ML20094N766
| ML20094N766 | |
| Person / Time | |
|---|---|
| Site: | Clinton  |
| Issue date: | 03/30/1992 |
| From: | Jamila Perry ILLINOIS POWER CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| IEB-88-008, IEB-88-8, U-601958, NUDOCS 9204070072 | |
| Download: ML20094N766 (12) | |
Text
.-
o Ilbn% Pw.cf Cutpany canon n.we r.wm P o lk's 070 Clintm IL 017/ 7 Tel Pl? [#B Ns1 ILLINSIS POWER
$3}A -30 ) _tP 80.130 JSP-164-92 March 30, 1992 Docket No. 50-461 Nuclear Regulatory Commission Document Control Desk Washington, DC 20555
Subject:
Clinton Power Station (CPS)
Revision of RQEp.QDre to Bylletin 88-08. Sunnlement 3 Dear Sirt Illinois Power (IP) provided its response to Supplomont 3 of Bulletin 88-08 by letter U-601492 dated August 17, 1989, and supplements U-601623 and U-601693 dated March 16 and June 22, 1990, respectively.
IP's response was reviewed by the NRC Staff, and by letter dated November 26, 1991, the NRC Staff provided its evaluation of IP's response.
Therein it was stated that IP's " response to Action 3 oi the bulletin does not provide sufficient assurance that unisolable portions of all piping connected to the RCS (reactor coolant system) will not be subjected to conbined cyclic and static thermal stresses and other stresres that could cause fatigue failure during the remaining life of the unit."
It was also stated in the letter that " inservice inspection (as IP had previously committed to do to address the problem of potential cracks in piping) is not an acceptable technique...for preventing such cracks."
The letter included etlteria for IP to consider in preparing an acceptable response.
This letter provides IP's revised response to Sapplement 3 of Bulletin 88-08 with respect to Action 3 of the bulletin.
IP's revised response, provided in Attachment 1 to th's letter, is based on the evaluation criteria provided in the NRC's November 26, 1991 letter and on additional, clarifying guidance provided via several telephone discussions conducted during January and February 1992 between IP and NRC Staff personnel, i.e.,
Mr. A.
T. Gody, Jr. (NRC Licensing Project Manager for CPS) and Reactor Systems Branch personnel.
Application of the above guidance has resulted in conclusions
)
.and actions significantly different than described in IP's v
\\
9204070072 920330 PDR ADOCK 05000461 G
_. ~.
U-601958 previous response to supplement 3 of the bulletin.
In particular, with respect to procosa piping connected to the RCS, IP previously identified six subsystems of potential concern.
Of these, 1RH-34 was datormined by analysis not to be a concern on the basis that the piping wolds would not be subjected to excessive stressos over the lifetime of the plant.
This datormination remains unchanged.
Subsystems-1LP-01, 1HP-01, 1RH-01, 1RH-03 and IRH~05 woro previously evaluated by performing a bounding, conservativo analysis of ILP-01.
IP's previously performed analysis of this subsystem yioided high
)
stresses and cyclical loadings with a limited subsystem j
lifetino duo to fatigue (relative to plant lifo).
- However, after considering the additional, clarifying guidanc.o obtained from the NRC Staff (particularly with respect to the distance betwoon the isolation valvo and the connection to the RCS), IP has now concluded that nono of the above subsystems are susceptibio to the cracking or fatiguo failuro caused by thermal stratification as addressed by Bulletin 80-08.
Thoroforo, it 10 not necessary to perform periodic inspections of the wolds in those subsystems as IP previously committed to do in its June 22, 1990 letter.
Provided in Attachment 2 is a summary of IP's previous analysis of subsystem 1RH-34.
As noted above, IP's analysis of this subsystem confirmed that it should not be susceptibic to fatiguo failure due to thermal stratification over the lifetino of the plant.
This summary is provided (for information purposes only) because, during the tolophone discussions conducted betwoon the NRC Staff and IP, it was determined that IP did not provido sufficient do. tail concerning this analysis in its previous responce to Supplomont 3 of Du110 tin 88-08.
Additionally, the IRH-71 subsystem configuration was tho l
subject of much discussion :>etwoon IP and the NRC Staff due to some' sin 11arity to the configuration addressed in Supnlomont 3.
Tnis lettor, together with the information provided in its attachments, servos to completo IP's responso to Supplomont 3 of Bulletin 88-08 and resolves the concerns expressed in the NRC Staff's letter dated November 26, 1991.
I hereby affirm that the information in this letter is correct to the best of my knowledge.
Sincoroly yours, P
Vice President WTD/alh WSI19 WTD16 l
l
.. _. -. _. _ _ -,. _. _... ~.. -.. _ _ _ _. _ -, _.. _,... -. _. _. _ _,. -.. ~ _ _
U-601958 4
Attachmento cc NRC Clinton Licensing Project Manager NRC Resident Offico Regional Administrator, Region III, USNRC Illinois Departmont of Nuclear Safety I
r I
l-
to U-601958 Page 1 of 4 Clarification of Response to NRC Hulletin 08-08 Supplement 3 in Response to the NRC's Ictter Dated November 26, 1991 Ilh M GliQl' lid NRC Bulletin 88-08, Supplement 3, documented an event at a foreign reactor facility which raised new concerns on thermal stratification in unisolabic piping connected to the Reactor Coolant System (RCS).
At this foreign facility, cracks woro found in piping connected to the RCS.
The cracks resulted from thormal fatigue caused by hot vator, which was drawn periodically from the RCS hot leg, leaking through the packing gland of a Residual llent Removal (HilR) valvo.
The hot fluid flowed on top of the cool fluid in the pipo and produced a temperature gladient between the top and bottom of the pipo resulting in thermal stresses on the pipo.
The valvo leakage and resultant thermal stresson woro cyclic due to the thermal expansion and contraction of the Ri!R valvo disk.
This event is different than the event documented in the original NRC Bulletin 88-08 where thermal stratification resulted from leakage of higher-pressure cold water into hot RCS water.
The NRC has requested that the three actions in the original bulletin be addressed for the event documented in Supplement 3.
These actions are as follows:
A.
Action 1 - Review systems connected to the RCS to datormino whether unisolable sections of piping connected to the RCS can be subjected to stresses from temperaturo stratification or temperature oscillations that could be induced by leaking valves and that woro not evaluated in the design analysis of the piping.
For those addressocs who determine that there are no unisolable sections of piping that can be subjected to such stresses, no additional actions are required.
B.
Action 2 - For any unisolable sections of piping connected to the RCS that may have been subjected to excessive thermal stresson, examino non-destructively the wolds, heat-affected zones and high stress locations (including geometric discontinuities) in that piping to provido assurance that there are no existing flaws.
C.
Action 3 - Plan and implement a program to provido continuing assurance that unisolable sections of all piping connected to the RCS will not be subjected to combined cyclic and static thermal stresses and other stresses that could cause fatigue failure during the remaining life of the unit.
I to U-601958 pago 2 of 4
[
i Attachmo' ; I to Supplomont 3 of Bu110 tin 88-00 identified various approaches that might be used to address configurations like the one that existed at the foreign reactor and provido continuing assuranco that fatiguo failuro would not occur during the remaining life of the unit.
Ono approach was to reviso the piping arrangement to minirize the offects of thermal stratification by moving the valve "sufficiently" far away from the hot sourco.
An indication of what was " sufficient", however, was not givon.
As a result, when Illinois Power (IP) performed its ovaluation of potentially vulnerable piping configurations at CPS, IP adopted a very conservativo approach which did not considor the distanco betwoon the valvo and source.
The conclusions of IP's analysis were transmitted to the NRC on June 22, 1990.
IP's responso identified wolds in subsystems ILP-01, IllP-01, 1Rll-01, 1101- 03 and lidi-05 as a po':ential concern based on the conservativo analysis.
The i
analysis datormined that those five subsystems had a fatigue life of four years with the occurrence of stratification.
1111noin power indicated that Action 3 of Bulletin 88-08 Supplomont 3 would be satisfied by adding these wolds to the Clinton Inservice Inspection (ISI) Program such that they would bo inspected onco ovary two refueling outages.
It was felt that this would provido continuing assurance that piping / wold fatigue would not go undetected and would permit action to be taken prior to the occurrence of fatiguo failuro, thus mooting the intent of Action 3.
By letter dated November 26, 1991, the NRC indicated that IP's response to Action 3 of Bulletin 88-08 for Supplomont_3 "does not provido sufficient assurance that unisolable portions of all piping connected to the RCS will not be subjected to i
combined cyclic and static thermal stressos and other stresses that could cause'fatigua failure during the romaining life of the unit."
pursuant to this evaluation of IP's responso, ovaluation critoria woro provided in the NRC's letter to assist in preparing an acceptable responso.. Tnoso included a 1
criterion which provided an indication of what distanco betwoon the isolation valvo in the subject piping and the hot source (RCS) is " sufficient" to alleviato concerns.
A botter understanding of this-avaluation critorion was gained in subsequent tolophone discussions with the NRC Staff.
It was thus confirmed that the concerns presented-in Bulletin 88-08 would not be applicable when the isolation valve was greater
+
than 25 pipe diamators from the hot sourco.
This critorion forms the basis for IP's revised responso to Action 3 of Bulletin 88-08 for Supplement 3, as discussed below.
l l
._.--.._ a, _
_... _ _. _. _ m _ _ _ _ _.., _ _.... _. _. _. _...
to U-601958 Page 3 of 4 i
)lfV10l' LROlWE d
As shown in rigure 1 (p. 4 of 4 of this attachment), the configuration of the five subsystems in question consists of a locked-open gate valve, a check valve, and an isolation valve.
It is postulated that a small amount of leakage could flow past the check valve to the isolation valvo and then past the valve disk and out the stem of the isolation valve.
This leak would slowly heat up the disk causing it to seat tightly.
Leakage flow would then cease, allowing the disk to cool and subsequently contract.
This cycle would then resume after the disk cooled.
This cyclic phenomenon would reduce the life of the associated piping between the reactor pressure vesse' (RPV) and the isolation valve due to intigue.
If, however, the isolation valve was of sufficient distance from the hot source (the RPV in this case), enough heat would be lost to the environment and enough mixing would occur such that there would be insufficient heat available to drive this cycle.
The distances between the RPV and the isolation valves for our five applicable subsystems are significantly greater than 25 pipe diameters.
This puts these five subsystems outside the scope of Bulletin 88-08, Supplement 3, based on the evaluation criteria provided by the NRC.
Increased surveillance through IP's ISI Program as discussed in IP's response to Action 3 of Bulletin 88-08 for Supplement 3 (letter U-601693 dated June 22, 1990) will consequently not be necessary.
CONCL1) Slot.!
Based on the NRC-provided evaluation criteria, Supplement 3 of Bulletin 88-08 requires no further action for the five subsystems noted above.
E_UFSYSTEM 1Ril-34 It should be noted that not all piping which connects to the PCS at Clinton meets the 25 pipe diameter criterion.
Those configurations that do not, however, have been analyzed as discussed in IP's June 22, 1990 transmittal, and they have been shown not to be a concern.
Most notable of these cases is subsystem 1Ril-34 since it is analogous to the caso presented in Bulletin 88-08, Supplement 3.
Although this subsystem was shown by analysis not to be a concern, it was brought r in discussions with the HRC due to its similarity with the case that led to supplement 3.
A brief discussion of the analysis used to evaluate this subsystem is therefore provided in Attachment 2.
Attachmerit 1 to U-601!P,38 l'a g e 4 of 4 I
Locked Oport O
Isolation Valvo RPV
-lA H~
x v
~
L1~**- h '*
2 Illgh Low Pressure Pressure w
SUBSYSTEM L1 L2 Beinarkn L +L >>25D ILP-01
>25D
>>25D 1
2 L +L >>25D 111P-01
>25D
>>25D 1
2 L +L >>25D 1Ril-01
>25D
>19D 1
2 L +L >>25D 1Ril-03
>25D
>>25D 1
2 L +L >>25D 1Ril-05
>15D
>>25D 1
2 Figure 1
to U-601958 Pago 1 of 5 j
c l
Analysin of 1RH-34 l
STRATIF11;D_.l'ipW PijDipMU/dllagAgrojmd1 Thermal stratification is a phenomenon resulting fr om the lack of mixing betwoon the stagnant fluid in a horizontal pipe and the incoming, relatively hot, very slow moving fluid from any leakage source.
The donsor cold fluid occupies the bottom portion of the horizontal pipe while the more buoyant hot fluid occupies the upper portion of the same pipo.
Accordingly, the top side of the pipe, which in hotter, would expand significantly more than the cooler bottom side.
This croatos an upward bowing if the horizontal stratified pipe was simply supported, and a downward bowing if the horizontal stratiflod pipe was supported as a cant 11over.
With the exception of a stratified wator/ steam interf ace surf ace, the surface interface betwoon the stratified densor cold fluid and lighter hot fluid is not generally distinct due to the heat-conduction betwoon the fluid layers.
In addition, the leakage flow disturbs this hot / cold interface surface and croatos standing wavo-like surface oscillations within it.
This oscillatory motion of the hot / cold interf ace surface, scanning the pipe inside wall, generatos a localized thermal transient which is commonly known as thermal striping.
This thermal striping generatos a concern with regard to thermal fatiguo cracking.
SUBSYSTEM IML_J_4_ _ ANAL ySIS 2
Subsystem 1RH-34 begins at the connection to the 20-inch Reactor Recirculation (RR) Pump "B" suction line and ends at the penotration anchor for containment penetration IMC-14.
The portion of piping considered as unisolable from the Reactor Coolant System (RCS) includes the portion of the lino from the RPV up to Valvo 1E12-F009.
A simplified schematic is shown in Figure A (p. 3 of 5 of this attachment).
Analysis of subsystem 1RH-34 was concorvatively based on the following scenario-A.
With the system stagnant at a temperature of 130'F, leakage develops through the seat of valvo 1E12-F009 and the stem of valvo 1E12-F008.
This starts the stratification cycle.
B.
The leaking water, at a temperature of 550*F RPV temperature, flows at the top of cold water in horizontal piping runs.
t to U-601958 Pago 2 of 5 C.
The total number of stratification cyclon based on a conservative analysis was 15,000 over 40 years.
Striping, thermal fluctuations at the hot-cold interface, was calculated to occur at a rato of 200 cycles por stratification cycle for a total of 3,000,000 striping cycles.
1 liESULTS.,91 AN ALYS LQ por the analysis, bonding moments duo to bowing and the
[
resulting thermal stresson, as well as localized thormal stressos due to striping, woro calculated.
Those strouses woro then combined with design-basis loadings.
Both structural and fatiguo appocts were then ovaluated, with the following resulta.
l A.
Loadings at the connection betwoon the RH piping and the 20" RR "B" suction line woro unchanged.-
B.
Piping stressos and usago factors remained within
~,
the allowablo ASME Codo limits.
c.
Load incroanos on supports worn ovaluated and were found to be acceptable.
1 D.
Load increases at the containment penotration (IMc-
- 14) from.the subsystem 1RH-34 analysis were analyzed and woro found to be acceptable.
NOTE:
The first attempt to qualify the ponotration itself was dono using the fatiguo cycle analysis discussed previously.
This resulted in unacceptablo'fatiguo loading.- Tho original fatiguo cycle analysis was revisited and a number of assumptions woro datormined to be overly conservativo.
In the now analysis, consideration was given to cooldown of the lino and heat up of the valvo disk as those r
mitigating offects were not included in the-original analysis.
A re-evaluation was periormed to more-accurately model-the fatiguo cycles.
A graphical comparison of the two analysos is given by Figures B and C (pages 4 and 5 of this attachment).
Based on the more accurato analysis, 6,739 stratification and 375,407 striping cycles would occur.
The ponotration was successfully qualified using the fatiguo cycles from this more accurato analysis.
Subsystem 1RH-34 was not roanalyzed using the more accurate fatiguo analysis since the results using the original analysis were acceptable as is.
i
......~.
_._..u__._..._.__.____.._.,.;_.,
_.. ~
to U-60195fl l'a g e 3 of S e
l'e ne t ra tion
._ Inolation Valve Inolation Valve 11:12r009 11:121'000
/
- >4 j
low I'ronc.uro V
A A
liigh l>ressure L2 N
V
.L) 1/.>c);od Open SUllSYSTEM Li L2 ITEMAIUG 1Ril-34 1/2'
~ 16' L1 i L2 <2SD NOTI::
L1 10 a llorizontal Ilun.
L2 in a Vortical Run.
Thin cano han boon analyzed for thermal stratification and found to be acceptable.
Figuro A to U-6019';H Page 4 of S o
I original Analyuin i
Cycle in annumed to renuma immediat.ely o
-+c 0---
O
W t
cycle repeatn itself every 24 houra
/
Temperature difference between top (hot) portion of pipe and the bottom (cold) portion atabilizen.
Flow in accumed to otop.
Fluid flown, thermal stratification / striping occur, pipe heata up.
Leak begina.
Piping is subjected to continuoun fatigue cycles.
Cycle restarto every 24 houra.
Figure 11
to U-601(JS8 Page D of 5 o
More Accurate An11yain Imakage
- begins, cycle reGumeG.
o a
c, o
A,, Q-
.V Water & Valve Dick Begin To cool Next cycle Valve disk neata due to thermal expanaion of disk.
-Fluid flown, thermal stratification /otriping occur, valve dick heatn up.
Ioak Begins.
Piping experiences thermal fatigue cyclen induced by thermal atriping for a 7-hour duration.
A 45-hour cooldown occura.
During this time period the piping doea not experience fatigue cycles induced by thermal striping.
The 7-hour cycle restarta after cooldown.
Figure C
.