ML20207A115
Text
__
- ~_--
UNITED STATES
[
\\
C
,[ '
,[;
NUCLEAR REGULATORY COMMISS'ON
{
e W ASHis 'OTON. 0 C. 20455 s
1
4.,,,, /
8 liG6 i
MEMORANDUM FOR:
Gary M. Holahan. Director t
Operating Reactors Assessment Staff t
FROM:
Gus C. Lainas. Assistant Director Division of BWR Licensing l
Sl,' EJECT :
ANOMALOUS RECIRCULATION LOOF FLOW BEHAVIOR Ih JET PUMP BWP FLANTS 1.0 thTF000CT10F I
i In early 1985 unexpected fluctuations in indicated recirculstion loop flow rate were experienced at the Pilgrim huclear Ocwer Station (PNDS) following sn outage to replace the recirculation system piping. With core power and flow rates near rated conditions, the indicated drive flow rate in recirculation i
loop "A" would suddenly incresse about 2 percent, remain at the higher value for a short tire and then return to the original value. The occurrence of this event was randor with an average occerrence rate of about 3 events per hour. The curation of the increase ranged from a few seconds to about ten minutes. An average of about 12 percent of the time was spent at the higher flow rate.
For the "B" loop, the amplitude of the fluctuations was about the same out there were only about thres events in three hours with a maximum high
~
flow duration of about nc minute. The high flow rates in the "B" loop occurred about one,wrant of the tire.
These fluctuations were accompanied l
by fluctuations in core flw rate APRM readings and individual jet pump i
differential pressures. The 2 percent increase in loop "A" flow resulted in t
abect a 1 percent increase in core power.
For the low probability occurrence l
of simultaneous increase in both "A" and "B" loop flev rate, the core power increase would be about 2 percent I
In a repert to Boston Edison (Ref. 1) General Electric attributed the occurrence of these fluctuations in the recirculation loops to a bistable flow pattern at the header cross in the recirculation loop discharge piping. The flow pattern producing higher piping losses and, hence, lower flow rates i
involved formation of a spiral flow in the arts of the circular header. This pattern resulted in a measured increase in the flow rates of the jet purps fed by the header inriediately above the cross which was more then offset by reduction in flew rate of the other jet pumps, General Electric noted that this behavter was found in subscale laboratory tests.
In Reference 1. General i
Electric also provided an evaluation of the safety significance of this change in plant behavior. They concluded that the short duration 2 percent power increase was insignificant and that the 1 percent power increase from the loop i
"/." flow fluctuation was within it.e 2 porcent power allowance used in the
Contact:
C. Graves RSB, x29461
( pqcD 90l&Mf uw r/sY
t 2
safety analyses.
It was also noted by GE that the NRC in an August 1980 memo (Ref. 2) recognized and accepted a 2 percent variation in power. The General Electric Company report was reviewed by INEL personnel and Dr. Pater Griffith of PIT under an NRE technical assistance contract. A draft of the INEL
- r. valuation which essentially agrees with the GE evaluation is currently being reviewed by the staft.
Fecently, two other cates of anomalous behavier of recirculation loop flow rates, similar to that found at Pilcrim were encountered. At Vermont iankee, where replacement of recirculation loop, piping was completed recently, the licensee rc.tified the resident inspector on August 11, 1986, that the plant was experiencing recirculation loop flow oscillations. The preliminary test results indicated about 1 percert increases in icop "A" flow rates (Ref. 3),
ho fluctuations were found for ;oop "S".
At NUCLENOR, fluctuations of about 3 percent were feurd in their "B' recirculetinn loop (Ref, 4). At Vermont Yankee there vere major repla',ements of the recirculation system piping, which included changes to the design of seme piping compo*ents. At NUCLENCR, there was partial replacement of the pir' 1 system.
The anortlous recirculation loop behavior apparently can result frcm recirculatier riping system replacements of various types which may also be made at other SWP plants as a countermeasure to the intergranular stress t
corrosien cracking. (IGSCC) problem. Hence, a brief review was initiated to detemine if the flow anomaly wa encountered at other plants. This mercrardum reports on the current status of the review and presents recoverdations based on the limited data available at this time.
0.0 DISCUSSION 2.1 IGSCC Froblem and Countermeasures In 1951, intergrarular stress corrosion cracking (IGSCC) was found near r
weldretts in large diameter piping of the recirculation systems of some BWRs t
(Ref. 5).
Extensive inspect!cn of BWR piping systems since that time
- ndicated wide variations in the anount of IGSCC, ranging from minor cracking tc significant degradation of piping. As discussed in Reference 5.
counterreasures to IGSCC include replacenent of affected piping with more resistant material, use of improved primary coolant chemistry procedures and reduction et tensile stresses en inside piping surfaces near welds, in some plants, cracked weldments are reinforced by weld everlays. A number of plants have replaced or plan to replace part or nearly all of the recirculation system. The replacements may range from a simple replacement of local sections end components with no major change in design or involve new compenents and extensive replacement. The piping replacements could result in anov.alous recirculation loop flow behavior such as that which occurred at Pilgrim In this regard, it is noted that the present understanding of the 4
mechanisms involved does not permit prediction of the occurrence of the phenomenon or of the magnitudes and duration of the fluctuations resulting from the piping system changes.
I
\\
I 1
r'.
3 2.2 Pilgrim Event 2.2.1 Description of Event (Ref. 1)
~
At Pilgrim, the major charee occurred to recirculation loop "A" which exhibited sudden increase: ili flow rate of about 2 percent of the rated value. The event occurrence was random in time with an average of about three ever*.! per hour. The duration of the flow increase ranged from a few seconds to. ninutes.
For loop "B". the magnitude of the flow rate increase was also about 2 percent. The evert occurrence was also random, but with an averane rate of about one event per hcur and durations at high flow rates of less than 1 minute. The tire spent at the higher flow rates was about 12 percent for loop "A", but less than 1 percent 'or loop "B".
The increases in loop flow rates were accompanied by increase. in ccre flow rate and APRM readings.
Changes to the pressure drop for irdividual jet pumps were also observed.
However, the direction und magnitude of the changes for individual jet purrps varied with riser location, as noted below in the discussion of the mechanism fcr the fluctuations.
The 2 percent increase in ficw rate for one loop re:ulteo in about a 1 percent increase in reactor power.
2.2.2 Proposed Mechanism for Recirculation Loop Fluctuations In Referencu 1. General Electric proposed the existence of a bistable ficw pattern, criginating at the 28 inch by 22 inch header cross in the rec' culation pump discharge piping, as the source of the flow oscillations.
The low ficw rates were associated with the presence of a swirling flow in the i
i header arms.
For this pattern, there was an increase in flow rate for the 1
riser cirectly abnve the cross which was more than offset by the reduction in flow rate throuch the other risers as the result of the increased pressure loss in the header arins.
The individual jet purrp pressure drops followed this behavior. General Electric also noted that subscale laboratory tests exhibited this behavior.
In his review of the General Electric report. Dr.
Griffith concluded that two almost equally likely flow regibes could be i
present and that the systar could switch spontareously from one regime to the
[
other. He suggested tha: the initiating cause could be a particularly large turbulent fluctuation.
It is noted that the proposed rer.hanism is a qualitative explanation of the i
behavior at Pilgrim. There i: r.ot basis for predicting whether or not the occurrerce of the fluctuations will r:n 't from a particular piping change.
There is also no basis for predicting U e frequency or duration of the flew i
changes.
The above references do not m :uss the bounds on the magnitude of the flow changes.
2.2.3 GE Evaluation of Safe +v Significance of the Pilgrim Event i
Reference 1 provides a General Electric Co.apany evaluation of the safety i
significance of the recirculation loop flow anomaly at Pilgrin.
St. ice the
]
l
4 safety analyses include a 2 percent allowance for power uncertainty, they concluded that the 1 percent increases in pcwer resulting from the flow increase in the "A" loop were not significant. The 2 percent power increase resulting from simultaneous increase in flow rate of both loops was also considered as not significant as the result of the very low probability of cccurrence. The effects of the fluctuations on piping vibration, fatigue usage of vessel and piping components and flow and APRM calibration were also stuc'ied and found acceptable.
The concluded that ne Technical Specificatiens would have tn changs, but recorrended some changes to operating procedures to l
avoid viol nion of w e Technical Specifications.
2.3 Other Events
'!anhe was notified by On August 11. 1986, the resident inspector at Verm 6
the lictnsee that the plant was exhibiting a bista.e flow anomaly similar to thet previously found at Pilgrim. A major replacecent of recirculation system Jiping had been conrleted during the previous plant outage.
Preliminary results provided in Reference 3 indicate that the "A"
recirculation loop exhibited flow increases of about 1 percent, with associated increases of about 1/3 percent in Core flow rate and 1 percent in pow 6r.
No fluctuations in loop "B" flow rrte we*e observed. The data of Reference 3 suggest that loop "A" is operating in the high flow mode roughly 30 percent of the time.
Reference 4 provides some rcsults of anomalous recirculation loop flow betavior at NUCLEh0R. Recirculation loop "B" exhibits flow increases of about 3 percent.
Fluctuations were not observed in loop "A".
The magnitude of the flow fluctuatier and frequency of or rrrence decreased with decrease in loop flow rate.
I From telepncne conversations with General Electric it was learned that I
anomalous behavior of the recirculation loop flow rate was also observed at Leibstact, a bhp /6. Fluctuations occurre1 in both loops. The fraction of the time spent in the high flow mode was about the sare in each loop. Since Leibstadt is a new plant, the anomalous flow behavior did not result from piping cplacerent.
l 3.0 C0hCLUSIONS AND RECO M NnATIONS a)
Tbt anomalous recirculation loop flow behavior that occurred at Pilgrim l
Nuclear Power Station after an outage to replace the recirculation system piping has been observed at several other BWR jet pump plants (Vertrent l
Yankee,NUCLEh0R).
1 4
b)
The bistable flow pattern suggested by General tilectric Company as the mechanism for the observed fluctuations in recirculation system flow 4
]
rates and jet pump pressure drops is plav'ble. However, current understanding and availability of plant intorination does not permi*
prediction of occurrence of anomalous behavior following a given r aing i
system change or estimates of the magnitude of the fluctuations anc average fraction of time spent in the high flow mode. The limited data i
available suggests significant differences in these values.
1
e O
0 5
c)
The General Electric Company evaluation indicates that the anomalous behavior at Pilgrim is acceptable.
However, the results for Pilgrim cannot be generalized : predict the behavior and acceptability of flow fluctuations at other plants.
d)
It is recommended that ar information notice be de', eloped gnd forwarded to EWR licensees advising of the above and recommehding that any iaentified flew ancmalies be appropriately reviewed.
4.0 REFERENCES
I 1.
"Safety Evaluation of a Retirculatier. System Flow Anemaly at Pilcrim l
Statier.", General Electric Ccepany Report MDE 470385,DRFA00-0I713 l
Revision 1, November 1985.
O.
Perorandum for E. J. Brunner et. al. from E. L. Jordan, "Discussion of l
Licensed Power Level" (AITS 14580H2), August 22, 1980.
3.
TELEX from W. Fayrond and T. Silko, Vermont Yankee Resident Inspector Office to H. Abelson, NRR, August 13, 1986.
l 4
Perorandun for J. Funches NRR from H. Faulkrer, IP, "Reports free Spain i
on the Occurrence of APRW.s Instebility and Flow Fluctuations at 6 EWR,"
l July 29, 1986.
5.
"Praft Revision to NUPEG and an Implementing Gereric Letter " R. M.
Eernero, July 1966.
/(}
l us C. Lainas, Assistant Directer Division of BWR Licensing CC:
R. Berecio i
l W. Houston P. Leech M V. Rooney E. Jordan L. Phillips l
l l
1 l
l