ML20148D375
| ML20148D375 | |
| Person / Time | |
|---|---|
| Issue date: | 07/11/1980 |
| From: | Harold Denton Office of Nuclear Reactor Regulation |
| To: | Michelson C NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD) |
| References | |
| NUDOCS 8009160606 | |
| Download: ML20148D375 (5) | |
Text
'^
I)TSTRIBUTION u
Central File NRC POR Ann Ferguson HRR #4059 JUL 111980 Terry Shultze 0EEB Rreadi_ng MEMORANDUM FOR:
Carl Michelson Director Office for Analysis and Evaluation of Operational Data FROM:
Narold R. Ornton. Director Office of Nuclear Reactor Regulation SUBKCT:
BWR KT PUMP INTEGRITT This reply is in response to your memorandum of 'BWR Jet Pump Integrity,"
dated May 23, 1980. You made three recannendations for NRR consideration, namely:
1.
Replacement of all jet pisap held down beams with an improved design beginning no later than Spring.1981.
2.
Evaluation of the potential for, and magnitude o#, water hammer type leads on jet pump integrity.
3.
Evaluation of the potential for damaging jet pump vibration and fatigue failure during the initial LPCI injection or subsequent long-term cooling modes.
We will respond te each of these in ture below.
I.
Recommendation For Rep 1weement:
With regeril to the first recommendation, we agree that the held down beam bars (NDB8's) shouS8 be replaced and are considering three alternate schedules the shortest of which would conform to y6ur recommendation of the Spring of 1981.
Before further addressing the schedule question it is useful to understand the work that has been gefing on in this area since february,1980 Subse-quent to the MDB8 fai!vre at Dresden Unit No. 3.' the BWR licrisees have been performing UT examinations of the N088's during refueling outages and have been replacing these beams showing crack indications. In response to I & E Bulletin 90-07 the licensees have provided written justification for continued operation until such time.that these inspections are made. In addition. the ! 8 E Bulletin identified specific in-service inspection /sur veillance procedures te be fallowed to provide early indication of jet pump degradation or failure. MRR ar4 i & E are coordinating their efforts to review the licensees' responses under a Task Interface Agreement (ROI 80 issued on March 19,1g80. This Task includes in part the technical review of I
proposed design changes, and the review of the implamentation schedule and installation procedures.
8009160 h [
t a
JUL 111980
- v. n..... w..
The detensination of a realistic schedule for installing the modified !!OB8 design in all operating BWR's should consider the need to complete the generic tasks of design analysis. redesign, corrosion testing and test evaluation, and the plant spectfic tasks of licensing review and approval, material procurement, manufacturing, delivery, and insta111.tfon. The pre-liminary results of the Dresden 3 analysis and evaluation were received on June 20,1980. Since the licensee and GE at still in their preliminary analysis and redesign phase, other licenses w itments to a schedule requiring installation of the HMS design t.
nning in the Spring of 1981 is probably optimistic. Futhermore, we do not believe the problem is serious anomah to warrant a special shutdown to replace the Hm8's and thus are inclined to allow BWR licensees to install the new HD6B's at the first scheduled refueling outage following completion of the required ifcensee and staff activities noted above.
An alternate schedule that is also being considered involves initial' replace-ment of all HOBB's in BWR/3' plants with existing spare BWR/4 HOBB hardware.
It has been observed that only the older BW/3 plants which have smaller HDEB's have exhibited crack indications from the UT inspections subsequent
_to,the Dresden HOBB failure. The combination of a smaller WR/3 HDB8 design and the longer operating time indicates that the WR/3 plants should be~given first consideration in the replacement '-heduling. Since the existing BWR/4 HDBB hardware is interchangeable t the existing BWR/3 HDGB design, no major mechanical replacement probl*
Id be anticipated.
However, if we use this approach, it would require a second replacment of the HOBB's in the BWR/3 plants with the new HOB 9 design when it becomes available. The major disadvantage with this dual replacement approach is the double radiation exposure for the work force involved in dual replace-ment operations. Should a timely schedule be established with respect to replacing the existing HD88's with the new HDB8, the.less destreabic double replacement program involving double radiation; exposures could perhaps be eliminated.
~ ~,~
A third alternate under consideration would be to work with the licensees and GE to establish a minimum reslistic HD88 replacement schedule consistant with a maxinva engineering and construction effort. Thus this third alterna-tive would be an all out crash effort as compared to the first alternative.
If it is determined that such a schedule could be reduced to the same approxisate time frame of the other two alternatives discussed above, the staff will pecbebly receamend going this way.
The staff plans to resolve the schedule requirements within sixty days of receipt of the final Oresden/GE analysis and evaluation report.
II. Evaluation of Water Hammer Type t.oadst Your specific recausendations regarding water hasser and vapor condensation induced vibratory loads, recommendations 2 and 3. have not been examined in the past, although we have. worked on similar generic and plant specific prob-lens. We agree that the potential for water haemer always exists in systems l
\\
. - - _ - - _ _ _ - _ _ _ _ _ _ _ _ _ _ _ - _ _ _ - - -. _ _. - - - _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ ~
JUL 11 Gyg l
C. Michelson lI I
In such cases' ll in which physical contact between the phases can occur.
l steam entrapment can be postulated with subsequent rapid condensation.
3 MRR has recognized the potential safety significance of water hamper irrences in nuclear plants and has developed a dual approach to dealing
, this problem. The first approach that has been applied is to establish ti.s conditions that have led to water hammer in reactors and to require This approach has measures to prevent the recurrence of these conditions.
j i'
been taken with the Westinghouse steam generator water hanner problem in which bottom discharge spargers were replaced in loany plants with top discharge spargers (J-tubes) which had the effect of delaying the inter-action of steen and auxiliary feed flow, thereby reducing the incidence f
of water hansier.
J The second approach taken by MRR to deal with the water hasser phenomena This has been to establish a specini generic task group under TAP A-1.
task group has reviewed a considerable amount of information on this long Work on this activity has been standing and highly complex subject.
under way for over two years and has reached va+ s stages of completion.
The development of a calculational method by w" to establish estimetas of the magnitudes of water hommer loads, as n nded in your memorandum, is presently part of this generic effort. Her it is to be recognized that a comprehensive calculational tool is wa
.de the prm ent scope of the NRR effort and realistic calculations ts
,1 mate the :nagnitude cannot be accomplished with the present state-of-che-art. This limitation of analytical capability tends to result in the use of ;:onservative or v
bounding calculations which are so overwhelming to real systems as'to provide no realistic basis on which to evaluate the safety significance or on which to propese design changes.
Yourrecommendationandconcernregardingwaterhasuserinsiiijetpumps will be fontarded to the TAP A-1 special task group for their consideration j
l in the review and evaluation of this generic issue.
III. Evaluation of Vibration and Fattwe Failures:
1 Recommendation 3, of year memo, addressed the need for aa. evaluation of the and fatigue potential for damage to the jet pumps resulting from ' vibration dering both the short and long tone core cooling periods ~ fb11owing a Loss-Of-Coolant-Accident (LOCA). During these time periods the jet pumps may be As a result i
surrounded by a steam / vapor environment in the annulus region.
.b' discussions with your staff (5. Rubin), thra separete pheno vibrations: 1) induced vibration by forward 'LPct flow in the unbroken loon.
induced vibration by backward LPCI flow in the broken loop and, 3) jet liu)sp excitation (respor.se) reselting free collapse of entreined vapor.
2 In your memorandum you referred to past WR single loop operation and jet pump
^
.. w
r M 11 M80 C. Michelson i
We agree testing programs in which high amplitude vibratiens were noted.
with your statement that the conditions experienced.during single loop.
[
operation and jet pump tests may be similar to those that would be esperi-D this res Mct this enced under the LPCI conditions _postunited above.
experience may provide useful,information regarding the potential for
~
vibration and fatigue fafieres.
The Operating Experience Evaluation Branch and the Operating Reactor Branch No. 2 are preparing 1stters regnesting infomation from licensees This relative to these tests and available operating experience data.
infomation will be reviewed by the staff, to detemine if additional tests and/or calculational results on the vibration characteristics off the jet pumps are needed to complete our review.
In regard to the third phenomena involving steam entrainment and steam bubble collapse; if entrainment were to occur, it appears likely that the The subse-steam bubbles would be dispersed and extremely small in size.
quent collapse of bubbles would probably not have the sh As bhenomenonj but vibrations could be a result.
scale of the potential described above, no attenSt to quantify these water hammer effects can be perfomed at this time, but your recosusendation and conc Also the NRR water hammer generic task group will information request.
be alerted to your steam bubble collapse concern, and requested to consider this phenomena as part of their ongoing generic activity.
Saussarizing the above discussions, the staff believes that the Licensee of Dresden Unit 3, and GE have' responded to LER Mo. 80-4/03t-0, concernin jet pump failure, in an acceptable manner.
proposed by the Licensee and E provides adegwate assurance to pemit c l
operation of the plant while a replacement HDBB design is being persued. -Th staff has decided that MD88 will be replaced in all WR's.
l designs are developed and tssted, special inspections are be, ag performed on t
the plants and no plant will be allowed to operate with pgen HDB8's.
i
@uE requirements for the replace-The staff is considering three alternate sDetermination of the schedule requirements ment of the HDSB's.
on a systematic evaluation of the variables described herein, with majorScon emphasis on public and plant employee safety.
will be established to confben withis the minimum time req;;1rements consistent with a maaimum engineering and construction efferk, The selection of the schedule requirements should be not later than' sixty" days following receipt of the aforementioned Dresden/8E Report.
m
-_____._._.___mmm-.,,---
,..,.,,.m..,
m.
p vg,,,,,-,w,,,,,p-_,-w,g,,,,..--w,y-.
,__..,es,
s JUL II S3 C. Michelson To provide assurance that the NRR sctions described above are followed in a proset and efficient sanner, the< Project Manager for Dresden 3, R. Bevan, has been assigned to coordinate alt tasks described herein.
Dr.gis.at m.d g?
u ).E.Datag j Harold R. Denton, Director Office of !belear Peactor Aegulation ec:
D. Eisenhut
- 1. Vollmer D. Ross S. Manager R. Mattson M. Ernst C. Berlinger OEEB Members W. Mills J. Collins R. *aer K. Kneil V. Noonan T..%vak T. Ippolito E. Bevan W. MGroff R. Minogue V. Stallo R. Budaitz R. Fraiay
\\
R. R1 995 W. Dircks nrA (1 Y I
\\
?-
f carict).. CST:0EEBk [. DST: :3. C.f AD,
, [Di :k..).k NM. hr NRR..'
)'
! R. Riggs
.,!.C. Berlinger,M. Ernst
!R. Mattson
,! E. U s e iH.
cnton SURNAVE -.
17/^d80
{'.7/ 4E0.- -.
77
-[7
.,;../.7. /30
../gy30
.f./$80 l7/f/80......
i7 -.
3 NRC FORM 31819-766 NRCu 3240 D tJ.S. GOV E RNtJENT PR ** TING OFFICE: 1979 289 369 y.-...
.m..
- ~" - - -^- - ------ --
~ - o Harold R. Denton down component (s) can be designed, developed, and implemented.
In this regard, during a meeting with the staff on April 16, 1980, GE indicated its intentions to begin work on such a long-term fix.
However, GE's representatives iere not prepared to comit to a specific schedule for this work nor has GE formally done so to date. Since the current remedial licensee actions will still allow plant operation with an indeterminate number of potentially partially degrade 6 (cracked) hold down beams (with reliance on daily jet pump operability surveillance to detect impending jet pump failure), we recommeno that GE and/or the affected BWR licensees promptly be requested to submit and commit to a schedule for modifying the beam design to corre-t the identified shortcomings.
We further recomend that their schedule call for implementation beginning no later than the Spring,1981 refueling outages.
2.
AE00 recommends that' steps be taken to evaluate the potential for, and magnitude of, water hammer type loads on jet pump integrity.
We have found from our review that the potential for, and magnitude-of, water hamer type loads on the jet pump 1800 elbow (and hold down beam), which might occur during the initial phase of LPCI injection
(
(when the recirculation system piping may be filled with steam),
have not been considered. Such water bammer impact loads on the elbow (beam) may be momentarily higher than the blowdown and quasi steady-state hydrodynamic LPCI injection loads already considered in the jet pump design and beam f ailure analysis. A load pulse could be postulated to occur ar a result of the liquid water slug initially moving at a higher velocity in a steam filled piping than it would subsequently when the piping is completely filled with liquid. In this regard, plants still using LPCI loop selection logic (e.g., Dresden J) would be of greatest interest since they would involve the largest LPCI induced dynamic forces. Accord-ingly, we recomend that steps be taken to evaluate the potential for, and magnitude of, water hammer type loads on jet pump integrity.
3.
AE00 rec:ximends that appropriate steps be taken to evaluate the potential for damaging jet pump vibration and fatigue failure during the initial LPCI injection or subsequent long-term cooling modes.
Our general examination of jet pump cesign/ analysis and testing has resulted in the finding that no analyses or tests have been performed to determine the vibrational characteristics of BWR jet pumps during the initial LPCI injection or subsequent long-term cooling modes.
That is, no quantitative assessment of the susceptibility of ;fet pumps to camaging flow induced vitrations has been undertaken for jet pump operation in a steam or vapor surrounded environment, which would exist for larga recirculation line pipe breaks. Under such
~
2_
3-Harold R. Denton f
circumstances, the potential exists for vapor.entrainment and sub-sequent collapse of vapor bubbles causing excessive vibration and subsequent damage or. f ailure of one or more jet pumps while opera-
- ting in the initial or latter stages of emergency core cooling.
Additionally, backflow would occur in the idle loops which could result in damaging vibrations. Accordingly, we reconsnend that appropriate steps be taken to evaluate the potential for damaging jet pump vibration and fatigue failure for the aforementioned, conditions, since BWR operating experience has shown the tendency of jet pumps to experience higher amplitudes of vibration during off-normal (e.g., single loop operation) conditions.
We believe the above recommendations, which are intended to assure the integrity of BWR jet pumps, are consistent with NRC staff actions taken in connection with other similar operating experience prompted issues which involve LOCA potential and ECCS performance. Such issues include BWR Nozzle Cracking, BWR Core Spray Sparger Cracking, and PWR Steam Generatur Tube Integrity.
In the case of BWR jet pump plants with core spray sparger cracking',
for example, the staff has reviewed licensing analyses which give no credit for core spray (CS) heat transfer function even though
(
stress-deflection analyses support the conclusion that cracked spargers will retain their structural integrity during a.DBA.
- (.
Thus, even though structural analyses show the sparger would be expected to remain intact following a LOCA/ core spray injection, significant staff uncertainty exists as to its capability to per-
)
form its spray function as designed. This position has been taken 1
oespite CS system redundancy. Conversely, with regard 10 the jet pump hold down beam cracking problem, a ratto oflLOCA blowdown.
loads to normal loads also supports the conclusion that a cracked beam will withstand the blowdown without failure.
In this' case,.,
however, the staff has not required that one or more failed jet pumps be assumed for ECCS performance analysis purposes. This comparison is not intended to illustrate any inconsistency in i
staff actions which relate to the potential for LOCA/ECCS induced mechanical failures resulting in the loss of safety function of important safety equipment in the BWR system. The two are con-trasted 'instead in order.to substantiate our belief that there is a need to develop and document a comparatively higher degree of assurance than presently exists (via additional analysis / testing) that jet pumps will remain intact to perform their intended safety 1
c w
e-.
,,o-m,r--r-w es i.,
-w-vw,s,-e,--
w
-y6,-,.,-y----
3+--
m, e
,w,,+nm.
y
=
9 -
,[
e*
d jo UNITED STATES -
g NUCLEAR REGULATORY COMMISSION
[
,e g
WASMGTON, D. C. 20666 MAY Z 31980 g*..../
N MEMORAN'UM FOR:
Harold R. Denton, Director 0
Office of Nuclear Reactor Regulation FROM:
Carl Michelson, Director Office for Analysis and Evaluation cf Operational Data
SUBJECT:
BWR JET PUMP INTEGRITY Since the beginning of February of this year, NRR, IE, and AE00 have been investigating the causes and evaluating the consequences and imediate corrective actions required as a result of the jet pump f ailure at Dresden Unit 3, LER No. 80-4/03L-0, Docket No. 050-249. The failure was determined to be caused by progressive intergranular stress corrosion cracking of.;e jet pump Inconel X-750 hold down beam. The deepening crack eventual b ied to hydrodynamically induced overloading and breakage of the remaining liga-ment at the root of the crack. This recent Dresden 3 experience, tocether with other previous jet pump integrity-related problems at the Dresden and I
Quad Cities units, has prompted AE00 to also review, in more gene al terms,
(
the SWR jet pump mechanical design and its related design basis. Our. review examined the DBA-LOCA blowdown and ECCS mechanical loadings which have been considered by GE and evaluated by the staff. The AE00 analysis anc evaluation of these operating experiences and design bases, thus far, have resulted in the formulation of the following recommendations which we are now asking you to consider. Additional experience may give rise to further recommendations.
1.
AE00 recomends that GE and/or the affected BWR licensees be requested to submit and comit to a schedule for replacement of all hold down beams with an improved design, and that said schedule call for implementation beginning no later than the. pririg,1981 refueling outages.
5 To date, for the most part, vendor, licensee, and NRC staff discus-siens, evaluations, and actions have been appropriately directed toward the imediate safe operation of currently operating BWR jet pump plants. We agree that the specific concerns and issues that were addressed are consistent with current regulatory needs and requirements. Furthermore, we agree that the actions being taken by licensees provide reasonable assurance of safe plant operation in the short-term. At the same time, however, AE00 considers the jet pump integrity analyses and evaluations that were performed and the actions taken to be adequate for only a relatively brief interim period until such a time when an improved jet pump hold
~~
c,
+
Harold R. Denton.
functions if such credit is to be given in the long-tenn. The need for this assurance appears to be especially true in the case of BWR jet pumps which potentially could be susceptible to connon l
cause/ common mode failure mechanisms such as has been discussed 4
above.
/
1 6
Carl Michelson, Director i
Office for Analysis and Evaluation.
of ( serational Data cc:
C. Berlinger E. Brown D. Eisenhut T. Ippolito E. Jordan
(
R. Mattson S. Rubin
(_
D. Vollmer i
i 1
Du.