ML20236E408
| ML20236E408 | |
| Person / Time | |
|---|---|
| Issue date: | 07/15/1986 |
| From: | Shao L, Vollmer R NRC - PIPING REVIEW COMMITTEE |
| To: | Roller W DETROIT EDISON CO. |
| Shared Package | |
| ML20236E258 | List: |
| References | |
| FOIA-87-666 NUDOCS 8710290210 | |
| Download: ML20236E408 (9) | |
Text
-
"6.
d y
ok._.
f bM RP.!'
M 1 & 1966 Mr. Williwr. R. Ruller Plant Superintendent Monroe Power Plant Production Organization P. O. Box 731 Monroe, Michigan 48161
Dear Mr. Roller:
During our visit to the Monroe Power Plant to review the Unit I reheat pipe rupture you agreed to review our report for accuracy and for any proprietary inforination that should be omitted.
Enclosed is a copy of a draft report that we plan to issue as
- Official Use Only" information. Please let us have your connents on the subject.
l As you may know, the NRC Piping Review Comittee has been reevaluating piping requirements for nuclear plants in light of new inforination and operating-experience. Our aim is to establish a more rational, while still safe, set of piping requirements. In this reevaluation the comittee found that tome of the existing requirements have led to unnecessary and expensive piping appurtenances making the piping difficult to inspect and, in some cases, more prone to failure. These requirements are being modified to take advantage of our improved understanding of the subject.
Also, as part of our ongoing reevaluation of piping information, our research staff has been studying Tailure modes in piping including cracked and degraded piping. This research will provide a data base and basis for evaluating the leak-before-break concept which, if applied, would permit the elimination of massive pipe whip restraints. We are also evaluating criteria for extending life of nuclear plants which include the effect of aging on plant materials.
It's because of these concerns that we have been following with great interest the information on recent pipe failures in fossile power plants.
Especially in those cases where rapid breaks have rasulted. Although our preliminary review has concluded that these failures are probably creep related and would not be expected to occur in nuclear plants which operate at temperatures too low for creep, we have not been able to obtain actual data to support these' conclusions.
i:
10290210 871020 L
W. R. Rollar g I
4 To confirm our assumptions, we would very much like to have one of our contractors, the Oak Ridge National Laboratory (ORNL), so an indepth evalu-ation of a piece of pipe that has been removed from a pipe failure, such as the Monroc Unit I reheat line. This evaluation would include sophisticated metallogr6phy and mechanical testing.
G. M. Slaughter and R. Swindemen, the principal investigators have considerable experience ir this field and are well known to your materials people.
If you could make a suitable piece of material approximately 10 inches square from the longitudinal weld region of the Unit I reheat pipe available to us to investigate, we would be pleased to make all of the data and results available to you. Further, since our only interest is in confirming that the failure mode is not applicable to nuclear plants we do not plan to publish the data and information of a proprietary or company confidential nature would be honored in these studies.
Please let us know if Detroit Edison would be willing to supply a section of material for our study. We would be happy to develop a cooperative plan of action for your review, if necessary, on the detailed ucrk to be done and
~
disposition of the subsequent results.
Sincerely, Driginal Signed By:
Richard H. yollmer Richard H. Vollmer, Co-Chairman NRC Piping Review Courittee Lawrence C. Shao, CcwChairman NRC Piping Review Camr.ittee Enclos are: As stated D1STRIhtnION:
RES Files ATaboada LShao RVollme'.
NRC PRC files Disk: SHA0 i
Title:
roller NR PRC NR H
.C
, 1mer 6
ada:cr LS au R ro 6/ (/86
};7/86 6
/86 f
L
n-_--_____-
[ : g, i.-
's j
li i
taboada(SHAO) 6/13/66-Draft 3
SUBJECT:
DETROIT EDISON'S M0hRGE POWER PLANT STEAM LlhE RUPTURE
'On January 30,1986, at 8:07 am, a large steam line ruptured at Detroit i
Edison's Monroe Power Plant injurying 17 of the plant personnel. An NRC team e S P' -
visited the plant on February 12 1966 to review the incident, observe the pipe failure and, ascertain any information,that might be applicable.to safe
- (y,/ gf,y) '
operation of nuclear plants. Enclosure 1 has a list of the attendees.
\\
The Detroit Edisun facility, located in Monroe, Michigan, is a 3000 Mwe, coal fired plant with four turbine-generator units, each rated at 750 Mwe, and has steam conditions at 1000'F and 3600 psi. The plant, which began comercial operations in June 1971, operates in a base loaded condition, and provides approximately 50% of Detroit Edison's capacity.
[VThefailureoccurredinahotreheatlineofUnit#1whichhadoperatedfor 97000 hoursf $n$t il has a Babcock and kilcox, once-through, supercritical boiler and a General Electric turbine and was producing 720 Mwe at steady state operating conditions at the time of the failure.
f Several minutes before the failure a steam leak was discovered coming through
{
the pipe insulation. Observers reported that before corrective actions could l
be taken the leak rate was observed to increase rapidly and culminate in an explosive pipe failure. The resulting shock tripped the turbine, toppled a block wall nearby and caused the building to shake for approximately 10 seconds. Four minor fires were started and the building was inundated with a cloud of steam, fly ash, coal dust and insulation particles. Although there were no serious injuries 17 people required treatment. Six people were admitted to the hospital for such proti1Ns'as cuts, bruises, broken' bones and.
respiratory problems. All of the injured people were back to work at the tiae of our visit.
l
i
)
Clean up of the facility to a condition that permitted nemal traffic took 11 weeks, primarily because of precautions required to remove the f
. asbestos containing insulation. Certification i. hat the area was free of
)
asbestos required analysis of over 500 samples. During all of this period the other three units at the Monroe plant remained on line.
pipe Description The failed pipe was a horizontal, 32-inch diameter,1.505-inch thick wall, hot, reheat line, designed to 1_002*F and 750 psi and operating at the time of y
failure at 1000*F, 730 psi steam conditior.s. The pipe was constructed by bending and seam welding El Cr-1Mo alloy steel plate to ASTM A155 (P-22)-1966 standards; manufactured by the Dravo Corporation and heat treated (annealed at 1700*F) by National Annealing Box Company. The seam weld was made by subarc welding from both sides using 4 or 5 passes. Radiography of the weld was required by the purchase specifications. However, the radiographic records are no longer available for review. Preliminary chemical analysis of pipe sections I
removed from the failed pipe confirfns that the material chemistry meets specifications except for minor variations; slightly highet manganese, and copper content in the weld metal.
Failure Description The pipe failure occurred in a 40 foot horizontal spool piece section at the edge of the seam weld located at the six o' clock position. The fracture was a
" fish mouth" rupture approximately 20 ft. long and approximately 6 ft, wide at its widest point (See Figure 1). The crack arrested at a circumferential weld at one end of the spool. piece. At the other end the cracks appeared to simp.yl run out of driving force and stopped in the normal pipe run. The force of the~
steam release' broke several pipe supports and put a substantial bend'in the 4%
pipe.
(SeeFigure2.)
The pipe appeared to have failed along one of the weld metal / base metal fusion A visual. inspection of the pipe rupture revealed that much of the weld l
rones.
metal was completely intact, and the outline of the weld beads was clearly j
visible on both fracture surfaces which were oxidized. Except for the last several feet at the ends of the crack, the fracture surf aces appeared to be brittle with a minimum of shear lip.
There was no evidence of transverse i
cracking, gress pitting or other corrosive attack. Detroit Edison
)
p[<*
reported that nonmetallic. inclusions 'were found in the weld area. A g
g
. preliminary microscopic examination by Detroit Edison disclosed that the '
1,',wg j,(
failurelinitiated in the center of -the pipe and grew to the outside' pipe $M ge/
surface before fciling.
l Several key sections of pipe, including the region of probable fracture initiation, were removed for evaluation before our visit. These specimens are j
being tested to verify chemical and mechanical properties and to aid in
. establishing the cause and mechanism of failure.
Hopefully, the results of
)T these examinations will be made available to the NRC and the public in time.#
M General Observations The pipe failure at Monroe had several similarities to the Mohave pipe failure j, p (SeeSECY-85-275). Both failures occurred explosively ir. large diameter (c 4,, gy.
horizontal reheat lines after long tenn operation at 100M and relatively
~
similar. pressures (730 psi and 580 psi).
Both failures occurred primarily along a seam weld and opened up fish mouth fractures that were approximately 20 feet long and 6 feet wide. Both failed pipes were of the same vintage, made to similar material specification and designed to ASME/ ANSI B31.1. However, the Mohave pipe was made of in Cr-i Mo material and the Monroe pipe was made of an improved 21 Cr-1Mo material.
g y w g g Q,,(ti,j/S b)J~,:lvre, h.
As a consequence of these failures the electric power industry has instituted extensive inspection programs on piping in fossil plants that operate at conditions similar to the Mohave and Monroe plants. Over 60 plants have been inspected and at least five units have or are planning to replace piping in steam reheat systems because of suspect conditions and/or metallurgical evidence of aging and initici creep damage.-
Despite all of these inspections and several industry conferences on the subject potentirPfailure mechanisms are not well defined since the specific cause of the Monroe and Mohave failures are not yet kngwn to the power industry or the technical comunity. No failure report or detcHed analysis information on either incident has been released by the utilities. Howe i
is generally assumed that the failures /were brought on by(e.
creep rupture in the weldment at locations containing manufacturing flaws. There is some evidence l
L
i.
1
-that the fusion zonf of welds are more prone to fail in this manner.
~
Other 4
failure mechanisms have been proposed.
Although the utilities have not' issued any information on the cause of failure,
+
an incident report on the Mohave pipe failure has been produced by the National &
.N Board of Boiler and Pressure Vessel Inspectors. They concludgthat the 6-I longitudinal-weld contained embedded cracks that had existed for a long time prior to failure and that embrittlement of the material associated with the g
cracked area) supposedly due to aging,~as well as the presence' of carbides and l
sulfidesplayed[amajorroleinweakeningthepipejoint.
Q+k [?) pre swab)
Comparisons with LWR Piping SA;q The staff continues to believe that the failed Mohave and Monroe piping and g
their operating conditions are not typical of the safety related piping in
' nuclear systems that were reviewed by the NRC Piping Review Committee.
In particular, the 1000*F operating temperature is significantly higher and makes g-behI?)
the alloy steel material susceptible to creep rupture and creep fatigue J'
g
' cer m.
problems as well as toiurther metallurgical changes not expected at the lower b/'/M "a operating temperatures of light water reactors.
Further, the Chromium, molybdenum materials used in the reheat lines are not used in safety-related systems of light water reactors.
g e%L In safety-related systems, light water reactors use austenitic stainless steel (V,<
y.
E and carbon steel for piping materials which tend to be tougher and more weldable.
In addition to material and operating temperature differences, there are other factors which miti against such s failure. These include:
(-
o Improved design and analysis o
Improved welder qualification o
Better control of pre-and post-weld heat treatments A higher quality construction nondestructive examination p.
o o
Inservice inspection--now being instituted in fossil plants o
Leak detection requirements o
Generally more and improved quality assurance
e' 4
l
$' unwary Failure of high temperature piping in fossil plants along longitudinal seam welds af ter long term operation appears to be a not well understood generic problem. The failure mechanism is most likely creep rupture at pipe weld l
locations with properties deteriorated because of inappropriate welding procedures and/or because of aging of the material and possibly corrosion. The Monroe f ailure is superficially similar to the Mohave f ailure. However,-the.
material used was an improved raterial and the operating conditions were more closely controlled, making the reason for the f ailure more puzzling. Since the plant condition, material, design,. fabrication, inspection and quality assurance are so different, the NRC staff and our consultants continue to believe it is unlikely that the failure mode is' relevant to light water reactor piping. The staff expects to follow developments of these pipe failures closely-and are attempting to obtain material from the failed pipe and other
~
similar piping to more carefully evaluate relevance to nuclear plant piping operation, l
f i
f l
l l
l
m-7; t,;t-r.,
3 ' ;
_i!
g,',
r
'n EN' CLOSURE 1 YISITORS TO DETROIT EDISON'S M0hR0E PLANT ON FEBRUARY' 12, 1986 R. H. Vollmer, IE
~
W. Hazelton, NRR 1
A.-Taboade, RES J
P. Shosnan, ACRS N
C. Lundeen, UT C. Czajkowski, BNL n
W 7
4 i
l
1
\\
~
9 s.
. :rg i
t
. [ [,.
N~
/
^
A
( $'"
!f
,f.
.. ~
., v,
.y i
.g s-i i
i A
{
_t
~
l e\\
., l.r.c r
^
.s.
.. hl.
" ?.
.4~..-
.\\., g
.,..... _..% ':, % p' ',....,.
3 h.,.
- [,.
. ;f',
I r
. s
~~
.j,*
Fig. 1 Unit do.1 No*+h Hot Reheat Steam Line Failure
[*<#:'p^m.,c;* Q. E 3,. l.,
- ..C'.c.~'.-.
--3.
p..'.Q g '
?'
...~'
,, 7 y
.Q. ~.b
'g., ?[
st. ' c' \\, a., -.; :
s
..s
.~
h,,*
g 6
_ ~
,e e
~
~
[_ '.
~
l 9
- d :[ l g'l:.Ql I.
..n.
y h_
r ll'.
~
~
=
1,
~
't sy..
1 i.
.$ w' s'.
- 7..
P
, as.
1. ;. 4, p
()
%,; e
=. '.,,
sj --
..+.
.;, -, ' '., ',4
.f.
l' l
- 5*
l,
-. y.- s g-i 4'.l '.
'.' ;A. &&
-)
. y ~, T.,.
_3 ?]
r i
rig. 2 Steam Line - View after pipe rupture showing bend in pipe
_ _ _ _ _ _ - _ _ _ _ _ _ -_