i i.J x g ,, % TN -p mq%i Q
m fi v. u -* ~ ",,L.l
- { !}"t ~
h
~
. El l r, j
@1 _,_ _-_ ._,m _ _ _ - -
.g__ _ - _ _ _ _ _ _ _ _ _ _ _ _
==- _ , _ _ _ _ _ _ _ _ , . . _
STEAM GENERATOR FEED PUMP TURBINE FIGURE 10 (no 41
- nww :
55 I ;
l .__
,["
Q;i t4558L . _ .
U,rk Q.. . ^
%s'f. ; -
., ti
=)
I ;:
fv apu l 3
J -
J-[ - m l,, i.n.us.t.a
< - . Jou..at
..s,e ,ou...
';o N n : I>' , n p'~'.,,, **' _ , _
_ g i _: 'L. ,
L/
I't I
~
b .
gy
_ f& *' }\ . /
?o gk l i
_4 i r i
. . o..
n a ,i -
4 e
[
_lij
~
j__lj f 1; 11 ' ',
i
- wr 4f
L..................._..;
otL sum 9 p : ,
SECTION A-A STEAM GENERATOR FEED PUMP TURBINE FIGURE 11 (REE 4}
l _
l l
t j '
L P. STOP VALVE STEM LEAKOFF l
j -tow PRES 5URE STE AM CHEST L P STOP VALVE ASOVE SE AT OR AIN
' g p. coy. < > < >
TO CONDE N5ER i FROM UNITS 2&3 '* '
\
- f^E AKO 5 -
OL AND L E AK OFF Y O M AIN L P. S TE AM SUPPL V .
WT GL AND CONDENSER I-l - -
l l St '
N II "" -
' )
TO GLAsep CONDE909 Eft L P. STOP VALVE BELOW SE AT OR AIN TO CONOENSER ;
i
. 6 .
f ~
H P. STOP VatVE
- i LP STEM L E AMOFF ;
l , , -
II E l d ' ' H P GOV VA1VE 8 HP STEM LE AMOf f j
SST STAGE H P GOV' 1 CVL. DR AIN yg g.
p i
TO UNsT5 2&3 L P STEM
< > VALVE H P L E AMOF F STEM
- m .E Att.O R, .s
" ^ " " !
- TURSONE \ s g j W e_ _
i ENHAUST LEAMOFF m' -
! a 3 3 3 E SUPPLY j 25 PssG " II C ' " 8 U $2 g, g j p CONOE NS ATE g, , , , 5 . POGN T l DR AIM s a e e
f % H P L OOP I l s P. GOV. VALVE BELOW 3 l I CONTROL VALVE _SE AT DR AIN
" ,,, s,,t ,,c SEATDRAm O CONoENSER DC D" ~ ,STE AM FROM l y
i PRE S SURE MAM UNsT i GAUGE ,
~
}
FEED PUMP. TURBINE GLAND SEALING STEAM a LEAKOFF FIGURE 12 (REE s) i i
4
- - , , - ,,-.,.n .n- , - - - , ---- - , , - - - - - - ~ , ---- - - - , - - - - - - - - - - , - - - , - - -
!
- ACTu ATORs i JL JL
$vCL Sv0P 1 kIdh TURumG NO 2 GtAR eE ARius ?.',WC TC pij PE NO.1 ARING THRUST OVERSPEED OVERSPEED BE aRIN T RIP TRIP RESE 1
- % P,} tu vs _sc_ 'c T G. .p/
, ,, SP fU
- e .IEt --
,yj W E ' ' :k g Pt ] ,
E 3)
I g ,
', , RIP W IGHT TC g g, ,*t 63 *t U Obs ,g 3 J3_ DD 3 SVCL -
I% PUMP - O tv-1 33 C Sv0p M IN N
- 5 OD 23 @D j 0 -o 'd L 2
-63 Sv0P E.H. RE' SERV 0eR O gBO O #DRGO \O EOp 63
' ' E OvERSPEED SEARmG ; LW3 l
-_: 1 P' ' f uT. O P h 2 3_ 4 8 SOL AT ON ,, ,,
1 ACCUM. ' '
TO TUR84NE EhHAUSI ogggp HP Ost -* Oct L ystitR COO 4 E HS H P. 5iOP vntvE{watyg, ,
63 ,4
{ Qq C u < . _
33 33
, , g G N OO OS
. t H. RE S. 63 63 gg :
'M P. OPER ATmG $ t OT2 s3 S$ ,,C =:
43 83 EOPT2 # #"
I SE Rv0. 13 lev.AtM MOP 2- 1 g[,
I u - IT D MOP S-2 MOP 2 -2 - I
\ PUMP EOP
, s uma ,
FI ' '
O 4 SI
} A88 l
Ij
=
$ FEED Otu E3 ,
p vtNT ' '
L P. Gov.vAtviE WT.F. PUMP 1 1 T hec d'* b vtv . ,&
O ' v. x1R Ft 20
- OLL O' '
20 OL "
Q- - -
El
[l '*"' '
yg eggp iT - 20 20 ((
GvC Gv0 OOW PO4 l OR m R< *
@ j MOPJ , MOP;1 j. /
4 == -
l E'
, ROM -a_, e i I ' '
__j "'
1 4$ , 3 C
00T; I STR AINER-J VENT Illl HOtf '
]M M{ 4
+
JOM =]/g MOP-2,
- - d
', g y , MOP- 1 ,
) "
u g
'1" servo. ORAM Log !B II !I '
I ,t , Ms. " ll c EO' cs G- ODC 50 l u P. Gov. watve mRo. L - tat cv0R s y f ton,,scE Ost RE SERvOr1# .
i PUMP of ARmO Oil. H 't tEvet At ARM
)
! ELECT RO HYDRAULIC HIGH PRESSURE 8 LUBRIC ATING LOW PRESSURE TURBINE OIL SYSTEMS n o m m i,,cP 3, 1 r
PILOT COMPRESSION VALVE SPRING OVERSPEED SEAT TRIP WElGHT OIL RELEASE /
OIL VALVE SEAT / OVERSPEED RELEASE /
1 ; "
RESET -
TRIP LEVER VALVE HANDLE g . ,
\
b L f !
xgkg f ,
I*sbs s
A "'
"h < 1 -
TURBlNE (jt hg I J
~
'5;$ ROTOR f VMzzzgoz k 3
._ j, w$]-mr 3 .
E -
s Jg!jy' h.
A
- DRA!N ll R
@^ 1
? -
a TRIPff RESET j\=- = PILOT VALVE MANUAL HANDLE .
- PLUNGER TRIP
~
PLUNGER H.P. RESET OIL CONNECTION :
- 1 v
TRIP OIL N I
CONNECTION BEARING BRACKET BASE l
OVERSPEED TRIP MECHANISM FIGURE I4 inte a p 5 _ _ ____.__________ _ _
APPENDIX B TURBINE AND PUMP AS FOUND CONDITION WITH PHOTOCRAPHIC RECORD
This section summarizes the as-found position of the turbine components. The decvings reveal the general arrangement of the component parts in both olcyation and plan views. Alpha characters on the photos and drawings cpproximate the as found positions of the component parts.
A legend is provided, on each of the drawings, to identify the parts. A ph2to account is also included to further illustrate the as found positions, j
l l
S3/TCANA/a B.1
~
u l l !
l Efg I = .
-[p- v3 !! *! l_ !i e
a ,i,IIl d
A "
a I.
g r
f@
ggs 5 e
- ]
2 kI8 8 ;
- I-e<s-s e -q1 b . 8 l l < g s 5ua91
, t s;eh=
E t=N s8 m,2 l[!!,gi*ll!!
r sin g g 1t - e
!1 -!a l . e!! - v!!g - -
i riliiils e v sov e !!
[
.tilIli!!illE!!lU!!!!! ,3 8 a n<
o@eee9@@@eeeG@@@eseeeee@e! l -
ee@@ @@@ @w 2
nrd, -
-_ ,' a.-.~n w
_ a F
a 2
i
,_ $jE (3 Od E8!
g
@ 6gf3
~
} $$}
Hx
~I Y !
l @ /$~[h EN ,a, $$lz
+7 . .- , o L g
@ !/G i O '
W u g
1T@ -
- u. y e([-(f[
39 E
- o. .
X.g -
&, @2 9b g <
e I v j s K .-_\W
~
s--pb
\ j..y c..'
.M ,
il.
i,3
_ k I lhg m.,. m e .. ..s u eo ~ ;ji a .
c # f W % ib e
-9 ,j f@
@ @r w- n._ '
a
LEGEND j "
h THRUST DEAmMG HOUSING-UPPER HALF (SOUDOC)
@ TuneesE SHAFT-SOUTH DO i @ ruur oc courtes Hus l
i ~~
I~'"" P#'
r-'"'
i
@ Tuais e cE*a =E'ax
@ uANUAL ovERSPEED TMP DEVICE 8
/ @ mAnuat oven .eEED Tne vAtvE
- u. ,.. .
n =' i w ;
__ C_j __ T ,: vi --- -"
i f ic.,
\ ; .
- . - - .+
l l
A' ;I-
)
1 <
l
(
- +
r~~~ } .
~
't r
g 4 y- c '?Ti f_
D ,I j i ,. h, r @ j Lj E f-pr p gly
-l 4 .
ibi / O6 1 i 11 -Nfw!
y i ! N. -
4}^), --.. ,
I
[ ]
' ;' GL_
Wr
@ s, @_
_, . . 3.- ..= . _ _ z __ ~:
up - _, . . . .
1 MAIN FEED 1NATER PUMP TURBINE LOOKING EAST TURBINE COMPONENTS AS FOUND POSITION FIGURE 2
y - _ _ _ _ _ _ _ _ - _ _ . . .
one M
W b
b '
g g!i.s ea s.* '
lhg*il[!
w Z
l I E !' I I yl l
_7 F- 2 l N ~
l s
- a. x ei l
- a. o .
e
- s= --. ,
,tY
.}h . . _
- U,!!
(fe ? , !
, ei P : _
E E
~ q
=
LEGEND ,
h PWOTED SHOE Joup.6AL BRACKET
@ mOTEo oE -- .R.CxET h SHAFT FRAGedENT FOUPO NEAR NORTH SCE OF mesR N h MANUAL OVERSPEED TRP DEVICE
- 0 <<Oe - r e HORZ. EXHAUST LfC TO CONDENSOR) r5/ - -
h TURBINE DIAPHRAGad VERT EXHAUST (FOUNO LINE, APPROX. IN -
=
-M 1
g
( .,!
%w-E 2 FT. SELOW TURBME) i H
h;4
- g :
l
+ 1 V A g s l-n:u-l-i 4 _
-j'
-!-[-- QQQQC+- =. r lL ,
=
0 'J I J I'
1
-Ii I sT X ! ==c+- - -
l a: . T ' r 4, _l E_r Ta
_ gg_ - 7,i-j'
(.1 >_
11 N \ [ ]
T@
,c @@
O I
MAIN FEEDWATER PUMP TURBINE LOOKING WE3T TURBINE COM*0NENTS AS FOUND POSITION FIGURE 4
l Ya ^
^
p -
I-a * .
_ r o '
6 i a, .
. 49 ll 4 h i
- tunne- ,
\ e = z-
.x =
' I ~
I l , , , ,j 9 s ' ' ' *
. e ., .
^
\ . . . - . - ,
l g < '% l
'4 ..
l photo 1.
The as-found position of the pump hollow coupling tube (part A) as seen i looking down into the north cnd of the turbine enclosure. l l
'7" i s j 4 .
l A {\
- l
. l I t I >
l I
i
)
vf 4, l-M.
. ,,3 ' f* ,
l '
- l _. .
g.
photo 3.
l
e l -l e,
1.
$ .- ~ ,. _- N l .g / % - ,
j; ..
photo 3.
The north turbine internal-bearing oil seal and parts of the bearing temperature and vibration monitors are visible on the north end floor of the turbine enclosure.
e
, o o.
,. g
, .- "L- ,
)3 \, - s.
~~ -
,/ .,,
/
, / .~ di A 4 j
} ,- f x .- ,
(o c rs [ *
'.). A photo 4.
The hollow coupling tube (part A), upper bearing housing (part C), and the turning gear assembly (part E) are shown as found in the north end of the turbine enclosure looking west and down.
J g
gg , :.
/
y *
) .
photo S.
n th north end of
%I[
e turbine e e osure lo k ng down and est.
- p' .; < , ;
7' '
y .
$4 s ,
kA
~-
1 ,
- l I
4 photo 6.
Utbin DC OSure IOokinb w e$t.
l i
y ..
1 y .
l '
2
-yk i,s .. '[
yx7 .
1 j j < . .
+ 4 A
I gyF;- V ,
j ;f; =, .~,
photo 7.
I dhs.k 1 The thrust bearing upper housing (part F) as found in the south end of the j turbine looking down and to the north. ,
l i I (t 9,'
~~'
~
Y a i C. # i !
(/e' 4
I[l .i ,,[*,,
- l l,1 * %
3,'[ .N 3 h
. .- .i, !
t..'.' m, ? >.- 4 6,...
.yA. :-
.i
[, ~ ,. :gy*b 5
, i k ,
,; ~
. L. r photo 8 The turbine shaft piece (part G) is a product of the shaft failure at the (
mechanical oserspeed coutitcr w eight assembly and the turbine main j bearing.
h
o ,.
.. . , . .m l
\
( ll , ' . ~Wr e
jf k s [
hsr 4
- I' x N','%
( 4
. ,s (j e i :
1
. t
- e. , .
' *- k %
- F I
se I n .
g ]
a H l
-, .I !
g l ,
v "f' ? --
I. ,_ l ' .# 1 &
- i
. ( >
?- - .p 4
_s -
j photo 9.
The turbine speed sensor vibration rnonitoring tree as found on top of an l clectrical panel (view is from the south enclosure door looking west. l l
I photo 10 Close-up of the abose photograph.
j photo 11.
South thrust bearing pads (part it') numbers 2, 4, 8, and 12, looking down.
and to the north.
4
(.
,Xl .
c ;
.,g pheto 13.
of the u ine s'k d ( ew' lookin down an to h es )
i ,
t .
. 7 c- ,. , ,,
l f""* "' , , - ,
g <; - 1u i l
n A . H '. %d e., , -'
photo Is. )
i llearing pads 3 and 7 show signs of heavy rubbing and smearing .
i !
l i
/. . I, l .
l
, i
...A l
i r
j t
rheto 14 llearing pad 9 was found behind the control coasole (view from the cast
' side of the turbine skid looking north).
1 C_ _ _ _ _ _ _ _ __ _ _ _ _ .
1
, l l.-
( e l
- ),
{
)
K' ' "~[ '
N1r
,. OY L * ,,
~
f 6
' W .
f h ' -f . ;
43 . ,
% '?
j
, er photo 15. '
Shown is the thrust bearing cavity mount (part I), pivoted shoc housing ,
(part J), thrust bearing mount (part K), thrust shim (part L), and the oil seal (part M) m,--~; ,n : ~.-
.mp3
~ ~-
,,.- , --T .g
), ,.
,,j ~, , , : .
~
, ' P.' . .
,d.-
gjb - .* . ., .
~~ '
.i(.
.i 4 ,
, . s .
. v s.. -
r'noto 16.
The thrust bearing maint (part N) and the thrust shim (part O) as found between two oil pumps.
B
..-m.----w-_++n-,,, - _.
- , , e*-%w-,--r.-._-_- .,_-_ - __. - e -- _ _ _ _ _ _ _ - . - _ _ _ - _ _ _ _ _ - - - - - - - - - - - - _ - - - - - - -
~+
y .
L
. ; . * . ., ~~ -
1 [\. .' 1 ,/,[
~ , 9 e . .-
\ ~ '
\ t ,f,; f y.d, , i .tt l f ,
i photo 17.
The inlet thrust bearing oil seal (part P) is visible above the upper bearing housing.
1
~' '
~
l l - ,
1
.t.
l 4,
~
[
j.v . i o.. . .
'v:yl*/ '
AZ <
. ./ ' S r , ,. -,
f ,
~
% ,, ,, i
- ?;
photo 18.
- A portion of the inlet thrust bearing oil scal was also found near the turbine shaf t piece.
t . _ _ _ _ _ _ _
. 1-s
}- **[
4,(
.. . .- - . k, i
g .. e
, V i i_
I .. e ..
\' '
.* ~,. ' l9
. V s
( ,
,. .g . .
qN g
pr , g <.
n K s - '
4 /
O
- { \
~
f,M' 4
i ~4
+
' g -
r.
8 o
- . , j,'C 4Q,, . ,'
y 1i - . 9 . k . .<i Y:{, JJ.,g;;,
n7 . , ,. 4%%
g[;', ,
- ~
F ; , ;' . i .*J
.g ;n# . n ,
g,' ge, ' *
- s
, #/ ... .
. ' % 6 fi
_,, i photo 19.
The oil deflector (part R) and the thrust shim (part S) can be seen between the AC and DC lube oil pumps (view of the south end of turbine looking down and west).
4
./ % ... 1 M r ,___.-- ,
c s l
, , %,y 4 ^i9 L , ,, a- . L -ip '- +' .- m A o , g,. .
- 7. <h
' '#/ !
.s t l7 s.
.,s; ,. r :,r ,g ;5 s - . t
\ i( ,.
' Y .: ? ..l 7',,%[k5- !
p .; l ~ f.. 5
. '. . ~ j <. ,u. .
l -
'y
.1, , 41 3 . '
\
@ 9 st- N. :
1
.o' j l
. I y :r:. . , ,
. I photo 20.
l The bearing cnd plate as seen from west of the turbine shaft center line ,
looking north.
1 l
-w- - ---e.ee--,as - w--ww .-,e.a,__ - - -w----.-,---------s-- -
l i
l s .
g
- e 1
e ,'
~
[ { g .g .,
s^h , , .
e %- ~
.. r' y v W ;.
.s
/ *
. j i- hfA .
,e - .
,, l i ,> . ;-
9@. . .
, s ., .
N 3 ,
f M [ *,* ' W -
i l&
.F','4 ,
97 g ' *k n a'
.t'. .
photo 21.
Ilroken bolts from the split bearing housing were found near the bearing oil return bellows (view looking north). t
.1
-,p I
h 1 li; - l l
e L !
I,*: *. *'. ,
7'
. *; , , 'Ii '
U,': y', ,g ,
. t* I b "'j ll
- m. g de n' 'M' --
l photo II.
The pivoted shoe journal bracket (part V) and the north-end oil seal !
(part W) were found just inside the enclosure access door. The view is from the northeast side of tiie enclosure near the pump looxing down and southeast.
l l
--m. _ _ _ _ _ - , _ . _ _ _ . _ _ . . . _ _ _ _ - - - _ _ _ _ _ . , - - _ _ _ - _ _ _ . _ _ - . . . . _ . _ . _ _ . ---_-._I
l 7, s r, 7,2, . .,y
?
s ** j.
, :::y ., , ,.
s , 9 ._ \
~
l t**. '.' . !
1
,' f #,'d,,e -
r
' H **
{
b .
l J';
.&..Kr '
e , .- , , ,
i
- k l ,
photo 23.
The turbine north end oil labyrinth seal (part X) was found in the north section of the turbine enclosure (looking down and cast).
4 cy y va- s-
, 3 i ,. h. ~'
l
, ;~ n . ;
-lf.') ,, .
Y(*
- d--:;;-:h)?Y,Q \
1' y ,
+t.
~
e , l i
<3 1
- $,j s
+ ,
l l
iY I g.
k
. ,. . [. -
1,
<t _j f
( , m . .] [
photo 14.
The pump drive coupling flange is shown after impact with the enclosure roof. Flange distortion is visible in this view looking down and ,
northwest. ,
i t j i
' . _ _ _ . _ _ a
\
r
', r
\ g.
\
o
$ :. ~ 4.,
x -
l
. l photo 2s.
The turbine turning gear bcodix gear (part Z) was found inside the northwest access door of the turbine enclosure.
' l i
l
- l l -
L . . :. '. ..
X '
i
/ 4. i photo 26. l
, 3 The turning gear (part AA) as f
- r. seen from the northwest access !
. .-y. , --
door (looking down and cast) l on the turbine side of the ,
enclosure, !
, ,l , I
, t a*
l 1
. N l l l
I j
l i
-~ "
1 .
- ,e. , .,
,' v l . 488( ' Og ': , .
Q 1 y
l . >
I j photo 17.
I Photo of the south outboard thrust bearing (part AE) looking down into the south bearing oil return bellows. l i
l k
1 .. ,.
! ?. I i, t , j t ,
f
_;s ;
! o .
1 t
l l
. . .e -dr l l -
l 3
5 "-
iAF ;,$ f l i' ', ^d i
,,> l;4 'j
'i l i
- // tc._d,.-
l
} . v. .> n: .
l ',.' d I
! photo 18.
't he as found position of the manual overspeed trip valve (part AF) view [
just west of the turbine center line looking north.
i I f
- m. -. , ,
e
,~. r :7
..=u '
EM i ?
,/
- r. -
_ r ,
f
- Q 'N ~ , o
,\
' n .
afs4 -
_y . . .
( _
, ,..:[ ~
'\ , *I f , v.. ,.
., .- u # ..
(
I.
~
..+ s .
?
- s. .
phote29.
A piece of the diaphragm (part AJ) was fouad in the L.P. turbine after the case was removed (view looking south).
l f
4 l F l
\
(/
\-t TE
- ) -
,/
s' .[: [K ,
- k; ,I f$f h l{ [ g l 0 .
~'
'l, ,
^ , ,
'* i ,
Y.lm, photo 30.
Photo of a gear from the turning gear assembly was taken from the northwest access door facing southeast.
i ,
f i
. ,M* .- (~''} WY.
- O ?
~ .l p
. 'g)f, ,
- f, {p; _
i , ~,- ,ht e * % *. . ' i <:~ r~ j) o # 1 . f
.~
e j j(/. l [, . l f*t '/ - w- ' p
.a~e, ~
a . rA - p ; *y, _j- ,o 's" ~
*g:.. . . - r
!!h *'o. '
ff ' !$
. / l! , ' f _ '
'i . .,
/ .
, t .
t
, s
, 94. ' >
3
- ~.. s
.\
t l
- * ,A l
(
~
~
% c
~ ~
. . 1:+ , ;,y . %:t.c :
i photo 31. ' A piece of the diaphragm (part AG) was found in the exhaust duct. I
- '!E' TV . l. M,[ , ~ . ___. .. .
j ,
,.-' s' I
~
l *
. if ' ~' '
L *' 1
' h
. t . L _ .) .xm q .rk-
! s .- i - Al I . .. f:\ - I A t + g ,, ', o ~ . l ' __l .- \ r ? . f . phote33. , Another piece of the diaphragm (part AI) was found lodged in the exhaust duct vertical run. i 1
- ,6 .
k
- . .;e
~N ms 4 '3,.
.. -, ex,.m y% g ,
-[, '
y M' .
" 't
.>~. gsRt Qy
..4 i,
*' d I"
f,- q j ,
/.
1 .
. , g ,
l
.. \
s
.E photo 33.
The overspeed trip device (part AF8) counter weight and shaft fragment were found in the south oil sump. d { i i I i i I l l 1 l I i j t t
APPENDIX C "VESTINGHOUSE TEEDWATER TURBINE STOP KALSI ENGINEERING AND THROTTLE REPORT, VALVE A NALYSIS UNIT 1 DOCUMENT #1566C, DATED AUGUST 2, 1988
( Page 1 of 57 WESTINGHOUSE FEEDWATER TURBINE STOP AND THROTILE VALVE ANALYSIS, UNIT 1 DOCUMENT NO.1566C, REV.1 AUGUF,T 24,1988 Prepared for Houston Lighting & Power Company South Texas Project Bay City, Texas Prepared by Redewed by Adu
'P. lYaniel Alvirez ], ---- M. S. Kalsi "ThD, P.E.
QA Approval by b ' Tod'd Hors't i UNCONTROLLED COPY REVISIONS Prepared Revie w ed Q/A Pages L No. Data by ty Approval Des <nption of Changes Affected 1 1 4 24M PDA MSK TH Corrected serier stretch ree Unit 2 Valves 22 24 l l ' KAL51 ENGINEE AING, INC. wc-ANcx cesoN r. e.AL"s.s as== c _ ,e we , > a4:< xc l l
Document No.1566C August 2,1988 Page 2 of 57 TABLE OF CONTENTS Page
- 1. Introduction 3
- 2. Results and Conclusions 8
- 3. Recommendations 11
' 4. Results of Investigations and Analyses 12 Appendix A Required Spring Load Analysis 23 Appendix E: Spring Preload Investigation 21 Appendix C: High Pressure Stop Valves No.11 and No.12 Damage Report 25 Appendix D: Bypass and Pilot Valve Flow Resistances a Appendix E: Spring Rate and Preload Test 49 Appendix F: Valve Subcommittee Summary on Root Cause Analysis 54 MAL 51 ENGIN EERING, INC. ,
..c- sc..cinosa..s.....
5
Document No.1566C August 2,1988 Page 3 of 57
- 1. INTRODUCTION AND SUMMAIW A feedwater pump turbine at Housten Lighting & Power Company's South Texas Project Nuclear Power Plant Unit i failed destructively during the loss of off site power (LOOP) test on May 25,1988. To support the overall root cause investigation of this feedwater pump turbine failure, Kalsi Engineering, Inc. was engaged to independently evaluate the possible contribution of the turbine steam supply stop and throttle valves to the failure. Ti.is report documents the results of this investigation and re ommendations for corrective action.
Units 1 and 2 each have three feedwater pump turbines, identified as numbers 11,12,13 for Unit 1 and 21,22,23 for Unit 2. Each turbine uses a high pressure (HP) stop valve, a high pressure governor valve, a low pressure (LP) stop valve, and low pressure governor valves. The high pressure stop and governor valves are of Westinghouse design and are incorporated in an integral body construction (Figure 1), whereas the low pressure stop valve shown in Figure 2 is supplied by Gimpel Valve Company. In order to stop the turbine, both the high pressure and low pressure stop valves mut close. The root cause investigation included a detailed inspection of the HP stop and governor valve internals for valves 11 and 12, the condition of the actuators, springs, and linkage arrangements used on HP and LP stop valves; a review of the photographs showing the condition of the valves and stem position after the turbine failu.e; discussions with varbus HL&P and Bechte! personnel, including Mr. Dennis Stark of HL&P Systemt. Engineering, regarding previous history of problems, a thorough analysis of the forces acting on the valve stem and plug under various operating conditions, determination of minimum spring preload requirements to ensure fail.close action; actual measurements of spring rate and tests to determine if therv is any wound up spring pretension in HP stop valve springs by a load cell; cnd determination of installed spring preloads based upon initial free length and stretched length under installed conditions. Additionally, analysis was conducted to estimate the effect of flow resistance of various clearances and equalizing passages so that any favorable or unfavorable pressure differences under normal clearances or degraded conditions of the valve internals on the valve closing forces could be quantified and taken into account. The overall conclusion from these analyses is that the failure of HP stop valve to elete is the root cause of the overspeed and destructive failure of turbine No.11. 'Ihe valve failed to close KALal ENGINEERf NO,INC.
.m -.s.cw : .a u .....o
, - -_. - - - _ - _ ,. _ _ - -- - . ,m
Document No.1566C August 2,1988 pt.ge 4 of 57 s.' ..
\ r
%. \ / l
% \ / /, .
Y & . N 'y/ ,< y"' I !./' ,...,', %, '
.. * ,.......%..... I ...
_ S. _ .,, LL.a/ ll, _ m
.n. _ . i ,. 1 rg .
m '" 8
/ /
' ~~
r ae kn. _e_
. . . n,
, . +. ,.r. -,%
cs, n.g t j,, u n
=
/ p '.7
.t .g. 7 p _ .s / i fpE-My-s Ng,/'1_/ "3 t.
i g-
.s'_' 4,. f .._ , i: .
Nex
, .L. l
.z i "
5.'.M'.4 M 4 f
. f.i: i ~ .
l Uf '4 % pQg . ; 6.
. i 3 ,. -
-- l M
C#MP
',1,,, [ C %
'g I e i
_.' E
-: .. M i --- c h ,1, dm 'J "Ti ; b i
I
- . V ,f /'
hM f
/g ! ]
/. ,
/ Ii Ll I il a.a d'< l ' i.- '"f=
i N' l i FIGURE 1 HIGH PRESSURE STCP AND THROTTLE VALVES M ALBI P.NGIN EERING, INC. ; w c-. :..c .o s. ...
x Document No.1566C August 2,1988 Page 5 of 37
- :. :: n. .- -
. s ,- - ' cc-t s ,*s..T s.
, n,,. s.
. e. . /. . f i , . . .. t s.-
w- s i s, .-
?. ...'.s......w..
., ..y...
N \ ;. . : .c. i . '
I. 4 , y , , , ,.*.4.. ,, .~-i.
.etr/
' v A X4y ( .....:...
yI / JQ ... .-i 3 .c ! j j # '. ----- . . --: e ,
, /
s-ps a o J, 5 su 1 e,=. ., ... s: ::.-
,7..
c ; f ,s v
.n .. a
[T'.RI
! 8,~ jN, p..5 $N . ',l C.' ' "'"
l i q l'.s M,C'. ' s aW'$av- h:. .s.r. ss- 2 _X ."~s. 4y' a u- a r- ' } t. v ,..si
- i i ,m
?:! - - c.
m:=-+ ;
.u .' .. .-c .
t t
'l
/>; ;q..-' .::-
y' .>
~
- g. ,,1.......
~ g ,, f i .j ,g:q., :. ~s .. ., .e...
,1 c.,
,3}a ._,
- a, .p g r.- . . . . . < .
i
,&x _ _. <
~*
t 1,
=s.
= .c, ,0t.:'os..
* . n sa!,
<es
.u..
- ,u ,u-m...
se. .. M ,gI . 4'l'*,,. ,'o . p ..- 4
$. j ,. w , 4 . p
@ 'T W ;o.fwu o,. o c., ,,
,?'.;l.t.
4H' ;'!.., 1
. j. g: a Ng E,--- M .&i.~N q -. $ 0k g i,
. i
. ; f 4 !
;i @. .. , ; -
d. i ?'-+- _ -- /Mx s w #.' v,t. ' - - l
......n.....,.....
.....u..n........ ......
i I-ii, _ _ x_ / .:
@ -- i ). !
l .
. i r _ ...- -.. ,i innte 1: vs.it L- aff_
= t. c 7,'."'..j,.*,',6*'-.
}
.g
.-l, . . . . . . ' , . c.:wss <--4 ,],3777 FIGURE 2 LOW PRESSURE STOP VALVE KAL51 ENGIN EERING, INC.
tc c..c usi.*w.6.
Docum:nt No.1566C August 2,1988 Page 6 of 57 l j due to insufficient spring preload caused by improper adjustments, it was found that no written procedures were available for field personnel making adjustments on these valves. l This is believed to be the main reason for insufficient spring preload on all the valves in units ! 1 and 2. Additionally, even though the linkages eennecting the stem to the lever arm through a I clevis a-rangement for HP stop valve No.11 were found to be in a corroded condition, which l possfoly contributed to galling of the stem in the bushing area, this was not the root cause of the valve operability problem. The review of the HP valves also showed that hard surface treatment will be required on the stems to avoid galling proble.ms that have been observed on neveral occasions in these valves. A review of the LP stop valve) was performed by obtaining the necessary technical data j regarding the materials, :ritical dimensions, spring preload and rates from Mr. Doug Nolen, field service engineer with Gimpel Valves. These valves have vertical stems and use an in-line spring arrangemer.t with no options for field adjustments, the weight of the plug is ustd directly to provide the closing force, and the valve design is such that the closing force increases as the system pressure is increased. It was concluded that the LP stop valves have suffie:ent closing force, and they are not suspected to have contributed to the turbine failure. The review was supplemented by actual inspection of the LP stop valve internals by Mr. Tom Day of Westinghouso, it was reported that all the internals were in an excellent condition with no signs of galling and that the LP stop valves had performed reliably ever since they had been installed with no history of ary problems. A review of the HP governor valves was also conducted, which showed their basic design to be adequate and their final position after this turbine failure to be consistent with what is expected after a loss of hydraulic and electrical power. Also no corrosion or binding problems with their linkages or internals were found. Based on the results of the above review and conclusions with which Westinghouse concurred, a plan of corrective action was formulated. However, during controlled tests following the spring preload adjustments, it was found that, in spite of the highlighted problem of insuf1'icient preload which had created the turbine failure, the springs were again inadvertently adjusted below the minimum recommended preload. This problem was caused by improperly written procedures. This resulted in the valves not fully closing during the KALBI ENGINEERING,INC. le*4 C seN * &4 CG % 3% Gb 4%A6# St &
Document No.1566C August 2,1988 Page 7 of 57 controlled test. Even though these tests were unsuccessful due to the failure to correctly implement the needed modification, they provided actual data regarding the total frictional forces present, which previously had only been estimated. The actual equilibrium position of the stem and the system pressure present in this test were used to calculate the in-situ friction of the piston seals and all other friction contributing elements of the system. The magnitude of the friction was found to be higher than that used by Westinghouse in their spring sizing calculation, it was also found that Westinghouse had used a lower system pressure of 1,060 psig in sizing the preload on the spring, whereas the actual system pressures is higher ' (1,285 psia plus 10 percent accumulation for relief pressure). To account for higher friction and increased system pressure,it was calculated that sp-ings will need to be preloaded to 2.0-inch instead of the previously recommended value of 1.5 inch. The new adjustments were made on all the lip stop valves; however,it was found that valve No.12 was still leaking excessively even though the stem was in the fully closed position after the adjustment for the increased spring preload was made. Upon disassembly and inspection it was discovered that an improper procedure of rotating the clevis had been used to increase the preload which caused relative rotation of the stem and plug that resulted in circumferential galling of the seating surfaces, it was confirmed by discussions with the field personnel involved in making this adjustment that this procedure was only used on valve No.12. Therefore, valves 11 and 13, which were sealing correctly, do not need to be disassembled to correct this problem. The results and conclusions from this independent review are summarized below, followed by supporting appendices which give the pertinent analysis or testing details. i KAL51 ENGINEERING, INC. Vice*AN C a. C S SGN 8. AN A.* S S
Document No.1566C August 2,1988 Page 8 of 57
- 2. RESULTS AND CONCLUSIONS Root Cause of the Failurts
- 1. From an independent review of the turbine valves, which included an analysis c(' ces required to slose the valves, measurements of critical dimensions, spring rates, and determination of any wound up spring pretension,it is concluded that the high pressure stop valve No.11 actuator springs did not have sumeient preload to overcome the blow out force on the stem due to a system pressure of1,160 psig which was present at the valve inlet. The root cause of the turbine No. Il failure during lor.s of off site power testing is, therefore, directly attributed to the failure of the HP stop valve to close.
- 2. It was determined that the as found spring preload adjustment on HP stop valves 12 and 13 were insufficient to ensure closure of the valves against the 1,160 psig line pressure.
Additionally, all the high pressure stop valves, even in Unit 2, were found to have insufficient spring preload.
- 3. The insufficient spring preloads found in all the high pressure stop valves are directly attributable to lack of written procedures and proper training of field personnel responsible for making such adjustments. As discussed in Conclusion 9, even the specified spring preload was found to be insufficient to cover the maximum pressure condition and actual system friction, both of which were found to be higher than the values assumed by Westinghouse.
Other Factors Affecting the Operability and Closing Forte Margins
- 4. The stem of the high pressure stop valve No.11 was found to have galling in the vicinity of the stem bushing. This galling is believed to have been caused by a combination of the weight of the lever assembly carried by t.he horizontally oriented stem and binding of the linkage which occurred as a result of corrosion.
- 6. Even though corrosion of the stem linkages on stop valve No.11 was not the root cause of the failure t,f the valve to close,it erested additional friction due to binding of the linkage
.lements which reduced the available closing for;e and the final equilibrium position of the plug. The actual margin of the effect of corrosion depends on the degree of binding caused in the linkage, which is hard to quantify. As shown in the detailed calculations, KALEI ENOnNEERING. INC. g vu. .w..use,s.. ~ ss
. .m
1 Document No.1566C August 2,1988 Page 9 of B7 the margin of spring preload was insufficient to esercome even the stem blowout force due to sy. tem pressure. Thus, corrosion effect in this valve was only secondary. Corrosion of the linkages should be prevented so that binding, stem galling, and the associated stem friction does not reduce the closing force margin.
- 6. Galling was found on the stems of valves 11 and 12 after the first disassembly following the turbine failure. The stem is made of 316 stainless steel without any hard surface treatment, which makes it very prone to galling. Nrd surface treatment such as nitriding shculd be used on these stems its eliminate the galling probleras that have been observed. The use of 316 stainless steel under sliding applications should consider the potential for galling. Although under normal operating conditions these valves should not experience any side load on the untreated sliding components, misalignment in the actuator lever assembly, the weight of the actuator assembly being supported by the stem, or line debris lodging in the small sliding clearances can generate stress levels that can create galling.
1
- 7. The valve bonnet uti'izes a small equalizing passage which communicates the upstream pressure to the area on top of the plug. A hydraulic resistance circuit model was made to analyze pressure above the plug and the effect of varying flow resistances of various equalizing passages between the upstream conduit and the top of the plug as well as the flow area across the pilot valve. This model also included the resistance of clearance gaps that are present between the plug guide bushing, the plug, the stem, and the stem bushing. The effect of tolerances of various parts as well as expected degradation of these parts on the resultant pressure above the plug for a range of upstream and downstream pressures was analyzed. The results showed that it will be desirable to increase the area of the equalizing passage to ensure that an unfavorable pressure difference across the plug, which would oppose the spring closing force, does not occur when the stem bushings have degraded.
Pmblems Due to lack of Adequate Field Adjustment Pmoedums
- 8. As stated earlier, it was found that there were no written procedures available for making spring preloau adjustments. Even after some procedures were developed, several problems were encountered in the correct implementation. During the first field M AL51 E NGIN EERING, INC.
. c .. .c . , c . e n. . . . . . n ,
Document No.1566C August 2,1988 Page 10 of 57 adjustments immediately following the failures, insumeient preload resulted due to lack of clarity and appropr4ta sketches in the written procedures. During subsequent field adjustmen% an improper method (which consisted of rotating the elevis while the stem is in compression and pushing the plug against the seat) was used to increase the spring preload. This procedure caused galling of the stem pi'ot to the plug and the main plug to the main seat areas on one of the valves. Due to this galling, tight seat closure could not be obtained even with the plug in its fully closed position, which allowed the steam turbine to continue to turn at a low speed instead of completely coming to a halt. There was no precautionary statement in the adjustment procedure regarding how to avoid galling. Requhrd SpdngIhload
- 9. After determining the root cause of the failure and reviewing the spring preload requirements with Westfrighouse personnel,it was decided that a preload ofi,200 pounds from the two springs, which entresponds to an initial stretch of 1.5 inches should be sumelent. This was based upon a system pressure of1,060 psig, a 10 percent allowanee for friction from linkages and piston seals, and a 25 percent closing force margin.
However, the minimura set pressure of the main steam safety valve for this applicatior 4.s 1,M5 psia, and an allowanee for 10 percent pressure above this for safety relief valve settings is needed. Additionally, frorn one of the controlled tests in which insumeient spring preload was inadvertently used, the in situ magnitude of frictional forces, determined from the equilibrium position of the stem under known system pressure, was found to be higher than the 10 percent value natumed in the original preload calculations. To properly account for the higher friction and higher pressures acting on the stem and provide suMeient closing force margin, it was calculated that a spring preload of 1,600 pounds, which corresponds to an initial stretch of 2 inches, should be used. Non I%blems
- 10. A review of the HP governo' valve, the LP stop valve, and the LP governor valves showed that none of these valves had any problems or deficiencies that contributed te the failure of the turbine by cverspeeding.
K ALBn ENGINEEReNG, INC.
- 1 s : . c ,a s s. *s.. E
1 Document No.1566C August 2,1988 Page 11 of 57
- 3. RECO5BTENDATIONS
- 1. Clearly written procedures and trained field personnel should be used to adjust the high pressure stop valve linkage to obtsdn proper spring preloads and travel.
- 2. The high pressure stop valvo springs should be preloaded by 2 inches, which is based upon actual measurements of friction in the linkages and piston and a maximum system pressure of 1 A13 psia (1,2b5 psia design pressure plus 10 percent margin for accumulation pressure due t: r.afety relief valves).
- 3. A new spring design should be made that will meet the spring force requirements and will allow full travel of the stop valve.
- 4. All the sliding surfaces of the high pressure stop valve internals should be hardened to avoid galling and high wenr. Hardening of these components can be accomplished by surface nitriding.
- 5. Spacers made out of corrosion resistant niaterial .such as bronze should be incorporated in the linkages to minimize friction and binding which can occur due to corrosion.
- 6. The area of the equalizing passage in the high pressure stop valves should be increased to ensure no adverse pressure difTerentials across the plug in the degraded conditions of the stem bushing. This can be accomplished by incorporating radially drilled holes in the valve cover.
- 7. As suggested by Mr. Gary Parkey of HL&P, provisions should be made to perform in.
service tests to determine the operability of the HP stop valves.
- 8. In general, a small mating angle between the stop valve and its seat can be used.
However, the use of an angle less than 30 degrees is not recommended due to potential sticking of the plug in the seat. It is true that better seating was obtained on high pressure stop valve No.12 after the mating angle was changed at Westinghouse's recommendation, but other changes made concurrently on this valve make quantification of the improvement due to angle change difUcult to access. This change was not made on the other valves and they are functioning properly. KALSI ENGINEERING. INC. w c -.w.. ei . c. . .... . . . .
Document No.1566C August 2,1988 Page 12 of 57
- 4. RESULTS OF INVESTIGATIONS AND ANALYSIS The results of the investigations and analyses performed on the high pressure stop valves are documented in the following appendices:
Appendix A contains the analysis performed to evaluate the required spring preload. Appendix B documents the findings of the inspection and investigation perfermed to determine the amount of spring preload present in the stop valves on both Unit No. I and Unit No.2. Appendix C documents the inspeetion of the valves for galling and potential malfunction. Appendix D contains the analysis performed to determine the effect ofincreasing the flow capacity of the equalizing passages from the inlet port to the top of the plug. Appendix E contains the results of tests and analysis performed to determine the spring rate and spring pretension of the tension springs used in the high pressure stop valves. Appendix F contains the summary of the review and corrective action formulated concurrently with Westinghouse. KALBI ENGINEERING. INC. v i m. s : .. a . ..s ., . . ... . . .
? ;
/,E f
/ /,' Document No.1566C Appendix A August 2,1988 Page 13 of G7 V.
t 4 APPENDIX A REQUIRED SPRING LOAD ANALYSIS 4 W. i 1 MALSI ENGINEERING. INC. w :-.sc.. e4 so% s *s. .. c.
- - . ~ - - - - , . , . - _ - - - _ , _
Document No.1566C Appendix A August 2,1988 Page 14 of 57 APPENDIX A REQUIRED SPRING LOAD ANALYSIS A.1 INTRODUCTION This appenc'ix documents the analysis performed to determine the required spring force to ensure closure of the plug under normal operating conditions and unde maximum design eenditions. To enstre complete closure of the plug, the spring should have sumelent load, including some margin to overcome the stem blowout force plus frictional resistances at the hnkage, stem, and piston sesls in the hydraulic actuator. This load should nct depend on dynamics such as plug inertia and now. The margin provided should be sumelent to overcome any expected increase in load due to degraded conditions of the valve and linkage assembly caused by galling and corrosion. The required spring force is maximum when the downstream pressure is highest (no now conditions). Even under now conditions the assistance gained by pressure balancing the plug l via the bypass holes in the valve cover becomes signincant only when the plug nears the seat. The stop valve's Oow is such that it does not produce eny signincant pressure drop until it is almost completely closed. Pressure balancing the plug should be regarded only as an enhancement to the elesing action and should not be considered as positive assistance, i Although limited test data was available to determine the level of frictional resistance that might be present in the valve assembly, data ebtained during functional testing of the refurbished valves can be used to quantify these frictional resistances. For a healthy valve, one that does not have corroded linkages, galled stem, called plug, or stuck piston, the known and dennitive forces opposing the closing motion are:
- 1. The stem blowout force,
- 2. The friction resistance of the stem in the bushing due to the weight of the assembly, and
- 3. The frictional resistance of the seals of the hydraulic piston.
KALBI ENGINEERING,IN C.
$.u -.s :. .. c . o s u .s.. e
Document Ne.1566C Appendix A August 2,1988 l' age 15 of 57 Summing these forces and adjusting them by the leverage ratio of the lever will yield the load required by the spring, which is: F= F 3+Fr -F L 3 p p s FB , = Force due to stem blowout
= 1,427 lbs for system operating conditions (Ref Section A.3.1)
= 1,717 lbs for system design pressure (Ref Section A.3.2)
FF = Stem friction load cause by th. ar assembly weight
= 47 lbs (Ref Section A.4)
Ls = Lever ratio between spring location and stem
= 1.6 (Ref Section A.2.1)
Lp = Lever ratio between hydraulic piston location and stem
= 1.11 (Ref Section A.2.2)
Fp = Hydraulic seals frictional resistance
= 100 lbs (Ref Section A.5)
F=3 ? # 000
- l'll}
6
= 819 + 111
= 930 lbs for normal operating conditions, and, F,=2.72';4'.n00x1.m
= 980 + 111 l = 1,091 lbs for system design pressure KALSI ENGINEERING. INC.
vi c .s c. c 2 . .s.. . . ___________________._____.___________________.____j
Document No.1566C Appendix A August 2,1988 Page 16 of 57 The minimum required spring stretch for two springs to attain this load is calculated using the spring rate. Smin = 930/(SRx 2) Sa = spring rate
= 400 lbvin (Reference Westinghouse Drawing 400A204)
= 1.16 in for normal operating conditions and, Smin' = 1,091/(SR x 2)
= 1.36 in for design pressure Although analysis shows that the additional spring load required to overcome the calculated l
l frictional resistances is only 137 pounds, testing 4 hows that the actual frictional resistances are actually 367 pounds for e refurbished valve as shown in Appendix B, Section B.2. Some margin, however, should, be provided above this as.found frictional resistance to ensure that the valve will close even under degraded conditions. The actual margin required is dimeult to quantify, but should be sufficient to overcome any additional resistance due to misalignment, corrosion, or galling of the sliding components, including the hydraulic cylinder. For this application, and kno ving that some uncertainties do exist, applying an additional 500 pounds above the calculated minimum force should provide an adequate l margin. l l Using 500 pounds as the minimum margin ofload to ensure complete valve closure, the final spring prutretch becomes Snom = (930 + 500) /(SR x 2) = 1,430 / 800 s
= 1.8 inches for normal system operating conditions {
i and, Snom*= (1,091 + f 00)/(SR x 2)
= 1,591/8/9
= 2.0 inches for maximum design pressure MALal ENGINEERING,INC.
co.w u.as .s.....
Document No.1566C Appendix A August 2,1988 Page 17 cf 57 To prevent spring overstress when the valve is open, the total stretch of the spring should be limited to Ave inches (Ref Westinghouse Drawing 400A204). Limiting the total spring stretch to Ove inches will result in slightly less opening of the plug but will not affect the flow performance because the valve's present travel is much greater than is actually required for full capacity. A.2 LEVER ARM ACTUATION RATIOS The stop valve is actuated by the spring and hydraulie cylinder through a lever mechanism which multiplies the applied force. The amount of multiplication is di.Terent for the spring and cylinder and is detennined as follows from the geometry of the lever (Figure A.1). 54 57,$ - -'
- ,fy
- 6
- o d.. ._.
../
1 - %- g ~.
/ ,
./
a) l
, h ':.'h .,f _
s' ,/* - l
\
i l 1 q' l I
'lf i
'kd l I 11 '
$.I ,i[id;9 JUtDdi
' 'l1l.{ilI9 [._-__. ,,l L _. . . - . - . _
,'t ' "
7A ' > %, 1) l ,
\
FIGURF, NO. A.1 l A.2.1. lever Ratio Betwn Spuing and Stem 14 = Ratio of the spring force to stem load a 32.375/18
= 1.8 KALSI ENGINEERING. INC.
u -. c..e.,w .s.....
7 Document No.1566C Appendix A August 2,1988 Page 18 of 57 s
?
A.2,2. Isver Ratio Between Hydraulic Piston and Spring LP = Ratio of the spring force to stem load
- = 36.0/32.375
= 1.11 A,3. STEM BLOWOUT FORCES When the valve is subjected to presst tre the stem is urged out of the bonnet by a force equal to the pressure Genes the crors. sectional area of the '., tem. This force needs to be overcome before the valve can close completely.
A.3.1. S' w Blowout Fom Under Nonnal Operating Conditions FB = As x PN As = Stem Area
= nx D2 /4 D = 1.25 inches a r. x 1.252/ 4
= 1.227 in2
."N = Normal system operating pressure
= 1,160 psig i FB = 1.23 x 1160
= 1,4271bs A.3.2. Stem 1Mowout Forte Under Marimum System Design Pressure The maximum stem blowout force occurs when the pressure is as its highest. This pressure corresponds to the relief valve settings that control the pressure which can exLt for the system.
FB = As x PM As = Stem Area
= 1.227 int Pg = Minimum Relier Valve Set Pressure
= '. .2r A + (10 4 x 1,285) (From HIAP System Spe;s)
= $ . . ds
, : 15)
*. '. L ee' E NGIN E EV t' ' s. i " O.~
~
.% .ceev.s.s s,ss~
Document No.1566C Appendit A August 2,1988 Page 19 of 57 A.4. STEM FRICTIOMAL LOAD DUE TO ASSEMBLY WEIGHT Frictional load at the stem occurs in this valve due to the installation orientation. The valve is positioned with the lever arm oriented horizontally with the pivot point at one end forcing the stem to provide the support to hold the arm in place. The load at the stem is determined using the v. eights of the components and their leverage effect on the :te n. I4ad due to initial misalignment of the assembly is not considered. W gp x 32.375 W L2 x 36 W s"W L+ W g+W e+ l8
+
18 The weights of ti.e components were calculated using measured dimensions or by actual weight rnessurements. WL = Weight of the operating lever
= 61 lbs WLt = Weight of the clevis links
= 14 lbs WC = Weight of the elevis
= 4 hs WsP = One. half the total weight of the springs
= 321bs WL2 = One. half of the total weight of the pisten lir.ks
= 91bs W 8 = 61 + 1A + 4 + T x 32 375 , 9 x M 18 18
= 1551bs The sliding resistance of the stem in the bushin= is cateulated using the weight supported by the stem and the coemeient of the friction between the stem and. the bushing.
Fr = W,a x Cr Cr a coefficient of friction
= 0.3 (for steel on steelin steam and no natural surfmee degradation)
= 155 x 0.3
= 46.51bs KAL 51 ENGIN EERING. lNC.
vu-*s c.. m .c.s r. s.. s s
Document No.1566C Appendix A August 2,1988 Page 20 of 57 A25 FRICTIONAL RESISTANCE OF THE In'DRAULIC SEALS The frictional resistance at the hydraulie piston is caused by the initial compression needed in the seals to ensure s.n effective seal under low pressure applications. Depending on the t>Pe of seals used in the hydraulic cylinder, the load can vary from 30 to 100 pounds. The upper limit load will be assumed. This data was provided by Farker Hannifin. Fp = 100lbs j 1 KALBI EN GIN EE RING, INC. w c -as . s :.s e. .si. s. I )
. . . _ .. .- .~ _. - - - . - - _ ..
Document No. 3566C Appendix B August 2,1988 Page 21 of 57 9 APPENDIX B SPRING PRELOAD INVESTIGATION KALSI ENGIN EF, RING. INC. ,
- s. :-*wa. : . : . s. asa. s s
T li Document Nr.1566C, Rev.1 Appendix B August 24,1988 Page 22 of 57 APPENDDr B SPRING PRELOAD INVESTIGATION D.I. INTRODUCTION This appendix documents the results of the investigation performed to determine the amount of spring pretension present in the high pressure sty valves before any adjustments were made and after intermediate adjustments were made. Initialinspections showed that valves on both Unit No.1 and Unit No. 2 did not have sufficient spring preload to fully close the valves under normal operating conditions of 1,160,esig, and under the maximum system pressure,1,285
+ 10% psia, as controlled by the relief valve settings including accumulation pressure, investigations performed after making intermediate adjustments again showed that the ,
springs on the valves on Unit Ns. I still did not have the required preload to fully close the valves. These intermediate adjustments, thought, did provide valuable insight into the actual amount of valve and actuator frictional resistance that should be added to tl.e calculated stem blowout force to arrive at the actual minimum spring preload. D.2. INITIAL SPRING PRELOAD INVESTIGATION The dimensional inspection of the valves on Unit I was performed by llL&P, and the inspection of the valves on Unit 2 was performed by Westinghouse and HL&P. As e.lculated in Appendix A, the minimum spring stretch needed to ensure complete closure of the high ; pressure stop valve is 1.16 inches for normal operating conditions and 1.38 inches for maximum system pressure. These spring stretches, though, do not include any margin for degraded valve conditions. In situ testing under known pressure and spring preload conditions shows that this mir.imum calculated spring prestretch should be ineressed to 1.8 inches for nerms) operating conditions and 2.0 inches for maximum pressure conditions. The as.found stretches for some of the stop valves of Units 1 and 2 were Unit No.1 Valves: Stop Valve No.13: 0.21 inthes (actual measured average) Unit No. 2 Valves: Stop Valve No. 21: 0.8% to 1.01 in. (indireet'y calculated) Stop Valve No. 22: 0.64 to 0.96 la (indirectly calculated) Stop Valve No. 23: 0.75 to 0.86 in. (approximate measurement) hAL51 E NGIN E E RI NG. INC. E.s. :., : 3 v. a. *s . . . . j ,.
Docum:nt No.1566C, Rev.1 Appendix B August 24,1988 Page 23 of 57 As can be seen, the as.found spring prestretch for the valves irivestigated as led than even the minimum required for normal plant operations. The spring pretension for Valves 21 and 22 were determined by using the original clevis and lever arm position dimensions obtained by Westinghouse representatives and correlating these dimensions to the subsequent spring prestretch analysis performed for Valve No. 23. l I Valves 21 and 22 could not be re inspected for actual spring stretch because they had already bean disassembled prior to re inspection of Valve 23. The spring stretch on each of Nos. 21 and 22 was calculated based on the assumption that the valve assemblies were identical in construction and that the only difTerence between them was the amount of elevis engagement on the stem. Observations of the valves on Unit i during operation also proved that the spring did net have the required preload to fully close the valves. V"ien manually tripped, the valves would stop before the fully closed position. Aner pressure was shut off using the m.dn steam isolation valve, the stop valves would then move to the fully closed position. B.3. INTERMEDIATE SI' RING ADJUSTMENTINVESTIGATION Re-evaluation of the spring preload was required when valve functional testing showed that, even aftir adjustments in the spring preload, valve No.12 was still not completely closing (i.e., the plug travel stopped before reaching the seat) aRer being tripped at tl.e normal operating pressare of1,160 psig. Investigation revealed that the spring had not been set as had been intended. The prestretch of the spring was adjusted to 1.0 inch instead of the desired 1.5 inches. At the spring prestretch of 1.0 inch, the stem was advancing to a position between 1/S and 1/4 inch before the fully closed position and then completely closing aner the MSIV valve was manually closed. Misadjustmtnt of the spring preload was attributed to inadequate adjustment procedures. As a result of this misadjust nent, a more precise and definitive adjustment procedure was devsloped. Applying the data obtained from this controlled test, the valve actuator drag can be more r; us ately quantified. Knowing that the spring preloat LO inch and that the valve KALSI ENGINEERING,INC. vi:-.sca.: ass.*s..,ea
Document Nr.1566C Rev.1 Appendix B August 24,1988 Pagt 24 of 57 traveled to within 1/4 inch from the fully closed position, the total frictional resistance, FR, can be calculated as PxA P*s us +F g 1 Then PxA s Fg=F-- 3 ,3 18 Wh e re. Fs = Spring Pri;ead
= SR x L SR = Spring Rate
= 600 lbs/in for two springs L = Spring stretch at 1/4 inchcs from fully closed
= 1.0 + (1/4 x 32.375/18)
= 1.45 inches P = Test Pressure
= 1,160 psig As = Stem Area
= 1.252x W4.0
= 1.23 in2 Therefore, Fg = (8M x U5b _1,160 _ l,g x 1.23
= 1,160 793
= 367 lbs This additional spring load required to overcome the frictiona3 resistance compares to the 137 pounds calculated in Appendix A. For design purposes, the minimum used should be 367 pounds instead of 137 pounds unless justificationi can be given otherwise.
KALSI E NGIN EERING, INC. v :-4 ca. es e os s asa, 3 m
Document No.1566C Appendix C August 2,1988 Page 25 of 57 APPENDIX C HIGH PRESSURE STOP VALVES NO. n AND NO.12 DAMAQE REPORT MALSI ENGINEERING. INC.
.i; sc..en as6 s ..
Document No.1566C Appendix C August 2.1988 Page 26 of 57 APPENDIX C IDGH PRESSURE STOP VALVES NO.11 AND NO.12 DAMAGE REPORT This appendix documents the results of two separate inspections performed. The first , inspection was performed after the initial failure of high presst ta stop valve No.11; the second inspection was performed after final spring adjustments were made and high pressure stop valve No.12 failed to provide a tight enough seal to keep the turbine from turning. C.1. FIRSTINSPECTION Review of high pressure stop valves 11 and 12 shows that these valves were in a degraded condition, which may have contributed to the final closing position of the stop valves during normal plant operation. Valve No.11 had corroded linkages and link pins and a galled stem, while valve No.12 had only a small gall on the stem. C.1.1. High Pressun Stop Valve No.11 Corrosion of the linkages and link pins on valve No.11 was so severe that disassembly of the unit required impact force to retate and separate the linkages r.nd pins from the elevis and 1;ver arm. The photograph on the following page (Figure C.1) shows the corroded area of the lever arm. Referring to Figure C.2, the stem was galled along the top of the stem for a distance of approximately 13/4 inches and extended in the circumferential direction about 120 degrecs. The galled area was on the underside of the stem when in the instslied position in the valve. The cause of the galling can be citributed to the weight of the lever arm assembly bearing against the stem, misalignment of the lever with respect to the stem, and binding of the stem due to the inability of the corroded linkages to rotate when the valve was actuated.
~l l ,_
~
-l M - _ . . _ _ . _ ,
\ % Gati nC*A n d FIGURE C.2 MALBl ENGINEERING. INC. _
b W
__ _ _ __ _ . _ _ _ _ _ _ _ . _ _.___._ _ __ _ _ _ . _ . _ _ . _ . ~ _ _ _ _ _ I i wr, m;v,;" t*; -# - - -
- l %$@;'D.
., y; '. 5 .],:
3"?,-* ' , . j I*T,0f-l\.._c.?'.l'9l[;t},,. ..
, . . . .y . ..... < - < ,
1, (..<.+....:
; ~ ,,
? 9 "~
l
- e . ,, A . . ..y' -, u.- % '* y ;i( ,. *.'.
3 1
;U ,.
l
;,.' ,: ' , g ,4. ;.'
g- , ! .< .,- [ . <Y.'
. , . : ty,.i .
<.=,..,.) -
- e. .
, . .. ,s . :
+
', \ - j g ~- . g t .' ' ' V' ,. j ,* ~
j
', ,y . . . .
';' '* ll g yy '-. l:;
} . , ,. . t; , ,
',. j,.'fQf ' '
- f ',.i h-:
5
-Qll { ,
,4 .,
i ) ,
', . . g. : Y ik' = 5; 7 ;, .
, , s
~. - y .
~ . .
Close-up, corrosion is visible on the lever arm to stem clevis links, v^cre they L were .'rozen to the center pivot point of the lever arm. l i I i l l j *- l , i l I ! \ - . I S} .
.~
i j - T' . .
\'
\ '
O $
+ ,
I t
~
E # ,_ - i e '
~ ~~~ - .
A --
- u I
i
?
II.P. stop valve linkage shown disassembled during inspection of damaged ; tuch.ne Rust is visible at the center lever arm pivot point.(view of the west side looking east). ; i 7 m-
Document No.1666C Appenix C August 2,1988 Page 28 cf 57 All seating locations were in good condition. No gallir g was found on either the pilot valve and se it or the main stop valve and seat. To pr,tvent corrosion of the interface between the links and the clevis and lever arm, a spacer , (Fipr C.3) made of a noble material such as any copper alloy (brass, bronze, etc.) can be meerporated between the corrading surfaces, incorporating the spacer will ensure that the joints will remain free so articulation is not compromised. This modification can be easily
, ccommodated by using longer pins to span the increased thickness of the joint. The spacer should be incorporated in all the high pressure stop valves to circumvent potential corrosion.
m
. , s -
L FIGURE C.3 . I During reassembly, the valve assembly should be checked to determine if any lateral l misalignment exi.ts between the lever arm and the stem or if the stem is supporting the weight (f the actuation assembly. The galling resistance of the stem can be enhanced by nitriding the stem and any other sliding interfaces. C.1.2. High Prvesure Stop Valve No.12 ! Corrosion of the linkages and link pins was not evident on valve No.12. The links rotated freely, and the valve was easily disas.embled. Only one concentrated gall was feund on the stem of valve No.12. The gall was located on the lower end of the stem (reference Figure C.4) where it is cuided by the stop valve. The orientation and origin of this gall is unknown. KALSI ENGINEERING,INC. vic w..: e.c..s.....
n---------- Document No.1566C Appendix 0 August 2,1988 Page 29 cf 57 Y 2
/ '
\ _ _ . _ _ o N oeep Cay
-P.5 4 4 FIGURE C.4 All seating locations were in good condition. No galling was found on either the pilot valve and seat or the main stop valve and seat.
Duri..g reassembly, the valve assembly should be checked to determine if any lateral misalignment exists between the lever arm and the stem er if the stem is supporting the weight of the actuation assembly. C.2. SECOND INSPECTION Review of high pressure stop valve No.12 shows that t) gall marks found on the stem, pilot, pilot valve seat, stop valve, and stop valve seat were the result of turning the stem ag ainst the clevis while under spring load. When making the final adjustment to obtain the desired t spring preload, the ' tem was rotated under spring lead. The load on the stem was transferred to the pilot valve and stop valve. Because the above mentioned components are made of E stainless steel and have a very low threshold of galling, even this relktively low contact stress , was enough to create galling. [ Galling on the pilot valve and stop valve was circumferential, limited to the seating interface, and had indications of rotary tearing of the material. Galling on the stem was axial in orientation. It was in locations correrponding to the small clearance between the stem and stop valve and at the interface between the stem and the bonnet bushing. Galling of the stem can be attributed to detached material from the pilot valve getting lodged between the stem and stop valve with subsequent valve operation causing the galling. The galling on the seat can be proven by analysis to have been caused by excessive contact stress as shown below. f i l KALSI ENGINEERING,INC. 7 vi:- se..:,,asc.s ....
Document No.1566C Appendix C August 2,1988 Page 30 of 57 C.2.1. P0ot Valw Contact Stnss The cortset on the seating surface of the pilot valve is calcu?ated using dimensions obtained during ins,rection. i tifs i b FIGURE C.5 Centact Stress, SC SC = F / AC F = Applied Force at the stem
= 1.8 x (SR x b)
SR = Spring rate for two springs
= 800 lbs/in L = Spring Stretch
= 2.0 inches F = 1.8 x (800 x 2.0)
= 2,8S0lbs AC = Contact area
= n x 1.1875 x 0.125
= 0.466 in2 SC = 2,880/0.466
= 6,180 psi This contact stress is greater than the threshold of galling for most stainless steel mating surfaces, hiost unlubricated stainless steel will begin to gall when the contact stress exceed.
4,000 psi, with some combinations beginning to gall even at contact stresses as low as 2,000 psi. To prevent galling, the stem should not be rotated under load. If it is necessary to rotate the stem under load, it will be necessary to enhance the interface material at the pilot valve. Nitro),ardwning both interfacing surfaces can reduce the tendency for galling. MALSI ENGIN EERING,INC.
.u u .. c[ :. . a. * ... . .
Document No.1566C Apptadix C August 2,1988 Page 31 of 57 C.2.2. StopValve ContactStmas The contact on the seating surface of the stop valve is calculated using d:mensionb obtained during inspection, j Gall , t A L
~
$YG ,
r s V32 FIGURE C.6 Contact Stress, SC Sc = F / Ac F = Applied Force at the stem
= 1.8 x (SR x L) fR = Spring rate for two springs
= 800 lbs/in L = Spring Stretch
= 2.0 inches e F = 1.8 x (800 x 2.0; ,
= 2,8S0lbs AC = Contact area 7
= n x 4.12S a 0.030
= 0.389 in2 l l SC = 2,880 / 0.389 !
= 7,404 psi This contact stress is even greater than that found for the pilot valve and should, therefore, also gall. The contact width developed between the stop valve and stop seat is narrower than in the pilot valve because the contact betwe3n the surfaces is line to line due to the spherically shaped seat, The same procedures used to prevent galling in the pilot valve can be used for the stop va?ve. f l '
l M ALSI ENGIN E'E RING, INC. .
... - ..=.. _ ...
l
Document No.1366C Appendix C August 2,1988 Page 32 of 57 C.2.3. Stem Galling Two different and unrelated Lt.Ils were found on the stem. Galling was found on the stem at locations corresponding to the penetration through the stop valve and at the penetration through the bonnet bushing as shown on Figure C 7. G.11ing at the interface between the stem and the stop valve is attributed to detached material froni the galled rurface of the pilot valve getting lodged in the small clearance between the stem and the stop valve. The lodged material then initiated galling when the valve was schsequently stroked. Galling at the interface between the stem and the bonnet is attributed to side loading on the stem. Tne gall marks were found on the underside of the stem and could have been caused by the weight of the lever assembly bearing on the the stem er by misalignment of the lever arm i as sembly. Nitrehardening the stem can enhance the resistance of the stem to these types of galls. fG s)/ #
/So ll
/ _ _ _ _ _ _ _.
\ _
i j FIGURE C.7 ) i i KALSI ENGINEERING,INC. v u w. . : . . w i. .s.. 3.
Document No.15060 Appendix D August 2,1988 Page 33 of 57 l l l l [ 1 l l l APPENDIX D IWPASS AND PILOT VALVE FLOW RESISTANCE.c l KALSI ENGINEERING. INC. wi t .w., ci e r,s e. .sa. ss
Document No.1566C Appendix D August 2,1988 Page 34 of 57 APPENDIX D BYPASS AND PILOT VALVE FLOW RESISTANCES Dol, INTRODUCTION This appendix documents tho analysis performed to evalue.t4 the effect ofincreasing the bypass now area between the inlet port and the tcp of the plug. Increaciag this Oow area will create a positive pressure between the body cavity and the outlet port, which will assist the closing action of the f. top valve ur. der flowir.g conditions. Under no flow conditions, when the inlet and outlet ports are at the same pressure, this increase in now area will not provide any assistance and should not be counted on when sizing the minimum required spring force to close the valve The increase in bypass now area was recommended by Westinghouse as a positive assist in the closing action after the stop valves failed to close on demand. The flow area was increased from one hole to eight holes na shown in Figure No. D.1. Quantification of the benent of increasing the now area was not available from Westinghouse and was recommended as a modification based on positive results ebtained en similar problem valves at ancther po ver plant. The valve at that plant had etee sive stem leakage which did not permit the top cf the valve plug to be pressurized. Increasing the bypass from one hole to eignt holcs represents a gain in flow area of approximately eight times, but a reduction of about 100 times in Oow resistance. Although this modiflcation represents a substantial reduction in flow resistance, its benefit can only be quantined by comparing it to the flow resistance from the top of the plug to the outlet port. This comparison shows that before increasing the flow area the pressure drop from the inlet port through the bypass holes represented 83 percent of the total pre:sure drop from the inlet port through the oilot valve to the outlet port. After increasing the area the pressure drop scross the bypass holes now represent only S r,ercent of the total. For this new condition the pressure at the P er the plug can be expected to be almost eq aal to the inlet pressure, whereas before, the pressure at the top of the plug was elmost equal to the pressure in the outlet port. This increase in pressure above the plug should aid in the closing action under flowing conditions and there exists a pressure drop across the vr.lve. KALSI ENGINEERING,INC. Uu-.s a cms .s.. .n
Document No.1566C Appendix D August 2,1968 Page 35 cf 57 i
- a f!a K,si'% i .
l Qu e 800 / " N // 3
'f I
\ 1%
2 I c ' s<~
, i 1
., e i . . _ - -
~ .s N;n~, a.
o e a w h j l '.J i ii! ? i h % :P i !! L -C.. ' N ' _ Y. . - ' T R R 9",rW[fql1-- -
. ,p- __. M-r '
/ //, B '
-f -' ~t W -/' / / j!'Y.i?l D.;V /. /Y/.. .?!a
.Is(////A-L -
' / [<a 4r3
~- 'g '
, N/ , iM Q !
e /// N %W 'N I
////N2. w7 I n !' $b . 'G a .,N 'h ' %l (~~
t_ l 41-ee<% l
~
ln13 i ,s s'
'N.
b'. 1w-; (m>) - "l - 't. s, .
// xx b
]o~.g \wf er.
- N -ns FIGURE D.1 VALVE PT. ESSURE CHAMBER GEOMEDW bkw .
=
=
l [ l KALSI ENGINEERING. INC. u -.s a. ei , s . . s.. . s s
Document No.1566C Appendix D August 2,1968 Page 36 of 57 The flow resistance can be schematically represented as shown in Figure No. D.2.
% OrEn
& Y* W er, K Kv ' k ,is e
%U va w i
.Trv't t Ky , ,, ;
Vvvv ov4/et i FIGURE D.2 ~ VALVE OPEN HYDRAULIC CIRCUIT (See Figure D.1 for location of Resistances) Where the resistances are defined as 1:N = Needle Valve Passage ; KB = Plug / Bushing Clearance K}t = Radially Drilled Holes in Cover > KS = SterdBushing Clearance i Ky = Axially Holes in Valve Plug KA = Pilot Valve Kp = Pilot Valve Seat Port ; I KALSI ENGINEERING,INC. vec-.sc..: se.sc..s.. ss
Document No.15660 Appendix D August 2,1988 Page 37 cf 97 The now travels from the inlet port to the body cavity, then from the body cavity to the outki pu t. Some of the now also exits through the stem and out into the leak ofTfrom the body cavity. Since the aim of this analysis is to determine the relative amount of pressure that can exist in the body cavity and not the actual flow through the bypass, the reslitance through the valve proper will n9t be calculated. The equations used to calculate the f.ow resistances are from Crane Teer,nt cal Paper No. 4101 . D.2. FLOW RESISTANCES The resistances through the various paths will be calculated separately and then summed to produce the total resistances. These reaistances will be nondimensionalized (fi/D), but can be used to provide flow data by multiplying by the proper fluid constants. All resistances assume full turbulence and are equated to the velocity through a 0.25 diameter passage. D.2.1. Needle Valve Passage The needle valve, as described by Westinghouse,is provided *to admit steam to the chamber in back of the valve and into the valve itself. This relieves the pressure unbalance between the chamber in back of the valve and the surrounding chamber. Balancing these pressures prevents sluggish valae operation." This bypass, however, is so restrictive that it provides virtually no assistance or stability during the closing cycle and can only be considered useful during initial valve opening. To be efTeetive as a closing assist, the port should be larger and have at least as much flow capacity as the pilot valve; otherwise pressure can ,iot build up behind the stop valve. The resistance though the needle valve passage is calculated using the dimensions shown on Figure D.1. The total resistance is comprised of a straight run, two 90 degree turns, an inlet, and an outlet. 1 Flow of Fluids thrt. ugh Valves, Fittings, and Pipe, Technical Paper No. 410, Crane Company, NY. I mal Si E NGIN EERING. INC. m -..m .. n . a s e .s . ,
Document No.1566C Appendir. D August 2,1988 Page 38 of $7 Kg = Resia,tance through the drilled holes
= f x L /Di i i Li = Passage length
= 4.25 + 1.25 + 1.5
= 7.0 in D2 = Passage Diameter
= 0.25 (use the larger diameter) f = fully turbulent friction factor for 0.25 diameter
= 0 034 Kg = 0.034 x 7.0 / 0.25
= 0.95 K2 = Resistance through the 90' turns
= 60 x ft (Ref1)
= 60 x 0.034
= 2.04 for one turn
= 4.05 for two f urns K3 = Entry Resi .tance
= 0.5 K4 = Exi? Resistance
= 1.0 Therefore the total resistance through the needle valve passage is:
Kg K+K+K+K4 1 2 3 a 0.95 + 4.08 + 0.5 + 1.0
= 6.53 E).202. How Resistance Betmen the Stop Valw and Bushing Flow to the body cav;ty from the inlet port can also travel through the narrow clearance between the stop valve and its bushing as shown on Mgure D.3.
KAL51 ENGINEERING. INC. w ; m:... ee .:.u au. s .
Document No.1566C Appendix D August 2,1988 Page 39 of 57
.e a
/Q
/s s
/s ^ erop V4ws i
/ 8.b a ds Ruthin)v,s s v
\ ,
0 005 -o- ~
= 4.2 5 0 /a.
TIGURE D.3 STOP VALVE AND BUSIUNG CLFARANCE l KB = Flow resistance through clear in- + Intry Res. + Exit Res.
=-
f, x L, *
- D2 Where L2 = Clearance Length
= 3.3 in.
D2 = Equivalent Diameter
= radial clearance for very narrow annulus
= 0.005 f2 = 0,025 for a relative roughness of 0.003 (e'D)
, _0025 x 3 5 + 05 + 1 0.005 '
= 19.0 Equating to the velocity 2through a 0.25 diameter passage yields KB'
- KB x 0.25 4/ D2'4 Where D2's Equivalent flow diameter
= (Plug Diameter x Radial Clearance x 4)o.5
= (4.25 x 0.005 x 4)o3
= 0.292 in KB' = 19.0 x 0.25 4/ 0.2924
= 10.
KALSI ENGIN E Eft lNG. INC. ,, vi t s c.. cs sos s aw . . I
Document No.1566C Appendix D Anr.tst 2,1988 Page 40 of 57 D.2.3. Flow Resistiince Through Radially Drilled Holes In Cover The holes were added as shown on Figure D.A. to increast the flow into the body cavity. The total capseity of the eight holes is determined as follows j C N t.r* N Drlll- y - ~ l 8 Ncles -,R'.Q.- ?/9 ', 0 Ev.sM$ k// A FIGURE D.4 PADIALLY DRILIJE HOLES IN COVER K }{ ' = 1/N2 x g3; Wh e re, N = number of holes
=8 i
KH = resistance of each hole f3 xL 8
= + Entry Resittance + Exit Resistance D,
L3 se 0.75 in. D3 = 0.25 f3 = 0.034 for fully turbulent g H , 0 034 x 0.75 +05+1 0.025
= 1.507 K }{' = 1/(8 2):1.507
= 0.024 Since these holes are already 0.25 inches diameter, no equivalence is required.
KALSI ENGINEERING, INC. vi:-*s x c..oss. w ..+
Document No.1566C Appendix D August 2,1988 Page 41 of 57 D.2.4. How Resistance Through Stemilushing Cleasuee Pow through the clearance between the stem and the bushing is exhaustd into the leak.off piping system. The Dow capacity is added in parallel to the now through the pilot valve.
;y- l Leokoff yi N c=kM c.t:3 f / '{ -
7 FTCURE D.5 STDiBUSHING CLEARLNCE i Ks = Flow resistance through clearance + Entry Res. + Exit Res. f,xL,
= +05+1 Da i
Wh ere L5 = Clearance Length
= 1.25 in.
3 D5 = Equivalent Diameter i = radial clearance for very narrow annulus
= 0.003 f5 = 0.03 for a relative roughness of 0.005 (e.@
= 003 v Q + 05 + 1 .
0.00P !
= 14.0 l l
F KA1.51 E NGIN E E RING, IN C. w c -as :.. : . . s., i .s. . . .
Document No.1566C Appendix D August 2,1988 Page 42 of 57 Equating to the velocity 2through a 0.25 diameter passage yields Ks' = Ks x 0.25 4/ D 2'4 Where D2' = Equivalent flow diameter
= (Stem Diameter x Radial Clearance > 4)0 5
= (1.25 x 0.003 x 4)0.5
= 0.113 in Ks' = 14.0 x 0.15 /4 0.1134
= 335 D42.5. Flow Resistance'Ducugh Arial Eoles in Stop Valve The stop valvi has thret axial N leso which allow communication between the body cavity and the pilot valve as shovm on Figun D.6.
Fj b se.. 3 holes.
} )l \
2n A Ni% N\
~
\
\ i
\ l ,
Uf? ~ {/// FIGURE D.6 AXIAL HOLES IN STOP VAL 5T 11UG Ky* = 1/ N 2 m gy MAL 51 ENGIN EEMING. INC. "
- - u ..:. 2.a..s... s
l Document N .15660 ! Appendix D August 2,1988 Page 43 of 57 Wh ere, N = number of holes
=3 Ky = resistance of each hole f,xL,
= * ' " ' ""**** " ""*'
D6 Lc = 2.375 in. Os = 0.375 f6 = 0.03 for fully turl ulent K .\ = _0 03 x 2375 + 0.5 + 1 0375
= 1.69 Kv' = 1/3 x? 1.69
= 0.188 Equating to the velocity 2through a 0.25 diameter passage yields Kv" = Ky' x 0.25 4/ D 74 Where D7 = Equivalent flow diameter for the combined holes
= (D 26 x 3)o.5
= 0.650 Ky" = 0.188 x 0.25 /4 0.6504
= 0.004 D.2.G. Flow Resistance Through Pilot Valve Flow resistanes through the pilot valve v.ill be calculated for the condition where the pilot valve is completely open (reference Figure No. 7) and the pressure in the body cavity is urging it closed. This condition will allow the least pressure buildup in the body envity thus the minimum assist in closing the valve. The total pressure drop through the pilot valve is evmpr sed of that through the annulus defined by the valve and seat plus that through the pilot valve seat port. ,
KALEI ENGINPEERING,INC. vi c =4% ;a. s s LN E a %A.
- S S
F. Docunient No.1566C Appendix D August 2,1988 Page 44 of 57
% P,*/et P/yf FIOW '!B
\;
St.s t- 4 y y g., r s
,//h '!
FIGURE D.7 PILOTVALVE ANNL1US Since the now path is relatively short in comparison to the diameter, the resistance will be calculated using orifice equations. KA = Flow resistante throu.qh pilot annulus
= (1, 2)/(C2 x p4)
Where B and C are dependent on the ratio cf the port diameter to the equivalent flow diameter
= DE /Ds DE = Equivalent now diameter
= (D Ax H x 4)0.5 DA = 1.25 in.
H = 0.125 x sin (45')
= 0.088 in.
= (1.25 x 0.0SS x 4)0.5
= 0.663 in Ds = Seat diameter
= 1.25
= 0.663 /1.25
= 0.53 C = 0.63 for = 0.53 KA = (1 0.53 2)/(0.63 2x 0.534 )
= 22.96 KALSI ENGINEERING,INC.
s., :-m.. c e e as a. *s..m
\ Document No.1566C Appendix D August 21988 3 page 45 of 57 Equating to the velocity 2through a 0.25 diameter passage yields KA' = K xA 0.25 /4 Dg4 Where Dr = Equivalent flow diameter
= 0.663 K A' = 22.96 x 0.254 / 0.6634
= 0.464 To the resistance at the pilot valve is added the resistance due to the downstream portion of the pilot seat.
Kp = Resistance of the seat port diameter fp xL p^
= * " """
DP Lp = 1.5 in. Dp = 1.25 fp = 0.022 for fully turbulent g P , 0022 x 15 , 3 1.25
= 1.026 Equating to the velocity 2through a 0.25 diameter passage yields Kp' = Kp x 0.25 4/ Dp4 Where Dp = Seat port diameter
= 1.25 K p' = 1.026 x 0.254 /1.254
= 0.002 The total resistance across the pilot valve then becomes Epilos = KA'+ Kp'
= 0.464 + 0.002
= 0.466 KALB1 ENGINEERING. INC.
vi c - s c., en ss 6 *s*. .e m
Document No.1566C Appendix D August 2,1988 Page 46 of 57 l i D.2.7. Total Flow Resistances Having solved for all the component resistances in the now circuit, the total resistance between the inlet port and the body cavity and between the body cavity and the outlet port can now be determined using the schematic illustrated in Figure D.S. K' [y A33r D,-ain (,. F3 l< *' res / fF4 fn/e t: IC*I K' 4 0.079 p.004
?%cagjeg C. W A FIGURE D.8 HYDRAULIC SCHDIATIC OF FLOW BYPASS CIRCUIT Fer parallel resistances the genent equation can be written 2:
i '3' ]' A E* ~ r 1 > g, r p,y g,> g, y, 1* y
- y+...+ 7n
. < 's e < 7's s s y' s .
In this system, the Ks have been corrr ted to the reference diameter of1/4. inch, so that Ao = At o AN. Therefore, the equation simplines to
, 2 K=1 K, K, K,+...
1+ ge + g+ g4 . 2 Quick, Easy Way to Fsnd Flou Resistance in Hydraulie Circuits, Machine Design, pp 99 102, August 20,1981. MALBI ENGIN EERING, INC. v u -.s s .. . . ,s e. s. . . . .
Document No.1566C Appendix D August 2,1968 Page 47 of 57 l I For the circuit from the inlet port to the top of the plug I
. .2 KI= K, '
l 1+ [g + [g + [g + . . . )
. e s . .
Substituting the appropriate K variables yields the total resistance to be:
, .1 1*
~
N x ,, x, , 1+ p+ 7x
. s s ,
, for the case without the eight radially drilled holes
.t I
K, = 6. 5 3 -- 17=, 7
. N i .
K, = 2.015 and, for the case with the eight radially drilled holes
. 2 K i =&53 '
/se 1+g/sei3 2
- V 0c24 .
K, = 0. 0 2 The comparison between these two resistances shows that the pressure drop shculd be approximately 100 times less with the eight holee than with no holes. Calculating in the same way for the circuit from the top of the plug to the outlet port and stem leak off.
. .2 i" N x, s
g (# va*# ue) , MALBl E NGIN E Ebet NG. INC. vi:-.s a es r 2 . s. * ... m
- Y Docum:nt No 1566C Appendix D August 2,1fC] Page C of 57
, .2 i K O =335 ' '
/ 33s
, 1+ V(om+c usi ,
No = 0.4 2 A comparison of the resistance from the inlet port to the top of the plug against the resistance from the top of the plug to the outlet port and the stem leak.off shows that a more favorable condition now exists for pressure to build up in the body As a ratio of the total drop through the bypass and pilot valve, the pressure drop through the bypass has changed from 2.01 5 Old Ratio = 2.015 + 0.42
= 83%
to 0.02 + 0.42
= 5%
This change implies that the pressure on top of the pilot valve will always be within 83 percent of the inlet press are. The negative effect of this increase in pressure on the top of the plug is that it will now require more force to initially open the valve because the pressure on top of the plug will be relieved very little when the pilot valve is opened. In efreet, the pilot valve no longer acts as a pilot. It is possible to select the bypass area so that a compromise can be reached to achieve a more balanced operation overall, KALSI ENGINEERING. INC. vu s .. :. . w. .s . .
- Document No.1566C Appendix E Augmt 2,1988 Page 49 of 57 P
1 i [ l t j APPENDIX E SPRING RATE AND PRELOAD TEST i l 4 j 1 KALSI ENGIN EE RING. INC. us c-* .ca. c o . 6 4%. s e i
Document No.1566C Appendix E August 2,1988 Page 50 of 57 APFENDIX E SPRING RATE AND PRELOAD TEST This appendix documents the results of tests performed to deurmine the rate and pretension of the eprings used in the Westinghouse high pressure stop valves. These tests were performed P before the spring data were available from Westingl..use and the initial root cause of the valve's failure to close was being determined. Two springs were rate and pretension tested with two other installed springs checked for initial spring stretch. The two springs tested were from Sigh pressure stop valve No.11 with the springs of valve No.13 checked for spring stretch by gaging the gap between the coils. The springs were tested for rate and pretension by loading using a portable crane, taking lead readings using a calibrated load cell, and taking dimensional measurements frorn the top coil to the bottom coil ush.g vernier calipers. The measurements taken for the two springs are tabulated below with the results graphically represented in Figure E.1: Top Spring for Higi Pressure Stop Valve No.11 Imd SpdagIangth Spring Rate O 11.175 ; 203 11.651 426 414 12.173 415 602 12.622 41 6 81 0 13.120 41 6 414 12.160 420 . 0 11.175 Average Spring Rate = 419 lbdn : Bottom Spdag ibr High Pawasurs Stop Valve No.11 Imd Spdaglangth Spdag Rate 0 11.300 204 11.765 439 406 12.238 433 608 12.710 451 804 13.170 430 , 402 12.230 432 O 11.302 Average Spring Rate = 433 lbdn , l KAL51 ENGINEERING, INC. wc-asc ce w ,s.. ...
i , i Document No.1SowO Appendix E August 2,1988 Page 51 of 57 z _ _ _ _ _ T: =._. . ._. . .~_~
~-
._ :. _ _ _ , . _ -=_.==.__==
==. .. .__
==-- . _ . . _ - -
~~
_. __.--.~"_ - . _
=&~
~
___..____.__m__- L-- __f __ _ s- .._. ,
~ ~ . . -
.y-sy- ._: _~~ 7
- - ---- =49 =__ . - _ =
l soo . f.- - _. E = =_ ;
~
?_ ._?_
i .- W_ .__S~~I~OP Q~A0T: tom.SRRINK SPNNG 4 _--- w. 40o -
-~:-f
- {
i v - JrJO_ .
- . 1 -.-._.-_. -
e _ q ._ .f _ __
; u=9m .
,. =- : -_
.._.: .: == -_.
-- m__ a -_._,< .
/_ -
_/ _ _ _ - i
-s
,.__m
__.__.p__..__
; _ _g, , i g
i __. __q- . __ - - - - -
~--=_. _. _ . .-. _. ??L W S E M ?_ & '~~~~~.~___. - ~ f:.~ T~~
2 FIGURE E.1 I i i LOAD VERSUS DEFLECmON FOR ) VALVE NO, u HIGH PRESSURE SMP VALVE SPRINGS i
- } >
l The results of the spring load test show that the springs are not pretensioned, as evidenced by I the spring curve passing through the origin, and that the average spring rate of the two springs l l 1s approximately $26 pova.ds per inch. 3 i ) MALBI ENGINEERING,INC. m .. s . c . . ~. . .. . . . 1
- _ _ _ _ _ _. . - ~ . _. _ ._ _.
Docunent No.1566C Appendix E August 2,1988 Page 52 of 57 h The calculated spring rate agrees well with the tast and is calculated as folkws: K= Gd' 8D*N G = Shear modulus
= 11,500,000 psi for carbon steel d = Wire diameter
= 0.70 in ,
Dn Mean coil diameter '
= 4.0625 N = Number of active coils
= 13 4
- x. u. .o;0.20 '
8 x 4.0625 x 13
= 395 lbvin ,
The measured gaps between the coils of the springs (stillinstalled and undisturbed en valve No.13 were approximately 0.04C inch. These measurements we e gaged using a 0.020 inch thick metal scale. Multiplying the total number of active coils by the gap between the coils yields the approximate spring stretch, assuming that the coils were closely wound during ; fabrication. Spring Stretch = Active coils x coil gap Active coils = 13 Coil gap = 0.040 average Spring Stretch = 13 x 0.040
= 0 52 inch For this spring stretch, knowin; from the spring rate test that the springs are not wound with initial pretension, the approximate installed spring preload can be calculated for valve No 13.
Spring Preload = Spring stretch x spring rate
= 0.52 x 426
= 2221bs l KALE! E NGIN E E Rl h8G, I N C.
~
ve cov:.. es w s .~ . .e
, 1 Document No.1566C Appendir.E August 2,1988 Page 53 of 57
'.or twt springs in phrallel, the total spring load becomes Total Preload = 222x2
. 444 lbs l
This load is less than that nquind to nppose the stam blowout force which is h!!nimum Required Spring Force = Stem area a pressure / lever ratio Stem Area = 1.252x x / 4 i
= 1.23 in2 Pressure = 1160 psig Lever Ratio = 1.8 hiinimum Required Spring Force = 1.23 x 1160 /1.8
= 792 pounds I
L This minimum required spring force is based only on the stem blowout force and does not j include any loss in load due to assembly frictional resistances. 1 k r ] I KALSI E NGIN E ERING. INC. ve: avci.caso a.. ,s. f
Document No.1566C Appendix F August 2,1988 Page 54 of 57 6 L i i t t I APPLNDIX F VALVE SUBCO.\DirITEE SUhDIARY ON ROOT CAUSE ANALYSIS KALBI ENGINEERING,INC. vi:-*w.. os sos a .sa. e
Doerment No.1566C Appendix F August 2,1988 Page 55 of 57 VAUrt JUSC9911TTtt B M ERY OR ROOT CAUSE AAALYS15 MitM PREsamt STOP VALVES
- 1. The -alve did not have sufficient spring preload in the as found condition to ensure full closure.
- 2. If soring ' load pea original design is passent, the valve has sufficiem .rce to close.
- 3. stes ginding gallin and of the linkagefriction; increased due to corrosion resulted which fureher in side redated thelead,lable avai closin force.
Action Itama for W Stop Valves
- 1. Prior to Operetten o Develop vritten procedates and adjust linkages to obtain p*oper spring preload.
e Inspect, and replace or rework st es as retvired. o Insrect and cles.i linkages.
- 2. Lone Tore neccayendations e Enlarge the stop valve cover bypass flow area to increase eargin for degra6ed conditions.
o Ritride stem mate ial to impeove resistance to galling. e Incorporate spacers in linkage to miniette binding caaled by corrosion. e Verify that there is no steam leakage from the sten which accelerates corrosion of the linkage. e Replace springs to increase closing force meet ni under degraded conditions. ( 4 ti _
/
b hY JJK782 . I KAL51 E NOIN ZERING, INC. es c ,e..oss. . I
~ _ _
Document No.1666C Appendix F August 2.1968 Page 55 of 59 VALVE SJecest!TTtt SMWtf 0111100T CAu$t AAALT$15 LtM Pitt$$ Ult! STOP VALVES
- 1. The valve detien was fewed to have sufficient closing forces.
- 2. Field inspection revealed the valve to be in good condition.
- 3. Loose actuator mounting bolts, en found efter the incident, do not prevent valve closure.
Action Items for LP Step valves . 1. Prior to eseestion a home.
- 2. Lone fore Recomp4etiens !
o hone.
)
I i l l i ! i I I i I J 1 (.1 s t _ [M ;8 N N M 782 3 [ ; i L Y \ t KALSI ENGIN E E RING. INC. I l ve = asw ce sos .sa.. . I
~ . _ _ _ ._ r
Document No.1566C Appendix F August 2,1988 Page 57 of 57 I l t K AL.51 E N GIN E ERIN G, IN C. soc mo [ on 4=a. .oi.sce w .=,i arv ho Phlt 98 C.*vt. peCJt:T. Pa4Pa*t0 tv. M?t. P80J =0.: ptg M'1 _a - 40 0, to .- i F S$ $ r I eg[d see d k I. *
- 6=.
AL.: yaw <L l v % .j , - ..t : .rp.l~ % ) x '-
. wau \
J.a L s-pz J pra .L L.. y t.: ya J L L L:. f _ r f
$f~' })w* & $-.n--m %)w _ i J. % h p a- ,
l e WA 2, & Sn hwr: . kfm ( f&J e ! r I t I h i I l s-e-se c4 yay l
. - . - . - - - l f
4 KALSI E NGIN EE RING. IN C. l i f
e 1 1 APPENDIX D 1 l TURRINE SHATT RACMr.NT MISSILE ANALYS!$ 1 1 I 1 l l l l l l L
TURBINE MISSIL.E STVDY FROM MAY 25, 1988 FAI1ERE
- 1. Background Three Feedvater turbine pump assemblies have been installed in a line in a north south direction on the roof (El. 83' 0) of the Turbine Cenerator Building. During the test operation on May 25, 1988, the turbine pump assembly 11. on the north side, experienced Turbine damage. The Turbine shaft broke and a missile weighing 215 pounds was generated. This shaft segment struck the Turbine case (see Photograph 1). The missile then struck and damaged the steel T Section members of the turbine enclosure (see Photographs 2 4), was deflected and hit the pump discharge pipe (see Photographs 5 8), then fell to ground level hitting a well point protective rebar cage (Photographs 14 19), the secondary makeup tank (Photograph 18) and finally came to rest on the ground (Photographs 24 30). It is possible that another small 4*x4"x1/2" triangular missile existed which case out from the broken hollow coupling spacer. Figures 1 and 2 show the missile paths and associated equipment orientation. This study concentrates on the large shaft generated missile which emerged from the turbine enclosure.
There were two official test observers on the Turbine Deck during the Turbine missile accident. One witness observed that the missile emerged from the top of the pipe and flev cleanly to the ground without impact with the Turbine deck. The ather witness observed that the missile struck the concrete floor before it reached the ground. As neither witness's attention was solely concentrated on the missile but was also occupied with continued observation of the Turbine, it is the objective of this study to determine the shaf t missile trajectory. The rotational speed at the time the missile was generated from the Turbine is also calculated based on tapact damage observed.
- 3. Missile Trajectory From the evidence collected from the accident site, it is believed that the shaft missile hit the concrete floor before it reached the ground.
There are three points on the cancrete floor which were damaged by the 53/TCANA/a MS.1
l I tissile (Phot:grcphs 9 11). Th:y cre skotch:d with tha dimensiena co shown in Figure 3. Point 7 is a five inch diameter hole. This hole is believed to haye been created by the 6" diametes missile end. This hole I measured 3 1/4 inches deep and exposed a rebar which runs in a north south j i j direction, The concrete and aggregate in the hole was powdered by the ' missile impact. l ! Point 7A is an elliptical cavity with a long uxis of 5 inches and a short axis of 2 1/2 inches. The maximum depth is about 1/2 inch. This point is believed to have been hit by the 11. inch diameter flange on the missile. Point 78 is a small cavity in elliptical shape with a long axis of to 2 l l inches and a short axis of 1 inch. The depth is about 1/4 inch. This point is believed to have been hit during the missile bounce. I l The configuration of these points fits very well with the missile dimensions as seen in Figure 5. The; center to center distance between point 7 and point 7A is 17.50 inches which is essentially identical to the dimension from the missile 6 inch diameter end to the near end of the 11 inch flange shown in Figure 4 The missile orientation at impact is shown by Figure 5. Furthermors, the second small missile which weighed less than 5 pounds and was triangular in shape (4 inches on each side) is unlikely to have caused the hole at point 7 since the required velocity to cause the hole at point 7 would not be credible. In addition this small I missile could not have damaged the concrete floor at three points, because l it would have "skipped" away after the first impact. l A review of the available information suggests that the seqvence of events in regards to the turbine shaft missile was as follows: 1
- s. The missile was generated and ejected from the turbine sheft end l near the turbine pump coupling. The center line of the turbine rotor 1
is used as the missile source a.J as the origin of the coordinate system as shown in Figures 6 through 9, 1 l $3/TCANA/a MS.2
k
- b. The cissilo struck a secol T secticn acabor of tho turbins enclosure with an initial linear velocity of approximately 350 feet /second.
From the deformed T member, it can be seen that the rim of the ] missile hit the member. (see Photographs 2 and 4) l t
- c. The T section member was damaged and deflected the missile toward the i j
horizontal direction with a horizontal velocity of approximately 300 i p ]; feet /second. j
- d. The missile deflected by the T section seaber was in a vertical l position with the 11 inch flange on the top and the 6. inch shaft at the lower end. This lower missile end hit the pump discharge elbow l and the missile tipped over the pipe toward the turbine deck. j i (Photographs 5 8) r i k 4 i
- e. The utssile, with a total linear velocity of approximately 300 l l feet /second (45 feet /second vertically down). then struck the (
I concrete floor at an angle of approximatvly 9 degrees from the l t horizontal. (Photograph S) { 1
- f. The missile was then deflected by the concrete floor at point (
t i 7 and the missile flipped in orientation and caused point 78. The l ) missile then bounced between a 42* high handrail, which surrounds the i turbine roof, and a steel beam about 20 feet above the floor, which I i forms part of the deaerator support structure. As the missile parsed i ) over the hand rail it clipped a deenergized welding lead from some f
- ongoing maintenance work on the deserator structure. The cable was [
j partially severed by the rotational spin of the missile. (see Photographs 10 and 12) f ]' } I
\ }
- g. The missile passed over the handrail and hit a protective rebar cage surrounding well points on the ground. The cage was damaged when the l sissile crashed through the upper west side cage bar end the lower
] north side case bar. (Photographs 14 16) A hole approximately 3" l deep in the gravel covering the ground was caused by the missile upon i I i l S3/TCANA/a MS.3 i }
imp:ct (Ph:tsgrcph 17). Thoso icpect p31nts cro sh:vn in Figuro 1 cs points 2, 3, & 4 respectively. After hita:ing the ground the missile l was deflected into the secondary makeup water storage tank I (Photograph 18). A dent and cut from the spinning action of the flange was caused (Point 5). After the impact with the storage tank the missile bounced off the tank and hit the ground a second time (Photograph 22) near the southeast leg of the rebar cage (Point 6). The missile then came to rest af ter the spinning energy was dissipated in cutting a curved trench in the gravel and digging itself into the ground. The hole caused by the missile spin was on ' approximately 4" deep. It is labeled as point 1 (Photographs 25 30). The direction of the turbine spin was estabitshed by the digging action and packing of dirt against the coupling bolts. This eliminated the potential of turbine reverse rotation driven by pump , reverse flow possibilities. This eliminated the potential of turbine ! reverse rotation driven by pump reverse flow possibilities. This characteristic can be seen in Photographs 28 through 30. l l 3. Evaluation of Turbine Speed at the Time of Missile _Ceneration l 1 l l Missile impact is a complicated phenomenon. It cannot be described ' completely by existing theories. Even the measuremerts of parameters ' for '.he misrile strikes in the field are difficult. Therefore, the
- results of this study are approximata.
The missile velocity was changed in magnitude and direction during the course of its impact at the steel T Section seaber of the turbine ! enclosure end, later, tripping at the top of the pump discharge pipe. ! The slight impact of the shaft missile with the turbine casing is ignored. The missile velocity on striking the turbine deck was a determined by first determining the normal velocity component required to create the damage to the concrete floor using formulas for concrete i scabbing and perforation thickness.* l l
- Chang, V.S., ' Impact of Solid Missiles on Concrete Barriers," Journal of Structural Division, ASCF., February 1981.
1 S3/TCANA/a MS.4 l l
The norma' velocity comp:nsnt v:s then divided by tha sine cf the cnglo between the strike itne and the horizontal line to obtain the strike velocity. This yielded a missile velocity of 300 feet /second. The missile initial velocity van correlated to the strike v*locity based on the principle of conservation of linear monentum, and yielded a value of 350 feet /second. The determination of the turbine rotational speed at the time of the missile acciden': is based on equating the initial kinetic energy of the missile to the t ranslational kinetic energy of the missile rotating in a circle of radius equal to the missile eccentrit.ity relative to the straight Turbine shaft. An average valve of 15,200 rpm results. 4 Approximations There are several uncertainties in the parameters and assu:rptions utilized in this study:
- a. The formulas for scabbing and perforation thicknesses of concrete have been derived for determining the concrete thickness which is conservative for a known missile velocity (see Footnote on previous page). The formulas are for r inforced concrete panels. The Turbine deck at El. 83' 0 is a concrete floor with steel deck at the bottom side. The missile striking velocity is estimated from the calculated values, based on the6e formulas.
- b. The missile was deflected at the steel T section member of the Turbine enclosure and was tripped at the top of the discharge pipe of the pump. The changes in velocity magnitude and direction were calculated from a simplified model and assumptions.
$3/TGANA/a MS.5 ,
- c. The translational kinetic energies before and after the missile was generated ate assumed to be equal. The actual mechar. ism of the break of the Turbine shaf t is not reflected in this analysis, j It is observed that the missile ejection velocity is not on e.5e plane normal to the straight shaf t which is a reflection of effects of the shaft actual failure mechanism. l
- d. The accurate measurements of some parameters for this study from the field are difficult. The most difficult and important one is the Turbine shaf t bend angle for determining the eccentricity of the missile. The angle of the shaft prior to separation was estimated to be from 10 to 15 degrees based on examination of the turbine shaf t missile and the visually observable slight bend at the point of failure (See Photographs 26). l
- 5. Conclusions i
! Among the above uncertainties of the parameters, the most sensitive and l critical parameter is the turbine shaf t bend angle of Item 4 Assuming 10 degrees and 15 degrees with other parameters remaining the same, results of a turbine rotational speed of 19.000 and 12,700 rpm respectively are l obtained. The average speed is equal to 15., ' 9m corresponding to 12.5 ! degrees prior to ejection. ! l l t 1 I 5 i S3/TCANA/a MS.6 ' l i t
I i t 't* '
, -s. .
J . . . , 7 f w , . n '
.y , , , '
'; (
.s ,
4 + a
. 4 ,. :g
. t
.T' 4,- ..
',]e " .
l
. N l
' A+ *'..- ,
I I '
/j- * <
i '
' '. .% sT
- ' g' ,
z
-3 *
. . .uar . - ' x .- . - - _ _. z._ _. )
photo 1. The damage of the shaft missile impact on the turbine case,
. .e 6 .
gb
~
I phete s l The damaged T section on the enclosure of the failed tuibine, as seen f rom the cast door looking northwest. (Point T of test). i
l i l M } } 7 , ,
~4 s 1. .3 j y ' <? + 5 i ;* .
fi\ . i\ -
- '\ , ,
\
iw T ,
' t
{ ' 6' photo 3 . l The damaged *T* sectian on the enclosure of the failed turbine, as seen from the outside lookirg southeast. (Point T of test)
, , e' I
/
i rhoto 4 l The damaged 'Isection on the enclosure of the f ailed turbine, as seen from the outside looking southw est. (Point T of test) W d
I l \ l ) . ~ . _ ~ 7 - - ,-
?!
i l N . -2 c' i ( 6 h\ l
. , .s ;
, 'd j, s ;
!' s l
petos. ; I The turbine enclosure incurred a bent 'T'section and the insulation on the pump discharge pipe was damaged by the turbine shaf t missile.
, ~ . .
t
,, ,,rt
_~
', e IN l; .5 i ' i,* C y
e _ 7
*5 ,
j { I
\
'Il i
=~ ~
1 j p eio a ! The turbine enclosure and discharge pipe as viewed along the missile path. l l ! l -. {
--- . ~ . . - -- -
, =
g" % _ _. E l k;. c.; " bA ,
~~
,n -
,=q 1
'a
,3 .-
g , g .e lf ir I,' . N '< s .g l
- k. l a %; kY
{ photo F. 1 A pump discharge 90' cibow sustained damage from the turbine missile. Point P of text (view looking north). s
?
f* * ( ~,
.- [,' gg' ..
- ',
- Q'$
[, . p . , ,
- N- A '-
e
\
*2. ,
a PI s , . .. M' l
. .( ,
l t .s . ; pheto 8 Pipe impact damage close up (view look!ng north).
l j l 4 ,i , . . . , - - -, . - ,-.. .. ~ , i > (, . . . . l t .s l
'* %S* .
e , r'<e ,r , I g' ( y' q - j 9 j;
- i W , ) . \
l ! % _
*[f, , . , j f,'# T. (
J I , i . g [ ;. yr.e....,i i
. (,. . j e - f', , n '. l'
_..Y
. T,W: " .w L
\tf ,f;fi- .,e,% t- ~-
^
7 i y s' i 4, - .r, P t,- u: /,
. q- ' "" ' g% ' ' - ga ,
g.
,'o,(;.~.s
" +
/, .g, ,
, , N /, ,
. l ,. y. . ,
+; - -
l , ,
, ;s*
% 3' ' # i . rN g . . , , , , . , , 7 A ',
* . 1/ *
' , %_. *.. ?.,*%c s- >
, ,j ,
e, u-. ,.? , a
~.. -
3, .s 'g .. w T
. .,/
i 7 l
)
, , . _e . .N,
. > ,1 , P g, . , g a
.s -
3
, e.(
', g
- ^;
) J
..,i >.
.- i[t.* f l/s g -. t 7,
. A_. ('* . *k g ( , + * . -. . d. e is i-t .. .. h . 'v . :_ _ _ . m i ;h-t, ^4
( ('ime up of shaf t minile impact with the turbine dcek. Visible are points '7' and '
.s ' l i
j l , rp, y -/ ES '
~ =-
'~
N M. .p,. '
=
' ) l.
) . __ ) p . 1
~. ,
g s,4
- h '
, t, i ..
+
h .
+
a . 1 .-.4 g
, e. M, S ,
~
i, j
;.QW s %A ,1
'N- -
.,4 , ..
_.,9_.."_'...^_'t> -
;>t n View hd ing north. Point '7','M' and w elding nuchine are show n Io the right f t he a c1Je r n t he t.p ..I~ t% scs ond it s m i k e u p w a t e r t a rd
~
i i s ,, y ; r(,s "'-
* '<.=.s.:7 1' J '
~
. i k '.
~
~
.;: ,. 9 .
. s' ,
T '- ..' f ,' 9[.- ; .. ' If *. 3 , 1
. a /5
" '~ '
l .. f. - y {
. . $1 1 . ,.,
1
.t < . -s - . .
[
, .n ,
[ ..
< l
, , .. 9-photo 9.
Closc-up of shaft missile impact with the turbinc deck, Visible are points'7' and j '7 A'. l i 1
*) E I
M*#. JRt*MM .L . e g, ' i i n .. , .- -**.I &, i ( .YM 46 ' * ' , -- . f V , . . . , , . <64 ,,89 - . .g.n . sm .-- . t i
.py yp.H.- >
i J.$ M E
/
e,_
... ~
s
,l
'- Q. *
.s g-u ,ve -+ .: :
c l
.. a , , .
,', . , 4, , "? ,_,'.,.3.u -% . .
. pg .
. i
> . a p, - . - ~, .- ,
'* s '& * '.. q.% ', ,.
s e . .A l-,"', fe .. > & , 3 - - %, g4 :. . g s, - a' *~',, ,. r.
*~
?ge
,, . , Mk' - C
. s g m q,y,
- . w : .y .
W l (s
.c 'f
'...c,
, s' K. ' . 9
,e
. . ., .8 . . "
- ag % . ,. ,
, . -. e r-
~
Y .n p: .a ,,. yt * - e; n ,
. s f. ,- ,' . - - \ '
i s i
- 4. ; . ,.- 4,.- , , ,, y '
r i' L 2'- ' ~ I
' .[,...:.. .r: 5 c%~ _j photo 10. ;
View looking north. Point '7','7A' and weldirs machine are shown.To th; right ! of the welder is the top of the secondary make up water tank, l l
}
t
{. - , a", .. - i.. '. Ih' >_ f * i
'i
. - ' ,',h[.- - -' l
}[ !) .. . ' ' W,.$ ,, , '
., a. . .: w: . .
M Gl*%? . p :. Y ve
.'.%SYr'#g" ..:. [\
*T ^ ' -
l [,N ;y*.l,f ,
. , , U' 4 .#'
*=.
, j, , ,, .i , ~ - 3 .. .
*,e..
*,,,.y_ >
< t? ,,?> .; r- =
1,P ' . . . a%< ...
'+4.%
q ^$ e > ,. g' i * . . 4, - p h. %,}L y 4., . 4 ,' s
, :, , ,f . y . -
f: N n . .3,,f q r .- -
.. }$ ~ k fj' _f* t . ', : .
. I y . A ; ; - *py ;,y.W.
;g '
f fes,- . ,t 7 .' '., [ *. . ' ' 'g [s
- i j ~;.) , .s
- 54 -
4.L,*;;, 1y %%t ' f .' :% z.'.N v.' ,
.?;
)
l photo 11. l i Turbine deck impact points '7' and '7A'. Point FB' is not visibic due to ; powdered concrete. i r 5 m - *- em, ;- - l
,g eg ,.
g .gc j
'. f I
t j 6 g l .. l s n ami T ,
.~ . 7 t '
)
~~
4 p' m..
. . . >,t
_ s. m 3-3
. e l i
*b y_ )
===,, , !
( -.e I photo it. A welding lead was partially severed af ter contac: with the turbine rnissile.The lead was strung up to the descrator structure. View is looking east. I t
-... - _ _ _ . -. , _=. _ = , - . ___ ~ ._ . _ _ _ - - -
)
y ;. . m i { [y ,
; 01-I j -
t 1 . *% _
~
( rp s ,
,e' ' ~ . .Y. , . 's, I
,..s , .o ?. _
1
' dN '_ , 9, . ; ; ,, .
i y. , , ' ,. l f g
.. ga. :,
'7.'2 _; 7 M.[ _l' s , c,g; Y
.~ s j,rf u.
photo 13, 1. Close-up of the welding lead. The secondary make up tank can be seen in the I background. 4 , I l l - -. - ! I s I s l . i i l i l
-. g 4
g
~ t W ,
[ $
-~
g
~Q
; . ?
.N [?, '
Yh\ . . . . v; 4 '- Ci j photo 14. { View of the make up tank and rebar cage looking down along the missile path l i from the turbine deck. ! I I
I 1 l
-n r,.,;w ,
l
-W u-I l
r 9 photo 15. The rebar cage is reconstructed
/ with the broken upper and lower !
'i bars being held in place. Missile
* impact with the ground and taak 6 .:x .t is visible.
-;6q,..:; ,%W:/~ ,;
,,- r . ,c - . ,.
((. [ .,, .'q zcs1,
- k Y '
1 o .c.f ' .:w ,s . 3
, , /J 5 % ,
e
,~. T gej
'^
m a (.i .; , ;.1&*,n,,
- zM'[
- e- *,,E' ' ; 4' s
,%)D
[~ 27- ,' hU,'
$ m ; *by r.,?:*;
.i. * **-
- .y.
w., Q ,c[3g
./%. 7 ' y'
.N. -y; ' -M yl % & &.t}{rg .'.a ; e;7. ,% ,s;? , % r 4; ") 2I%'C.?,:. :
..rd - . n'- \
2 .
.' : '$. 5 [ .
photo 16, I i e i ! e2 ! w ,
< -[.'.Y u- .
* 'e M ,1 ~ ' ? , *N ~,
t '!-
, . ;g.% : f *9.' 'tl, y.,
,Q. ' C. A, p* p- .
w.1
, ,m .:: . ',. ,3g _:).,g jg -
EQ?' m, _ N' a
- .(, 's f
- .J -
- -::,.y"?.*jNf,'$
~ , ,; ;-
. f 1.*,f.,y.a, w < 4 1+ l-_;~- s .,. .. *5
- ~:
',"4:h] g 1,'
ik,*)* ;!.h ; (*k.. '
~,..n.,
j8 h,', [d [d I i i ! s
,-,--,--,--~,--.~,mm_
' ~
6)h.[$ ~~ 'f['1:.> 1llfN,f *k',.
,~ ; ' .. ,
~
* *. ~ .,
25 "q , II , 3 MM [ .#i s$'* , d # ". '-4: *
-G Nji. T.. ' cE ,(. , J k
' ~
. v ? $Pi.y gk ..
*. g . 5 l* . .; _
, c. .:. ,;.:.; . .. ,
O M .' ,M Y.)/p~ .- ((9"l ..g 1*-'.e .I ** " ,,
.e c . y t ;c. . "
^
. .. L : .
f O*
. V .ifd . 'l/
j.; ';.T. eij . (g. *.;!,n.) f
'{ ' #'
t - 2 . . , . w
- c4*, '.a' . ..[.[
=
,, -l . .., .,r' ' - .
j d*.
,, , 1 g ..
x . .
. (J ' o ,'.. ;
. . . . 1 -*
y, A $r . g- ', n.O. . g(e -
' l . . . . ... ' )
[.; ..,*...tJ e . \I - l e s . - 1 iJ -
#
- r ~
,, f 4
' o ,, * .. .
+
9 s , photo 17. Close.up of the initial ground impact (point 4). Note the crushed gravel. , l I t i [
. ,o .
f e I ) a '. ,e (
- s l
* ~
4 . l
- l 4 ,. t j
. I
- l 1
l s ! - ( l
' s ,
. wrA a . A*
photo 18. i 1 The secondary make up water storage tank with impact point 5. View along the f missile path. l r i I
. . _ . _ __ . .m __ _ __. . _ _ _m. . - . . _ __._. . . . _ . . _ _.
'l 1 I J l O 2
'.',f,.4 c l5 6 . . _' ; ?*'. t * '
sfs', sh- V , . . [, g
<& go
, .gr ,* . I* r .,- "f .,y.'
\.>s..
s
\.r ,
, :~. A -
.,.? .
l
%,, u,G .n a , .p \*, g s. , ' ~:? y,o ,y .~ty*m 7 :^ y*:y e
. * .e*^ P-\q u 'y -
.'**4, w- t ? Lj ,
_ g y .f' '" ~ w
.g'- (ms,-u O, , ,* 'q .
,- -w s'
.q;;f '.~ > ~; .&., . ., ~
.,VJ , n - E.' .j . )
^
.. .w-* *A. (
'* I s
,4
. *f ' , v ,r. ,, $p g,~:
p,M.,g. .%, .
% "J; s a. <..3\p.3,~ . . ,
e' m A G **c g g ,. 'q' %n x' ~g y pg 5
- - l w11+.f, A g. ~
x' -:.. Dyry; . .y ', G ; , et,%r.fgjf,. 4..n.~ ,e / ; J.a<v~- . ,'- ^*?eL *g% . l r:r e t. s . r I t'f..
- . ec: =.o.s f u a$ *- > s \
. e.' ' . , , . ,,s h' . .. ,k .
' '
- b , ,5 ,. .p , '~tc, .;a"). [: . , . . - , .v.,: h. .. - + b e-'.2 w'
i c, . ,.,w,2, n.%. .+m i i . .. ..,. ,
. .ww-. ; .
c s.
~..* . . .' 4 :-1 a r..* ~,s 2.,~,d; r
(: .. a. c-t*. .*
. e-
, . p_
- +.".;.. . 1.. .% . ~. 05 "v &?,? &.3. %p , R.tw Hw
. .%'s,.r.. . q, ,=,. . . .L, photo 19.
View along missile path, showing well point pipe (2' PVC) broke by a flying [ piece of rebar. The rebar .,:came airborne upon impact of the shaft missile. ! i i I
- --m ;
j - :[ , p j , , 'i
. .Y , f j # ;i1; {
i i 1 I 1 I
- ' ~
*f
'~
f
, . r.,, .s,,-; : .+W.sw ,, m
.j_,, . -i 4
- x. . .
Y ,N Q ~ -
. . y l 1
. w e . . .
photo 20. ( , The broken rails of the rebar cage arc shown in their as found positions. I ! i >
l \ g
* (
n- $<
., z . 'p'f;,. . -
'x y s, b ,O ' ~.
- A*
g *. ..h , s, . .
..9 <
q - .
. ,,i
! - . --'.. .4 ,*. photo 21.
+ '
. . . e . s %. 4 .
'h ' ., cg. *:l# . - The broken PVC well point E. *- ,' - N .*j .,,, , [
- 7,'; ,.,
(part 2A)is shown as found. y
.e
. . ... .- r v.?m'm;j.i
,,, 4 N. ' 6'. u -, ik*2 ,
- .y , ,
*.e.+ Jis*R '. ?p,x'h, . /a=%- - > ;. . _
4 l
, , , .a. f **'.". , - f.
.',? . ,Q. .. , .,y;-
. %).$:ll t, _.x,> >l .a
~].,
- 4. .
, 3 fy'. . Y Q.~.1 s
,.,a..- .w
,- 7 . p. . . . v y. :
r ,.s I
,'(.
g ' . *, ** *' 2 4 g .,f. '[ # 1[
* * '5
% ,"y [j y.
-t .-?*).. ' ~
.. . ,~ ," 4}
.,. 3,
..o.
,3 m % >,1,- ;, . . . - 4 .;
A ** % ..,- . g**'c, 7 . .' ,, g* .q. .jT ',* i, %E,W'/ t it C ; ':',s7',',s. .,,.,'. {;i,
!j f
- t. , < ; . q .;lH. ; *p, y . , . , ,-
. m. ,...,,..}
$ . N '
, , .p .. q . .. ..-w _.f. , . . , ; . ,, ~ . , ; ,
p r.- ,. 3, .
- n . . . .e3 9 y",,g. .,$. "# 4 - . /C i
. , *hsap , , ..~+2 , , . ." ,'y 7, .,;'t 1 . -
p*>.,{, v); I s.f:$ y*f)y' ,'-Y.g, ...
. .a 4 .
,' g . '
..'7 c .v
';', , ~e; 1 2~
,.c . .y -
J. x.,,f,su .. c w,. n..u.=. 6 a My .p, ., .n.. . L .q w . .g a.
- M*..a. t : a.sa
- v. o .e . < . . .
t i i T, l e$r ( 1 i
?
4 i (l t f l i photo 22. .a l t ; a i Initial ground .impact (point 4), tank y **-
- e. '1:"opC'J= ..
y ..,.. 7~ , , .s i impact (point 5), and secondary ground e S zS p m <s <r.. a ,., '*l .' ." , W* '.
+. * < 7, :r,,s>c
,.y
\. impact (point 6) as seen along the ; 4.,e ,r ,oJ.g ?. J,1;c, - I rebound missile path. e j 4 # b.. . H . S . ' > 4 " i ? f.% ( . l.; .XQ* . S M, @ki' l Y.
~
).
. . 3..lL r, ' a.s 'g '
., +7e *,,. '
WQ* 9- av,- ..2,6 ?*M',sM.* .s '4
** . . ,.n. % j , %,,, y* ._g.. b o d '.J/
l vv <: s>h# *9' f , 1mpe+ , w , v ' A* w, g *,, w i f v', ,
.r-m a', ,q._ .
w: w ..
'
- g*g' .,*. . . , j ,.c t
' {' p si; #"# y ,* h, b.,4
- l , q
*, e-6
- ' .9 1 . 'T -
l R g 34, %g I p,,4.c;,M.W, t 4 *.e r d. ' .'-y.c.M.
'%, i
,4j
,.P.*. % dd U d d '- , * ' '"we / =,. .b..W.,.
a" s *
,Y*g y h
4 lWh/dJff.9QW",W8,. p
.,c% r .
hsMN M,u &.
~.,.
~~, m, & .e.
f ') .,$s, ,.L.p . rm~-+
- v. m ,
.<.,u
\
7 * * '
.* ..:. Q. 't ** c. .l. ' . ~
i w%,' './'* $u d'g skeg:y,e%
. .. .x.. w r %c.:.u.W.
..*&,._ . :A ..
.. v s: @. *,4 .
1 v
l l l l ps., .wp ":. > .~ 6' ';u ?.s?;. ;.7 - ~:
. .. . .. . . . , f. . .,; *; *~ D : i n % +M .;;a n~. .',. r..;..
s fs
; .a
;6 . . .
.. . a.. . . - .
;f,*.: i, *. ' '$'$,. ;
\h ., -h l , _ ' , .:, ; 5
.}
>:- ,,ai .
%i
.;n.
. .;-' . ..r:; .. ; . 4 . ..- _ ' , .
9 7.;
, 'Y :' ,
'y
'\'.~. . ..-, _ 3
; : ,. p_..j
.w ,a n' ?. .'. %
.e
;n-
- ,V. ..
- ,; ' ,. :,1 ,.
l
,e nt e_
,4 e :
' N,A t
t photo 2s. [ l View of the final resting position of the shaft missile and the broken rails from ! ! the rebar cage, i i I t
.., j
, l I
' : 5*f 6f. ' '. ,+ :1 .
~,
4
#. . c
'f ,1 .
?.n c.-'.
+ ' i
. ,. .. ...:.., c. .. '
.g.'
*4 r ,. r.' l
.n ,. ..e,v ,
. v.
; , -g y., .,
- p. s. A,
- 4 ; -- . 3 ..
. , a .
i
- 4.'.z . w, ,2 s s .
- ?. .' '; .. '!4r.. ; '
.id' i
,,,',a, y- d - :..
.e- -
g w
.gsy q [$; . * ,; )f .
,,y,<._
. o, ,. ,
. *+ . u..
r, rg ,_
' cQ. s [
i I ' * ' ' * ( ._ q- ft o.
.. ..s '.s. .
l I #
'i, ,' . e6 , ,. -
i I j photo $4. t I I ! View of the shaft missile and half moon trench cut in the gravel (point 1). View l is looking cast. > L r 4
-m _ . - _ - _ . . . .. . - - _ _ _ . . _ _ _ _ _ _ _m-,_
l l I se n 1n ,;.,
, ; *;, :"? ,y.
'a . ..,,,,%~
?h?!*g(yT 4 N [b:s?'yt
[ * ; %e , j. . ;* . $ k < k.
- 4 s sN s
' *W , ; C % . > %:'
~p a Q fe*~r., . "*-.p. . .;,s . y+
7 ,.0...t ~ g~.
'*i:<
J. .s~ .% , , y "; j t' ,,,,,
*~r;
- j R' i c..; o a '>M . ,, *j. % e
. . \ ~, ,.* ,[ .he . ,:*3. ,.
- .'*j ,. - 2s ' ;,,,,c,. * . ,s x
*e ,e. 'i. fan.'"g'i: ~ ,c yR ,
[ ='f w w~a y , , . . . Ay-. , f: *'*'3
- 4. >*- .
*'a?Ie f , ~ .
~-
g'g e c.,,g,,
.c. +
. ..:,V'. .gN
.e ;.Mm a.j# .M' .*,Q. . , p. i l
n . p' .a ,,7 . ..
.a ;.r h., pf yy . , , - . %.y.&yqg *Q {YQ*% *\ N wn ,
-ysyn & ~~;*. *;4. < . s m. - ,
'QR= s.,m ~I'%fyy
..e,,. .g < '
, ,e p.( ,#.- R ~. .a
' : 4 ?..-y .p?t'N'A*. .,J I& . . .QQ- 5.
,p., e ;,,, *we A,+@,h r '
' 4 , s.% e%Mt' gl t,,/
* +
* '. . a w ;. q '* t' d ;?py;*'%. g t.'.QV; ,y&
, hW'bfR~, W '. ?* -n.v ,, N%' ,r + r ., sp W 1.S' 'Q, r. W; . J.
.;,)&
s,
'Y ,
% .. * - 9 d4 m g: h
- g
- . ,N,, :
g, '*. ,. q t, *
- t*;iA-li -
}>#.f>. . ;tj ' .
%*% r
. +'
'**N4.44 x
l ,
. a i
u- / / k '*N':. h p% . N
^ '
s.. w, .Q.4%
>. . .e-w,;f,1 ,~y '
%. . e; N W &'. a W,;.. Q , T N , '
1 , l 1 s . 7 ., v f. S ts % ;,%l'J ' k*
,M. M,g
... A '. .
aJ s.
'-4fig.ss 9, p -
i i O ','N ** g
^ ' 'bM - ~ 5 % *se I
i photo 25. I
- Close-up of trench arid turbine shaft missile,looking south.
f, i s
*y
\ .
<- c,;. (.. $, *. .
f-
,.,[
.. r;r ,<, 'g* / ,. **
p;h _
*. ,]. ,- 1
, <, ' 4
. ~ . .. .
y*
,'.g# a,
, *4 y * *% ' %,o+ ..*
* . \
s \ ,, , o*"",, 9.- g * . . . . ~ . .
' )Ql. ' " g .,
' 4' '
\
, 1r, ' ,
o *
*W,.., ., g , D, "tP f, ., , p;j. .' y, .
.E
\
~ ,*~**.t*
/.
s
% 6
' ~- ., .
\ %' ,**;;2d',
.',[d'.
p.' : @,
~ . 'A f' Rh. $ .,J - * 'A g -
~
g .. '.
,,,. 3 \
* ,i -
.s
- y y ' ,**h* %.
*d' ' %. *
<*"*".'*A \
y,. ff , 'N. g ,J%'. ,f, ,
's r '
'.+A
..q 'y \
f,.* *,*,,
, c% -
- e. ,, ' .
4 4, .
- N x
*s,, a My g' [
\
2,,,,-**./ \
^s
.A),: -'. y ~,.s,',%,s 4
d w-
~ ? ,,a ' 1' \
'* M<* ,,# _, \ .
\
s.. ~ ' . , .
. ?lm.* 4.'.
3
.[& a
_ f,
~ - s-
.s .,;;[$*'\ , '
*1
&.&'4
,. 1 '
- s. .g . .
p; s,
- 4kA /.'
$. W,M se t,n , t - r f*:x g lY ,s .s
,"Q N 3/*%p
- d' 1 f'f. ( *1 ' *,
\b . ,y s
.x ,& c%
f L ' v(.s' 1 ) ,, ,q)* p4 >
\
, a '{ '* . -
\
s ph.40 26. l Close up of the hole created by the spinning shaft missile. A slight bend is l visible near the break point. l t i l I
i ,, .
~-
.. m ,~,-:: u:s.~g..
.r m -
.x . .
a .4, . . . , . x. .~ -.. . ; p%; ,,. w,
- a- , ,. . o . , , .n .
r~ ..: .s
....x-. .: n y
.4.
;n,u .;
, w .. .ny. -
.v- . . . .
$I.?, h,{y. . . ,.~~ $ v? s. .
+u,f- - 4. . -l ;c :..~; . 4., ,,
i ~^ ' $~ W a ::. f; .-
?.1- 4 . U w. lsQ'" .,
. -f g , . ~.- .
s -+ , . . , . , - . L -
.d* '] [.e.
-it.' st,c.
T s, ..,..'x o p .-
, ,, ~.k.; : ., - n(l L,~' .
, .- w ..... -: . -
'?
-h$.$$!..a ?
a,ss >. . F ', Y
,-l,K.
~ . . . .
.g{fE , .e ,
v).s.,
. . . ..g, s, . .
& ~*p * ', ;.j e, \ ,,i,J . *
** d,: > ' "' J - P.$
Y o ~, s -Q. f,i . s
' tf4
.. ..'.'?<*a f , -
_ . +3 y photo 27. Dirt buildup on one side of the coupling bolts indicates the direction of shaft rotation. View of the turbine shaf t missile is looking west.
,p-,. ,,
- 4. .
.t' '
d
' , ', 4 - 6$i
.y . >
g .. . . .- t '. f.j .f _:.y". - ',$, w Q gj J, ~ . -4 4 j, I . .s, , [4 ~ ' f
#.(
i
.s , Q
't ,,-
[ ,. MI j,k ..
,1'-: N ,ll'f o
~ ..Ua ' s
*~ ff,. f
. ..g ,
.- t
- s. .~.< ,
. y
- i (
, sy -
. ar .
er
- b. [
,1,/
v.. p' 4 , z
.i .
q v .. ' .s ,ob, .': ~l ~.,{-V" 1 l
/J pheto 28 Closc-up of turbine shaft counting bolts and dirt buildisp.
_-_ _ _- . . . - - - . - . _ _ . . - . - - . . _ _ - . _ - . - . - . . . _ . - - ._ .- . = - _ - _ = -
'- -o,** , *' ,
.a '
. <
- re
-e p , , .y *
., a e, a
~ "'
,~ ,
~~
., Wp, **
* * .'e*m ; ,' ,
*, ,,,~***'
,,,. , s- f* ' s , .p ' . . _
_
- Ag */'o* 5f ~ f
= , .s . - p a . .
na
+6 %
r *- * .
. j *. g W
< #' . e ,
*= g Sp /*'*
j... _ , -
, .~.*
p '# ' e r, l l f i
*s, -J ,
4 t
/ ,
3- v, *4'.
< ( , ,
. s, f if I r ,
t
- e i ,
d
= i
~s
~ ~e photo 29.
Close up of upper portion of the turbine shaft coupling bolts and dirt buildup. View looking west. ! I I l s .. a ,';
- s.,,,' ,$$ ,
h'k '
- > - . -8 y~ }Y
,he et $,5 ',
f (('N s [+Nklk,"*f,,
.} '
l ,* . i ' .' ." . y ., - I' .,
-j'
..'-. p f ,A y ...s *
#;> o
o ., l j'* -
- e , .
s l ' , , ,*;e . "* s , . ,
' 5 , *,' . J s'
' 7 Jes,p?g%
l ' ' .,. , +.6, l f , t I', , ; ',
, 'd i fc ; , 'N -
. ))' .
i l t i.s 4,
,' y ' -
r p,','.'>,
' .')s.
, ,;sp e
- s ' prt..* . . . >,&% , , ) ,, s_.,: .
i p: ; ,
,s.
t i i photo 30. Close up of the turbine shaft missile showing bcnd in the shaft near the break. 1 t i
%%% ~'
m..,.,-._, , r . . , , y.. a za_, .,
.y i,'-
/ i e i
*o,.@0:i;;cy /
li
*,u,,. .
.. ~-
- ~3- ,
i
'. r
.J.
* ~~_
- . ,' >f x , I k .: W ~' *** *. R ~; .y.p . .,4...
1: n
~E.* *.*;
T* , .. 4 l e
*' ~
l ... pg';,-. N -_ l a . y - . e t photo 31. Broken PVC pipe (part 2A)in final resting place (view in front of makeup tank looking northeast). 4 \ 1 < 1 1 1., l 1 6 l t .) I I \ I i > I i i l l l i i i t i l i 1 l l - i l l 4 i
?
1 i
.i i l 3
3 4
APPENDIX E SOUTHk'EST RESEARCH REPORT "METALLURGICAL EVALUATION OF FRACTURED COMPONENTS FROM STEAM GENERATOR FEED PUMP DRIVE TURBINE NO. 11"
i METALLURGICAL EVALUATION OF FRACTURED COMPONENTS FROM STEAM GENERATOR FEED PUMP DRIVE TURBINE NO.11 INTERIM REPORT SvRI Project 17 5770-565 Prepared for Houston Lighting and Power Company South Texas Project Electric Generating Statien P. O. Box 308 Bay City, Texas 77414 Prepared by Nondestructive Evaluation Science and Technology Division July 1958 u ;
<weqx f=p#'
=
) SOUTHWEST RESEARCH INSTITUTE
{ f. [-
. e.3 SAN ANTONIO HOUSTON
METALLURGICAL EVALUATION OF FRACTUFGD COMPONENTS FROM STEAM GENERATOR FEED PUMP DRIVE TURBINE NO.11 INTERIM REPORT SwRI Project 17 5770-565 Prepared for Houston Lighting and Power Company South Texas Project Electric Generating Station P. O. Box 308 Bay City, Texas 77414 Prepared by Nondestructive Evaluation Science and Technology Division t l July 19S8 l Written by Approved by h// kW d.T Csw
'erald A.Laniping '~ B. T. Cross
(/ Principal Engineer Director Department of NDE Science and Research
+, -- ',: 3 -
TABLE OF CONTENTS l Pat
- 1. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1
2
SUMMARY
AND RECOMMENDATIONS . . . . . . . . . . . . . . . . . . . . . 2. 4
- 3. METALLURGICAL EVALUATIONS. . . . . . . . . . . . . . . . . . . . . . . . i 4
3.1 Evaluation of Blade 53. . . . . . . . . . . . . . . . . . . . . . . . . . . 4 3.2 Evaluation of Blades 42 and 38. . . . . . . . . . . . . . . . . . . . . . . . 5 3.3 Evaluation of Rotor Shaft Fracture . . . . . . . . . . . . . . . . . . . . . . P I P P W I t [ l II
LIST OF FIGURES Firure Eagt 1 IAst Stage (No. 6 Row) Blade Portions as Received for Evaluation ...... 6 2 Remaining Portion of Blade 53 Root Serrations . . . . . . . . . . . . . . . 7 3 Microstructure of Blade 53 Root . . . . . . . . . . . . . . . . . . . . . . . 8 4 SEM Fractographs of Blade 53 at Root Fracture Surface . . . . . . . . . . . . 9 5 Sectional Views of Blade 53 Root Serratic.is . . . . . . . . . . . . . . . . . 10
'~
6 Metallographie Sections of Fracture Surfa< e. . . . . . . . . . . . ... 11 l 7 Remaining Portion of IAst Row Blade 42 . . . . . . . . . . . . . . . . . . 12 l 8 Remaining Portion of Last Row Blade 38 . . . . . . . . . . . . . . . . . . 13 9 Fatigue Crack Grewth Features of Blede 33. . . ...... ...... 14 l 10 Turbine Exhaust End Bearing Journi. e'racture Surfaces .......... 16 11 Microscopic Features at Fracture Origin .................. 17 12 Microscopic Features at Fast Fracture Region . . . . . . . . . . . . . . . . 18 w l l l 111 e f . , , b
~ . .
- 1. INTRODUCTION On May 25,1988, the No. Il Steam Generator Feed Pump and Drive Turbine machinery of South Texas Project Unit 1 Electric Generating Plant experienced a catastrophic failure. The failure occurred during a loss of offsite power (LOOP) test and resulted in severe damage to the turbine rotor, turbine stationary casings, pump rotor, pump seal housings, and coupling. South.
west Research Institute (SwRI) was requested to evaluate the damaged components from a metallurgical basis. The metallurgical evaluation by SwRI was required to supplement cther ' analyses being per'ormed by Houston Lighting and Power, Bechtel, Westinghouse Electde, Stone and Webster Eng. seering Corporation, and Kalsi Engineering. SwRI's project activities consisted l of (1) a trip to the South Texas Project site to visually observe the failed components; (2) lab-oratory azaminations of two fractured last stage (No. 6 Rotating Row) turbine blades, (3) a microscopic enmination of one shaft fracture surface, and (4) participation in a meeting to provide recommendations for corrective actions to prevent future machine failure events. The corrective action recommendation meeting occurrW in Orlando, Florida, in the engineering offices of Westinghouse Electric Corporatics. A summary of the evaluation results and recommendations is presented in Section 2. A more detailed discussion of the metallurgical examinations is given in Section 3. I 1
- 2.
SUMMARY
AND RECORDENDATIONS The physical and metallurgical evidence indicates that the failure event (1) initiated in the steam turbine, (2) proceeded to fracture the turbine rotor shaft, and (3) resulted in damage to l the flexible coupling and steam generator feed pump. Because Blade 53 of the last stage (No. 6 Rotating Row) blades was found by SwRI to be frsetured across the blade root, :t is hypothe-sized that the blade failed due to tensile overload from turbine rotor overspeed. The loss of l most of Blade 53 caused subsequent damage and unbalance to the rotating turbine rotor. 1 1 The fracture surface of Blade 53 contained features that are typleal of a tensile overload and showed no evidence of corrosion or fatigue Anage. Blade 53 was the only blade that was found whleh failed in the root section. Blade 3 was observed by Westinghouse to also have erschs in the top serrations of the blade root; however, fracture of this blade occurred in the airfoll. The fracture surface of Blade 38 had features whleh indicated some fatigue crack growth, and further analysis was performed. The fractures of the turbine rotor shaft at the inboard and outboard journal bearings were the rwult of tensile overload in a bending mode. The journal portions of the shaft were bent in the same direction and did no show any signs of bearing seizure or loss of lubdestion. The fracture surface at the turbine exhaust end contained dimple rupture microscopic features at the crack origin. Away from the origin, cleavage type rupture features were observed. The fracture appeared to ' fan out" or grow in a fast fracture manner in all directions from the origin. There wes no evidence of fatigue or stress corrosion cracking. The macroscopic fea-tures of the inlet end shaft fracture were similar to the features of the exhaust end shaft fracture. The damage to the coupling and pump shaft appeared to be the result of the fracture of the turbine shaft at the exhaust end journal. The macrosecpic features showed endence that the couplirg spacer was pushed against the pump coupling hub or Dange and then was ejected ou%d The outward motion of the coupling and journal portion of the turbine shaft caused the unsupported nump shaft to bend and to damage the pump housing. The shaft fracture and the coupling did twt show evidence of torsional damage. The severe extent of the damage to the turbine and the urgent need to ascertain the rcet cause of the failure prevented the completion of metallographic examinations of all important fractures. Some of these fractures were examined cursorily, so additional examination work would be desirable. If additional root cause analysis is required, it is recommended that the metallographic examinations be completed, analyzed, and documented for the following fractures: l (1) All three shaft fractures l l (2) Coupling spacer fractures (3) Coupling flange bolt (pump end) fractures (4) Blade Rows 5 and 6 altfoil fractures. ! The fast fracture of a steam turbine rotor shaft is a rare event since quality materials and conservative safety factors are typically used for this application. An engineering study of the dynamic response of the turbine rotor is recommended to determine the result of a blade loas on the turbine rotor. The study should determine what happens aftar a blade is ejected at 2
~ _
high rotational speed. The study should ascertain if the shaft bearings should be a plala journal type or a tilting pad type for rotor stability concerns. The root cause for the tensile overload on Blade 53 should be determined, and corrective actions should be taken to prevent recurrence. l 1 l l l l l 3
- 3. METAL 1URGICAL EVALUATIONS Portions of Blades 38,42, and 53 from the last stage row (No. 6) were submitted to SwRI for metallurgical evaluations. Figure 1 shows the portions given to SwRI. SwRI performed metal-lurgical evaluations of the fractures of Blades 38 and 53 and observed the microstructural featurse of the turbine rotor shaft fracture at the turbine exhaust end journal. The findings from the evaluations and observations are presented in the following subsections.
as . Evaluation of Blade b3 The remaining portion of Blade 53 consisted of only two blade root serrations. The third serration and the entire altfoil were separated from the rotor and were found among the loose debris. The blade root serrations were found intact in the rotor slot between the rotor steeples of the last rotating stage. It was the only blade observed to have failed in the root section. Figure 2 shows the macroscopic features of the fracture surfaces. The fracture surfaces were l clean and were not covered with any deposits or corrosion product oxides. Metallographic l analysis, as shown in Figure 3, revealed a microstructure whleh was typical of a martensitic stainless steel blade material. No chemical analyses were performed to ascertain material constituents. No mechanical properties testa were performed to ascertain material strengths, , ductility, or fracture toughness. Information about the material constituents and properties was not critical for metallurgical evaluation and can be obtained later if deemed necessary. Scanning electron microscope (SEM) examination of the fracture surfaces revealed features that are attributed to tensile overload. Figure 4 shows SEM fractographs along the convex side of the blade. There was no evidence of fatigue or intergranular crack growth. Metallo-graphic sections were taken r,t two locations to observe the microstructural features. Fig-tres 5(a) and (b) show the cross sectional views which show evidence of plastic deformation in a tensile mode. Figure 6 shows the tearing that was observed along the fracture surfaces. The tearing is consistent with a tensile overload failure mechanism. 3.2 Evaluation of Blades 42 and 38 The remaining portion of Blades 42 and 38 consisted of the root sect'on and short length f of the altfoil. Blade 38 was fractured between 0.5 and 0.8 inch from the root platform, while Blade 42 was fracture between 1.5 ud 2.0 inches from the root platform. Figure i shows two views of the damage to Blade 42. Figure 8 shows the fracture appearance of Blade 38. The fracture surface of Blade 42 wu severely damaged and no additional examinations were performed. The fracture surface of Blade 38 had features that indicated some fatigue crack growth and further analysis was performed. Figure 9(a) shows the portion of the fracture ( which aahlbita the fatigue crack growth futures. A macroscopic axamina+ ion of the fracture surface found a flat por:Jon with distinct arrest marks indicative of fatigue. The initial stage of the fracture at the trailing edge consisted of a small area cAented at approximately 45 degrm to the blade axis. Thl is definite evidence that initiation ocurred by some high stress loading in contrast to in service high. cycle fatigue. Scanning electron nicroscope (SEM) examl-nations of these futures at three di#erent locations are shown in Figures 9(b), (c), and (d). These futures include zones of dimpled rupture (Vigures 9(b) and (d)) which indicate that j ersch growth wu from high stress with a low number of cycles. The fatigue crack growth was not from an insenice type of high cycle, low. amplitude loading, but was from 1*ds applied after the failuro initiated and luring the rotor coastdown. 4
3.3 Evaluation of Rotor Shaft Fracture Both halves of the rotor fracture at the exhaust end beanng journal were subadtted to Westinghouse Electric Corporation for fracture analysis. SwRI observed the fracture halves at the Westinghouse metallurgy laboratory in Orlando. Figure 10(a) shows the fracture half which was ejected and embedded in the ground. Figure 10(b) shows the rotor fr3eture half whleh was in the radius at the transition in shaft diameter from the steam seal packing region to the " bearingjournal. figure 10(b) shows that some mechanical rubbing or metal deformation occurred which damage.1 the fractare surface; however, enough fracture surface was undamaged to perform an adequate tzamination. The fracture appears ta have initiated at the 6 o' clock position and propagated by tensile overload until it reached a critical size at Location 1. Figure 11 shows I the SEM features of dimple rupture within the origin area. The failure propagated by fast fracture after reaching its critical size. Figure 12 shows the SEM features of cleavage rupture within the fan. shaped area ideatified u Location 1. The origin area was at the circumferential location that corresponded to the bend in the shaft. This location would have been in tension prior to the rupture due to the bending deformation. i The macroscopic features of the inlet end bearing journal fracture were very similar to the features of the exhaust end bearing journal fracture. No microscopic examinations of the inlet end fracture were performed and observed by SwRT while visiting the Westinghouse labora-tory. The fracture features observed by SwRI for both shaft fractures were those of a bending type tensile overload failure mechanism. s I
%4 I
l l l 5 , h
l l l l l l 1 s
.g,~3 p -~ s -. g ., ,
,,.,.--g,
#53
( 33364 (a) ms; A .in
~ ~' ,
e53 33365 (b) Figure 1. IAst stage (No. 6 Row) blade portions as received for evaluation. (a) Convex 'Jde (inlet face is on left. hand side). (b) Concave side (inlet face is on the right hand side). 6
y..-,
- r. . .
' b ,
c;'r - , Y' M, . _. ,. , . , 430 #42 d53 33368 (a) 1X k- ,
\ 0%
f\ , s'-L - f M-
.n 8 v., 5, 33306 (b) IX Figure 2. Remaining portion of Illade 53 root serrations.
(a) Convex side. (b) Concave side. Note that a small part of the third Serration is still attached. The fracture surfaces are clean with no deposits or heavy oxide scales. 7
. n: ., . .m n, r y,;.v. mn e , x, ., ~e , ,. .-,. . . p . tv
. d; ,,a:, s>c,, a, . n +. * -) . ' 4 f. . .; x,i,.cs.
e ,,O.: W ..-.r
- r bt v s + 3,, jg. ,o. N.', ,,3 , .j J?r x ,'-
- i. - , . s - ,-
1, 4 e a. ,..L . . . . , f. e. .hy.
*) 4,# Q4.9 A/.'y ,R:* *; i . 1,.. ', mt ', 9.
=
, ,s w.
*- ', . *,' f #
,%. .: , '3. .-;m'#.D is 'd .v ' 4,.-
- v. s <
w ,.s < " >x'c> , t
- _'a. .n/. e' Q <*)%'
y, . -P, , . V. ,. , ,, , y,s.: 4 : swr fa , y- .. a. . , ,f .. c , y/,. .^ s. ,v ., , . . . . , ,e p. l/. .,w. . .
' ' p .j. (,
g ,
. '.l . L ,, >, .w,<p@g.. m , ~ . ., .-,g. > p ',, 1:,, c .#j[ba*[v: ;< $,.j,m,g' ).
.. r . ; ., ~.,:.,
s . .
.n
+ .-.c'.,y.-
4 7, , , ey ,5 s, . s A4 ,
> ; .: $ v..>.
p G, . . .. '
. ;%, Q @, cm, -. e . .. ', o r r n ;3(,.p~ e f.3 ,
t.f. ,. - g*f ,v, . b t ,.2.l'. g. ; j . ' : J.%< . c ., .>;*.K. n ^*u*y:3,;,,Q. s
.. ,. ,-. ) n.
%,o e. w .. . **
)W.
4,( . , i qk,g: y',;pi.1.u,,.,4, .
., ,' 'f:*f v, . 3 4f .'lv-'s. , .%va;V,t,,w:2,
.et < .~.-
1 1 t. o.
- ? ;. ~ 5a .'; ^,0*q%"l, ! . ~ . -,a e f:. ,, ',., *
- - ...,~>
p< .~
,p .n,As ,s
).. . . , , , .
y t ,,v .
. s ,,.**.n. , .". ; *p*)*
o - i i -M. 6
.*8. .. s - 1,4
'*'h Cl t l h.* $, ' (W(..) ^
- Y,b.*;
, l .y 1, v<#.( -* v~ r y. >( G ' :Q y , 1n" P m .u,e . , * . . .. '. , ;,x . 1 ed. -
.. s ;n ..
v. p,,o
;p .c .3 .v g.,
i t t , 1
,p:
y
. .s pt .6,. 3. - c,3,,. ,... g-m,.y. e<s.^
y.e.
.t - .
. . , 4 ,,.m.,
., . , ',w t,/.. ,/.
y'Ay . . . y,:, . < ; .. . . . . o . . . .
.a f, .
+tm .< ,4,
.>. .. . S,e,g,i-e.,,.s L ~. .t -
s -
- s. s
- q. 3./ ,. <.e M:y,,.'..d f . .s / - s ; < . -:q .
4 u..
. Jy . ', '. ,;. ,M*
V i ~'*! ;,'. %;; %q .*y.l.: - L',.3elOZ.\'d.'*T.h.,yy9.h . d r :1 33421 (a) 200X l'?' %}
~
C ; W.[*& *-l.l!C' [. ., . ~, pC *- '.:-TR ^,W, a ;n .- >* &. ,'Q' ,3 gy' Q's.~ y,J c.f fR :. $ , ..
~
r e g <n. . g . . ~ a ecu.
.y p,,.~,. ....,a, /, , 5 e y,
.- y w .;, . .f. ,.A gt f. .r ^.g -G. .:,,
v.(. . p9s; ,','N,;r
-.. , g; ..
..r wW,7 ,~.,A , 4 jc.,,,, . ,
, . ~ ; . ;-
r
..r.m*. c, .
- u. .
, , - - <g om ,
. ,, j ,4- a,
.m. .r .?,. t, 8 . . . . ..
5 4 *.,
, , . . .N. ~~y.:y. . +m- ,s.<. .e.
*q ,cg. ,g g .. >v. ../..,,. ).%, . < >x , y
,,g ,,.... 34 ,* g:
^.ia.a, .
.r'.- s. , .. :.,
t, - ,. .
;,;, , . 4 ~t.~,;:,*e. .-tax * ,;.' /,e<y
,r:t ,.. ;.s.,
< . . , , + . : p ,,,w s ,p
.n ,"w.n. , . e.> ,. . . , . .v,n
;;,t :.*v,f<f>i..73 .a-. .
;,e m . p - c
. e , s.',s ..
. b ;o-:x'g, ,^r e !,gv-(T
},*D. -
-~^,
.y' . . v, . iy ~.y ; s o g .,,,.%. . ,'%.** yf. p 4
o ::. ,..,.,.,.6,.,_ y3.g, .. .- f 3 s. *'. h / 2/ .. . g.y4 L,,' d s/, r., t g . ., , , .e. , ?q.-
, . A,;l t(. <p, n M , I':3 :3sy;y u(t j ,
U 4)62,4 fQ4 ,;'.; I ,,.5,a % v4 o
, t
?' J 4,\,q~fg<d,k, l',,'Q 3 - , ' $,
- s. y
,j p,.g';y
, 'y,y 3 (.! y'i.43,
* -J 3_
- Q 1, ft; .,,
w ;-Q . L(p ,A,9 ep.M[j, 9
.'?. c , y?,, ' ,^ e l
\ ^ '. o . '
1,.
. .M..y
. . s s
/p"g!.m
. ,. *.o ,, ., . ,.,s!,.>r',% ,.
w . -:,q< . r'k(h ig+. , a - d. 4 s, l
. .; , ,%, ~c<,1,,.
t.+. s p. ,.:. i,g. ., y ,. .9. y .1 . , ? .,s,
, - m u ; * , .'
- 4 /qt.. ,. 8s - .,.es,n , ,.\. i ,,: j s, #.. .g.rp , 4'Q. o..r
- v. . p t'y ,,*.*e., ,y g ,.,- y. % ,<',. , . .
.. ., s . 'g~ 9.., g yj,f.j1---+ .
("i- h, .., s Wq,n,gyn.Drs.,Qp* % q R.s... '. / . ig,,?,:g, e[% ... of 0 7 9 s g;r%. td. '\* P*pe l
;; ,.l,
.sv - -
s 4
. 4 N. * '. A.%.>
l
.,mk.,.:t '.v < 1s .s>,s,* a *n
* '. t ( '
ho M .,e s v .vM . -3 ,. i.a, Z* , .sa, fM r.* '*'I,U AA ( l a,, 's &. . , 33123 (b) 1000X Figure 3. Microstructure of Blade 53 root, ta) A fine marten-side type nuerostructure was found. (b) The rnierostmeture 'us typical of a 400 series martenside stainless steel. 8
.. / L k__
. '~'j ' k .;' , , .
lfg3- , 1 ;,) fdI.1 p (!MM[ x , P,h. . -y f- dja ids l f' N 'hh hh.
! j " ?.1 F.4. .s? si
&q&qd ?;>:dp . d$ d M E-j q 3
}. 3 k' "
'W l k .t
- h. k ,I L
- 3
'4+'!, p.; h5(.. y wiw .! - jefe'
- b+.y l .
t kihi :F ' - 4 k 33"71 (a) 500X h' h* > $'b L $N %
.QUQMh
%4 W.49Mh$diQid L%?'d ge;;l.. s 4
.[gw kbhGf'
' Q i%th,,*\:,l I.
.-py ' ,
- Q '?O f Kv M'YM45ja; '
. t ; ) . 1 7.7 ~'%$
e._(., " Q
? e:
(< y y 4 j. A , h% )y <' ,
, t , . '.'!\ }.i.,p';.Ir{N j
L .:' '
]
k'
,R. t, f,y.-Q J,),,;.E] {
f f hlk2.< ; .'-] . s,' ??;Yc/." li kg) 33373 (b) COOX Figure 4. SEM fractographs ofIllade fa at rc .a fracture surface , 9
+ -
4 i
/
i c l$ ( y f p.
..~.
th :b C
.ff. .
sj ,
, 4 ,,
h.
- v. z
~. ' , l
. :- a -- : .-~ ..
33378 (a) 3X Qt Eu . y
, :, /C %% .
; ;- .d ;N
- c&; il,.
y
,c ,. \s ,. r a
tz, .Q y T '_ p .V GI o , , .r j ('tu d' Gff L L L.. ,.q, t , a 33324 (b) 3X Figure 5. Sectional views of Blade 53 root serrations. (a) Sec-tional view near the inlet end shows stretching at a steep angle ( from the second serration. (b) SecConal view at approximately i one half of axial depth shows similar featwas, 10
~ ________________ ______ ___ _ ___ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .
5-
. .c
. ~, . .
. , , -,4;
- s. ,
, t
~ / *L. .
.. , .- .~ :
A 7. .. s. v. a 33418 Ia) 100X t m E
= .f al e
j
< . f Y,
/
gY
". o e
6 1 0 8 er kjh lNf F unire 6 '.t i t iR it y l.a o < n. in c f fi u t ur e surface < a i a nil l (lil ir lilli ' ,Al x :!,g (lic f t.h t !)[ r sil((tte w is r,} >.c rT e,} g };jc }g jg cornatuit w it h a t( n sih i s cil. ..i I f u t u r e nau han; sin. Il
i
/ qv ~ ~ ?,
,cw *;
y
~
Ih P
*\
'\
. / ? a h ~' .
33396 4 IX >
' ~-
p e-vg _ ^'f
%. a..
33307 (b) 1X Figure 7. Remening portion of last row Blade 42. (a) SHe view with talet on left hand side shows distortions to platform at root and to taatling edge. Hacture was between 1.5 and 2.0 inches from root platform. (b) Top view with inlet on right-t hand side show, distorticas to leading and trailing edges. 12
g~ _ . _ . _ _ - - _ _ _. _ _ _ _ . _ . . . . _ . _ . _ _ _ . _ . __. _ _ _ _ _ _ _ _ _ _ _ _ _ . _ . _ _ _ _ _ _ _ _ _ _ .
!- s i
J L 1 i f
*i'rWv .,b
\
~
% w++~
f 8 C ,_ . .-- -l ,. t i i i l ._ i i i l 333SO (a) IX ( a s 1 i i
)
i I 1 , l i f',' (3%
,[;
i ,g : . gu o
, , , gi
; ~s -
q.9 r ) W- , t l l 1 1 . 33393 (b) 1X l
- r l !
! Figure 8. Remsining portion of last row Blade 33. (rs Side t
{ view with inlet on le4. hand side sho vs distortions to platform. j i Fracture was between 0.5 and 0.8 inch from the root platform. ! i (b) Top view with inlet on left. hand side shows various fracture surface features ofinterest, 2 ]
?
l i 13 1 t
!. l l
in 33508 (a) 3X
*y< w >. w,: " r. i :n t y s :n vig'
'ha &c '
[} 2 ; p6 hg[lN)/ it
. i ,s p ,/ < d.s..
o/).hNk'f) f.y ' t 7,k tJ
't ,' D e r ,g i'Tf/'t -
?
,sqkiS[8
#757'c! _.
I d.?Id(Y,$g 33500 (b) 1000X Figure 9. Fatigue crack growth features of Blade 33. (a) A portion of the Blade SS fracture surface exhibited fatigue crack l growth features. (b) SEM features examination at Location 1 whhh is 0.3 inch from trailing edge. l 14
m..
, s ,.
-~
+ ,
-s Tl @ **. -
y- ~ g - .,, .
~e , 3~
t.'s
%; 'sa. 4 g e..' .
g,N
'y p '. yfy
..b. .
1 33504 (c) 1000X
',#[
'e II IDI' ' Y D2 V er- N'y a.:yr;<.:b:%4:?,h
%n"*r&m&,ro,b*?![/C
,i.2 ;r 4ti .
o
.y;i
/p.4,)y,s 'fy
*
- y. .<? *.6
(;ly')y e r.M. ). ! * ' * ;; t. ,* e~
< i .J$ gtj ' .,e N.
33506 (d) 1000X Figure 9 (cont'd). Fatigue crack growth features of Blade 33. (c) SEM examination at location 5 which is 0.7 inch fmm the trailing edge. (d) SEM examination at Location 7 which is 0.9 inch from the trailing edge. All fatigue growth was the result of high stress low-cycle loads applied after initiation of the failure event, 15
a; %
- . l'e';e Q ,
j u;,: ,x .N jNi;.k8']Yf[', 3 U k's ez.
':s; 9...;\, y% ;'a f'q. s A 1h II' ,/ f .,
.h' $ )
i y, I, ~ =,V n ' '. .' . ', v .
' & -t i , ,' :i:
)ai' 'h> s i ; f: ?. ("
,a fll
.Y' !
E E01 0 E 012 (a) 0.5X
- 'U.h j; y '
l ,l '
'=
..g..
,,,= y r
l
/
, , '.Y . I9 l
t . g[-
- 4
~'Nl.,. ;!y. # v
$ o 3(2 ; . .
' } y G' .
, )) .
/,
l
- 1 L .
E E010GSS-012 (b) 0.5X Figure 10. Turbine exhsust end bearing journal frt:ture surfaces. (a) Fracture surface from shaft section which was ejected from the casing and became embedded in the ground. (b) Fracture surface from rotor in the radius at the diameter transition from the steam seal packing region to the bearing journal. Fracture origin is at the 6 o' clock position at Imation 0. 16
)
1 Y A ,Y.
.s ..
?. g,~. W~.?Yb $ m;-u? &s . .. h i. W '
1 N .? -
.QV q? -..;:. .;c y;e:..4;yy.:.'.:.R.
* <sjy ~V .n*q.#?m* d.
~
?Q,; * * . V .] , r %v.&.~g. . .
> ; . ..> < . n :'.< .% '
~ --
4.5);,e,% **! , 1-M* ' ;+6 y &4 r* {. a:'CWh 5 .s..' mve . ..
' QN's '
):.
-% m b
'rh &k.*. ;t $
s.C="..p hkh. A W.* t * = f.p Jf24 .M M J sT
- a b Z e c-s.ase QxM'~ M='.
E F01 0CA%-012 (a) 640X
~
55 5 A WfjA 9j'y g. ) ,c. q ~* y , M N- T -ly. % [. ? M
' ~ '
% . .u .
lll%L'#.%j *'stS*,
- f.*g ,o?p-4
-t *T
. .c .N a a g (a 1
- h.,. -
QY ,- ,- i' s -
. .y ~9 .;. ,,;sf 11
,sy -
i r t 4 ps , iT . .. Q g,e . ...' 4
,k E E01 0&% 012 (b) 100X Figure 11. Microscopic features at fracture origin. (a) Dirnple rupture fracture features were observed at the crack origin. At location 0. (b) A change from dimple rupture to cleavage rupture occurred at the beginning of the fan-shaped region in the lower center portion of Figure 10(b) at Location 1.
1 17 6
1
!- .@ s.~ d M .s, M,1 e g % %"
I c.
*i YAf ./ **".k !kkN) .
W't' j 4*e rt c ~,b ' * ,.
' . . , 7I; gj ' .
' / $'
% 'Aj[8{,,[,i y? f@.. j.7-;;'*!T8,.m he,sy ...
~
f.w#N.
. ;y q
'M ,
.2 b. %. sk"s'ft,;m =.f.Y$gf..
_ . .a E E0. 06SS 012 (a) 100X h A
- f,
;
- T" , , 3 ,, s p- ,
s
, 'id .
& Je > .f\r I .- '44 , .,e r, J'*gj f*fje .* L'.h >f, ;'"'[
8 , Mh .(W; \ A Qb, 4 - Oi.'k'AS?Zcas [k8$$^$ykke, .
?,
S
'g . k I Mk r kk'\ .be "'M Mr E E01.coss 012 (b) 50vX Figure 12. Microseupic features at fast fracture re,: ion. taiand (b) Complete chavage type rupture wa3 obsened in fan shaped region above the orien area at locadon 2.
15
f f ATPENDIX F k'ESTINCHOUSE CUS'.0MER ADVISORY LETTER 88 01 e l F s
+
s a b l [ r h 1 1 T-1 I O i t
+ . , .--_,-_- _ . , _ , . - _ . . . -
--n-... . - _ . , . - - . _ , _ , , _ - - . - - _ , _ _ - - - . - . - , . - . . - , - - -
I i O) Westinghouse $305). gn m,, Bectric Corporation n;.n u nrra: CUSTOMER ADVISORY LETTER 88-01 ST-WT-ML-933 I July 5, 1988 l l Mr. W. H.Lighting Enaston Kinsey & Power Co P. C. Box 308 Bay City, TX 77414
Subject:
CAL 28-01 Feed Pump Turbine (FPT) High Pressure Stop ValveNpr[ngPre-Tension South Texas, Unit 1 - 8.0. 15A5286 ; South Texas, Unit 2 - 5.0. 15A5296 l Mr. Kinsey: 4 R:cently it was determined that the linkage on the FPT high , pressure stop valve may have been such that the spring pre-tension may not have permitted closure of the stop valva. It is recommended that the stop valve apring pre-tension be ! varified and/or adjusued at your earliest available I opportunity. Attached as Appendix A is the recommended l i l procedure for verifying / adjusting the spring pre-tension on the FPT stop valves. Group A of Table 1 - Specifled Dimensions of Appendix A applies to your unit. Operation of the FPT with the !' J high pressure stop valve spring pre-tenwien at less than the > Cpecified valve in Appendix A may subtset the FPT to potential overspeed conditiens that may result Ln the destruction of the [ FPT and could present potentlal safety hasards to plant personnel. , ! t ! PLEASE NOTE THAT THIS PP.5 'tENSICN CHECK AND/0R ACATUSTMENT ' SHOULD BE PERFORMED ONLY AFTER STEAM TO THE TURBINE AND THE CONTOL GYSTEM MAVE BEEN WECURED OUT OF SERVICE. i i l l I e
, ,- ., -- - , _ _ . - _ ~ _ _ _ _ . _ , . _ _ - - - - _ . _ _ _ - _ - _ _
Mr. W. H. Kinsey July 5, 1988 If any difficulty or inability is encountered in cchieving the dimensions specified in Appendix A, please contact me at once. If you hava any questions pertaining to these roccamendations or if further clarificatien is desired, please do not hesitate to contact me. A copy of this letter is attached. Please acknowledge receipt } of ule information in the space provided and return the copy to us for our files. i Very truly yours, h f.H. C Smith Technical Service Manager jl Attachment cci B. Heery L Houston - EUS Tom Da e Westin house - Jobsite STP
APPENDIX A FEED PUMP TURBINE XIGH PRESEURE STOP VALVE SPRING PRE-TENSION VERIFICATION ADJUSTMENT CAUTION
- STEAM TO UNIT SHOULD BF. 8ECURED FRIOR TO START OF WORK.
- TME CONTROL SYSTEM 8HOULD BE SECURED PRIOR To COMMENCEMENT OF WORK.
(FAILURE TO COMPLY WITH THESE CAUTIONS COULD RESULT IN PERSO INJURY.)
- 1. Review the valve and linkage for proper assembly.
Reference Figure 1. Verify tha linkage is free from binding. at flat ground and
- 2. Inspect for assembly the twowith closing springs spacer (Item(Itam 2). 1)In stae cases the spacer may have been removed in making previous adjustments and are no longer required.
- 3. Loosen nut (Item 7) at the bottom of the spring assembly bracket (Item 8) to remove any existing pre-tension of the two closang s'prings as noted by looseness in the spring.
Verify that tse valve stem is at the bottom of its seat.
- 4. RemcVe the set screw (Item 5) fromUse the the clevis (Item flats on 4) theand loosen the sten jna nut (Itam 6) .
valve stem to turn the stem inside of the clavis and thereby alter relative Turn positions of the the sten lever and to ob*,2in actuator a gap of and closing springs. dimension "A" (see Table I)he spacer, or top of bracket of between the flat ground and the spring and the top of t where no spacer is existing. Reference Figure 2 Detail A. At no time should the valve stem be turned if there is spring pre-tension present or galling to the valve plug or seat may result.
- 5. Measure and. record the length of e gosed threads from the bottom Reference of theFigure olevis to the startB.of the threads.This ler.gth is not to exceed 2 Detail 1.00 in.. Remove existing spring spacers if needed to conform to this maximum length.
- 6. Retighten the jam nut (Item 6) and set screw (Item 7) located on the valve stem and clavis respectively.
- 7. Tighten the hex nut CItem 7) at the bottom of the spring I assembly bracket unt:,1 the flat ground and of the spring !
firmly contacts the bracket therehy taking up the full gap of dimension "A" (see Table I) provided in step 4 above.
- 4. Rameasura and record the length of e gosed threads from the
' bottom of the clevis to the start of the threads and verify the maximun allowed 1.00 in. noted in step 5 above.
- 9. Adjust the actuator stop to obtain a nouiral length of travel dimension "B" see Table I. Re-adjust the limit switches LvDT and valv(e dettings as) needed.
- 10. Any diflficulty or inability to achieve dimensions requested st. auld be brought to the attention of the local Westinghouse Tec.'nical Service Manager at once.
TABLE 1 SPECIFIED DIMENSIONS
- GROUP A GROUP B OROUP C GROUP D DIMENSION A 2.0 IN. 1.40 IN. 1.8 IN. 9.75 IN.
SPRING GAP DIKZNSION B 3.0 IN. 1.25 IN. 1.8 IN. 4.25 IN. ACTUATOR STROKE
*IF YOU ARE AWARE THAT THE SPRINGS IN THE STOP VALVES OF FPT HAVE BEEN CHANGED SUBSEQUENT TO 8HIPMENT OF THE FPT BY (W), PLEASE CCNTACT YOUR TECHNICAL SERVICES MANAGER.
_ , , - - - ____,__.2- - - = *
@~
,,;. i
' 0 'i j , ;
g g
~~\',. \
- ., s
'h, ,' u'f,$
', ' @x M -)....e
.. ~I . .-- M {
l rM / .! /
'o (( ,
I t
- g ,
8 . . .
- __ o a i l 3 '~
in . ^?,' ng \ i l*(ha w g * 'ggjg vc = a j O ns
~
_W "7"tq') v n --
) _
1;
~ -
, 5 1
\ ll!Ii!I!lli
.n..
@ -=
N
.,id g =n
/ S \ -
i ...
+ ., - - r - -
- --r i l1I
~"H ="
g, ,, t
\
s - g ;- o , I
. < a
\ ' C 1
r
- .l'l".".,
[ El i
~
3 s, , . . . , - gc sy- , 7 1 i.ii
, ,' g< . . . . Ln
._,y__ 3.,3 <,
- p. us ,. . hj _ #
s
=
A a/ / I
~'
Y f r=
/ l ,
_ _ _ _ _ _____ _________ ._}}