ML20040A397
| ML20040A397 | |
| Person / Time | |
|---|---|
| Site: | San Onofre  |
| Issue date: | 03/12/1981 |
| From: | Weeks J BROOKHAVEN NATIONAL LABORATORY |
| To: | Pawlicki S Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML17250A806 | List: |
| References | |
| FOIA-81-313 NUDOCS 8201210018 | |
| Download: ML20040A397 (5) | |
Text
"
I'l:0PRILI ARY 1lJ0l;M,",110*:
./
/,',.
'?.!! l > UI.', i iloill! '>. li.
di 2617 i 3 e-I.. n i~ r,t v o u: e a ir.,r.,y March 12, 1981 Mr. S. S. Pawlicki, Chief Materials Engineering Branch Office of Nuclear Reactor Regulation U. S. Nuclear Regulatory Commission i
Washington, D. C.
20555
Subject:
San Onofre Tube Sleeving Program:
Summary of Discus-sions at Westinghouse Forest Hills on March 6,1981
Dear Steve:
By telephone on February 28, you asked me to visit Westinghouse at Forest
._Hills to discuss the tube sleeving program proposed for San Onofre, with par-ticular emphasis on what is known about the leakage rates of the so-called limited leakage sleeves.
As you know, these have been proposed for use in l
areas of the steam generator where sleeves could not be brazed because of j
problems in brazing in the sludge pile area.
t M.
On Friday, March 6, I met with R. T. Begley, A. Klein, L. Nelson and R. g.
Wilson to discuss this sleeving program.
Much of the information discussed I believe to be Westinghouse proprietary information, and for that reason I have stamped this letter " Proprietary Information" until it is cleared by Westing-house.
During this meeting, I received a copy of the Westinghouse (Proprietary Class 2) report HS-MFSE-81-054, " Technical Evaluation Report for Hybrid Sleeve", dated March 1981.
I was advised that this report had been officially transmitted to the NRC.
It served as the basis of our discussions.
As you know, the extensive corrosion that has occurred at the top of the tube sheet in the San Onofre steam generators has led to a program to insert sleeves in these tubes in order to maintain the integrity of the primary cool-ant pressure boundary.
This approach was deemed necessary because of the wide-spread nature of the degradation and the fact that its location at the top of the tube sheet is such as Jgkg.,ijs identification by inservice inspection very dif ficult unless the w.st3 = more than halfway through the tubing.
The original plan was to sleeve all the tubes in a wide area of the steam I
generator, as described in the Westinghouse proprietary steam generator repair I
report of September 1980.
Because of curvature of the primary header, tne l
length of the sleeves that can be inserted in the tubes varies with pMf 4 ion in the steam generator, and only near the central part (where the longest sleeves 820121001e eig009 PDR FOIA UDELL81-313 PDR m
n
i Mr. 5. 5. Pawlicki March 12,1981 l
Page Two I
I could be inserted) has it been rossibt etto braze these sleeves in the area j
above the sludge pile, as described in "ciiat September report.
The dif ficulty in brazing the sleeves, where sludge is piled up against the secondary side of the tubes, presumably occurs because of extensive heat transfer from moisture trapped in the sludge.
This lef t the utility with four options:
l t
1.
Braze sleeves only where possible out of sludge, 4
2.
use a mechanical " hybrid" sleeve in the sludge area, 3.
plug all affected tubes, possibly with removable plugs, for sub-sequent sleeving, or 1
4.
not plug or sleeve, based on pancake probe data.
I The utility is now pulling a few tubes from the affected area of the steam gen-l orator that passed the pancake prebe inspection, to obtain a better basis for l
__s21ecting this last option.
j i
However, the present plan for areas where brazing cannot be performed out l
of the sludge pile is to use what Westinghouse now refers to as a Typically, where the sleeve is hydraulically expanded against the tube, the design load is such as to cause a small permanent yield in the tube wall, of a maximum of the i
order of 6 to 10 mils.
(To expand the tube wall to the yield stress requires i
approximately 1-1/2 mils increase in diameter.)
I received copies of figures showing average DPH values for the tube in the hydraulically expanded area and the hard-rolled area.
Typically, the hard-rolled area has no more than 12 additional DPH numbers (500 g load) over and above the basic tube prior to the expansion.
The sleeve, of course, is more extensively worked, but it is therrally treated to minimize stress corrosion cracking.
j Westin house has run a series of experiments to determine the leak rate of sleeves at room temperature and at high temperature, both as produced and af ter an axial fatigue test that causes permanent deformation in
Mr. 5. S. Pawlicki March 12, 1981 Page ihree the tubing.
This fatigue. test simulates het Mp and cooldown stresses, and the degradation that might occur if the tube should break in the area covered by
_the original sleeve, i.e., the area where the intergranular attack is occur-During a design basis accident, such as a steam line break, the vertical ring.
force on the remaining tube would be great enough that it would attempt to loosen the sleeve.
The forces required to 1onsen the tube from the sleeve have been measured and are shown in the above cited proprietary report to be well in excess of the forces estimated during a design basis accident without signif-icant increase in leaking rates.
Typically, yield loads are 2,000 to 3,000 lbs. axially, whereas an SLB would produce a tensile load of 800 lbs., accord-ing to Uestinghouse.
The leakage rates given in the proprietary report are all in drops per minute; the conversion factor is typically 70,000 drops equals one Typical leak rates are in the order of 1 to 3 drops per minute, even gallon.
This af ter the tube was pulled sufficiently to deform the original joint.
suggests that 1 to 2,000 tubes could break in the corroded area during a DBA The without the leak rate from this breakage exceeding 1 gallon per minute.
tubes in the test were given both cyclic loads to determine the effects of plant on and of f cycles, as well as the yield pull loads to determine their
' behavior during a desip basis accident, and in neither condition did the leakage rates become excessive or even significant at pressures up to 2400 psi.
Two aspects of the testing program were, however, not performed exten-sively.
The first has to do with the effects of thermal cycling (as opposed to mechanical cycling) of the tube a number of times to simulate startups and shutdowns.
The second was that relatively few leak rates were measured at 600*F and the 2400 psi pressure differential.
However, from the room tem-perature data, the effect of pressure differential is not apparent and cer-tainly not large (less than a factor of 2), and from the high temperature data given, the ef fect of temperature on the leakage rate for the given joint is also less than a factor of 2.
If all sleeve tubes should break during a MSLBA, we could tolerate N 240 drops / minute / tube, or N 100 times the measured c
leakage / tube.
Consequently, I believ? the tube sleeving program at San Onofre should be quite satisfactory in preventing major primary to secondary leakage should a number of the tubes break in the corroded area during a design basis accident such as a steam line break or a LOCA.
Because there is denting in the San Onofre reactor, it was pointed out that normal heatup and cooldown stresses on the area of the sleeve expansion joint may be compressive rather than ten-sile, forcing the sleeve tube together, not pulling it apart. The exact nature of these forces depends on the relative motions of the tubes and tube support plates.
It is only the steam line break accident in which large tensile stresses are present on the tube in the area where the corrosion occurs.
l
/
, ;t.
fir. 5. 5. Pawlicki
!! arch 12,1981 Page Four During our discussions, it was pointed ob" Mhat for a nunber of tubes at San Onofre where the braze was imperfectly made, they now intend to add I
expansion joint below the brazed joint, and still above the tube sheet, where possible, to provide additional assurance of tightness during a design basis accident. Figure 1 gives a sketch of expansion joint design, and Figure 2 shows the average DPH values for the sleeve and the tube through these joints as neasured in the laboratory.
Because of concerns over the role of residual stresses from this expansion in the tube causing primary side stress corrosion cracking of the tube at the point D' (where changes in the sleeve diameter may preclude routine inservice inspection of the condition of the tube), Westinghouse has taken great care to design the joint to produce the mininal residual stresses on the tube in this area.
[There has been primary side cracking in the tube-to-tube sheet expansion transition in at least two European plants.
In both of these, however, the Westinghouse understanding is that the ID of the tube sheet lioles is more than 20 mils greater than the tube OD (compared to 20 mils in this country), and therefore the amount of cold work going into the roll ex-pansion ir these crevices may have been significantly greater than in this country. J In any event, the hydraulic expansion, which gives the primary de-fonnation of both the tube and the sleeve, is performed in such a way as to minimize the cold work at any one point in the expansion (as revealed by the DPH values in Figure 1), and therefore Westinghouse believes, and I concur, there is a minimum probability that stress corrosion cracking will occur at point D'.
Care should be taken, however, that residual fluxes from aborted at-tempts at brazing the sleeves are removed prior to installation of a sleeve, wherever possible.
At the same time, however, there is a continuing need for quantitative information on the relationship between cold work and stress in the tubing and the time required to nucleate a crack in service.
Based on these conclusions, I concur that the plant can be safely returned to service with for the following reasons:
1.
They will prevent double-ended rupture of the tubes during a DBA from continued progress of the IGA at the top of the tube sheet by re-straining the tubes.
2.
Suf ficient data have been obtained to indicate that leakage will be controlled to negligible levels by the use of these sleeves.
3.
Care has been taken to minimize residual stresses in the tubes where the sleeves are rolled into the tubes, thus minimizing the probability of stress corrosion cracking developing in this area.
v-n
-+
f.,,,
f*r. 5. 5. Pawlicki 11 arch 12,1981 Page five
,J
[
] hope this gives you the infor
- . ion you requested; I will be happy to meet with you to discuss the situation.urther.
j Sinc 1
- ours, i
)
]hbW John Ti. Weeks Lerad r Corrosion Science Group i
JRW/dt cc:
W. Y. Kato R. M. Gamble j
E. L. Murphy j
San Onofre Files 0
l l
1 l
l l
l f
.,...,,--,.-c-e,
,.m-.-__m.,-
w
,...-..'-F
..u-.i PC O
-