ML20090L880
| ML20090L880 | |
| Person / Time | |
|---|---|
| Site: | Fermi, Grand Gulf, Limerick, Vermont Yankee, Duane Arnold, Farley, Crane File:NorthStar Vermont Yankee icon.png |
| Issue date: | 03/18/1992 |
| From: | Horinka M COLTEC INDUSTRIES, INC. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| REF-PT21-92 NUDOCS 9203230300 | |
| Download: ML20090L880 (5) | |
Text
__
C'olt:c Inductrics
$4"$7p;"
701 Lamon Avenue Deiort. Wisconon 535114492 Telephone 608/364-4411 Telecopier 608/364 0382 18 Harch, 1992
-Document Control Desk United Staten Nuclear Regulatory Commincion Washington, D.C.
20555 Subject Opposed piston Engine (8-1/8) piston Cracking To Whom It Hay Concern On the 6th of February 1992, through an Internal Quality Assurtnce audit we found a cracked piston casting.
Upon further investigation, it was found that three basic piston castings were involved from one supplier.
These were F16402983 (first received truc ACHE Foundry on 01/03/89). F16401372 (first received on 03/18/88) and F16106631 (first received on 12/02/87).
Since these parts make several pi, tons and subsequently many assemblies, their location, customer order number and quantity shipped will be included as Attachment 1 to this notification.
The defect has been evaluated as a mishandling crack, most probab.ly a hot tear caused by a lack of a riser fillet under the riser pad at the parting line _and compounded by the process used to remove the riser.
To discover the defect, which is predominately so tight that it cannot be seen to the unaided eye, a wet magnetic particle test was performed on all pistons in our inventory.
We examined 1,939 pistons and found 50 cracked, or a 2.58% overall defective rete.
During the evaluation, we attemptea to inspect pistons in the rough state however, since the defect is mid wall to 1/4 diameter subsurface, we met with no success.
Pistons had to be at least rough turned to detect any cracking.
We also found that. finished tin plated pistons could be accurately ex-amined.
Our investigation revealed that the cracks occur at the same riser location on each par'; number.
However, due to piston design, the "983" pistons have the crack appear at and below the bottom of the top set of ring grooves, at the 3 o' clock position, with the part number as the 12 o' clock reference point.
The other two part numbered pis-tons have ring grooves machined lower on the piston, such that the defect is found on the ring land between the 2nd and 3rd compression ring, predominately at 9 o' clock.
Ubif fi1j
/(
s wm a ce.e nn W,s we 9203230300 920318 PDR ADOCK 05000271
.S.
r t
Having completed inspection-of all on hand stock pistons or as-semblies, including several completed engines, the following rejection rates on AC!!E roundry cast pistonn were 16402983 - 657 inspected, 42-rejected - 6.39% reject rate.
16401372 - 627 inspected, 5 rejected - 0.801 reject rate.
16106631 - 655 Aqnpected.
3 reiected - 0.46% reiect rate.
TOTAL
,1939 inspected, 50 rejected 3.581 reject rate.
No other pistons cast prior to use of_ ACHE roundry were found to con-tain cracks.
The problem is confined to pistons made by ACHE Foundry.
Our evaluation was completed on 17 Harch 1992.
Our_ conclusion is that if such cracked pistons are installed in engines, two failure modes are possible.
First. With extended time of engine operation, the crack may continue into the ring groove such that a small pieca of the picton land could come loose.
Second, the crack may propaga*.e through and around'the wall so that the skirt would ceparate from the croun.
Either case could result in further engine damage.
Additionally, wo-s+111 have several hundred pistons out for plating,
. which willlbe magnetic particle tested apon receipt at Coltec, Fair-banks Horse, before they go to stock.
The exact date of return for thcae parts is not yet known.
f Effective on the.date of discovery, rairbanks Horse went to 1001 HT of all pistons.
ACHE Poundry han prime corrective action responsibility to produce castings that do not have cracks.
Fairbar.ks Horse has a secondary corrective action responsibility to prevent existing
" cracked" stock from being shipped to-all customers.
ACHE was con-tactedion the 7th of Februsry, an-ACHE Gales Representative was at Fairbanks Horse"on the 10th, and cracked samples were sent to ACHE on the-14th of February.
There have been several meetings between ACHE l
'and Pairbanks Horse._
A. Quality Assurance corrective action audit-was conducted at ACHE on March 11.
It is apperent that the: root cause is in handling and riser removal techniques.
Risers are now plasma arc
~!
cut,-instead of being knocked off-with a large-hammer.
Also, a pat-
-tern change has been made so that the risers contact the_ piston at a heavy-section, which-has not shown cracks.
This was;done.to minimize the potential'for handling damage.
We'are awaiting receipt of the sample cartings_ poured on 3/10 and'3/11 to assess the effectiveness of corrective action-taken by the foundry.
Results of the foundries cor-
- rective action efforts wlth-the pattern equipment should
- be known by
-10 April 1992._-
We will: continue to magnetic particle _ inspect pistons; 100%.
In:AttachmentL1 to?this report, we list the location of-known ACHE
- pistons juld any licensee who has pistons of unknown origin to Fair-banks Horse =at this tine (meaning those pistons ~could have been cast by Fairbanks Horse. General Casting, or ACHE; however from records alone we can not-determine who cast the pistons).
The utility is to remove _all ACHE-pistons-from-their engines as soon as possible.
LThese pistons nust be returned to rairbanks Horse for magnetic par-
.ticle_ inspection.
An RG number must be obtained prior to return.
a.-
I,_
- a. m.,-._... _...
__.,_._,,_.__.;_1_
-.~_m..
l Fairbanks llorse will replace all pistons found to be cracked, and return those that are acceptable.
The pistons involv e d a r e gl),11 tho s e 8-1/8 pistcnei supplied under basic part number 1642983, 16401372, and 16106631, which were cast starting with the dates of November 1988, February 1988, and November 1987 respectively. -The heat date can be f d on the bottom flat area et l
the piston skirt.
The supplier logo i s located inside the piston I
just above the casting part number.
I is to be emphasized that the above work need oniv be done to ACHE Toundry pistons with heat codes (dates) such as 0398. as an example._
Piston heat codes (dates) are identified on Attachment 1.
Fairbanks Horse or General Casting pis-tons will exhibit'a cast date such as 10-21-82, and can be readily segregated from ACHE Toundry pistons.
Suspect pistons i n enginec can be.identifled, using an appropriate engine inspection methodology check the bottom of the piston skirt for cast heat dates.
3 Sincerely.
a
/
N H. S.- !!orinka Hanager Quality Assurance Nuclear and Hilitary MSH/dkc Attachment Y
')
s
+
6 l-
'y p
-W
-9'yyg*1P-dfT yTg T-cy t ' - if-P-**,-*W bWY1'r wv M i=wv W v yw w v FF Tr'r7 wy-T+ww' pTdWTNmWPr-rw me-t--M-rTM-&w me'-+-'
me " astmm-*-'r--ma ree-mehim--wor-wwwa*weewge, man-w-we m mw'M-'N-A-*
. - - _ ~.... - -.... _ - - _.. -. ~.. -. _ _
3, Page 1 of 2 Part #
!!t t l i ty Plant Canting QLy gustomer li t. C o d e, j
16600$00 Alabama Power Parley 16402983 2
QP-4914 unknown 16603301 Alabama Power Farley 16402983 2
QP-4914 unknown 16608300' Arkansas Power ANO#1 16402983 15 P0211059 unknown f
16608301 Entergy Unit 2 16402983 9
P0901295 unknown 16608301 Arkansas-Power Unit 2 16402983 8
P0899262 unknown
~16608301 Arkansas Power Unit 2-1640?983 16 P0211059 unknown 16608300 Detroit Edison Termi II
-16402983 18 Nit-22"/290 unknown
-1G608301 Detroit Editon Fermi II 16402983 22 Hit-227291 unknown 16600468 GPU Nuclear.
Tl!I 1 16106631 2
0101258 unknown (Sold as CGI) 16401372 Iowa Electric Duane Arnold 16401372-10 S54313 7050,9051, l
9055,9054
& 9052 1'6102802 Iowa Electric Duane Arnold lG106631 11 S54313 9149,9150,
& 9172
_16611224 PEco Limerick 16402903 lof26 LG606667 9109_
1G611224 PECo Limerick 16402983 5 GS279725-000037
- 9108, 9109 (3),
& 9110 16608300 PECo Limerick 16402983 1
LS630611 0057 or 0089 16608301 PECo Limerick 16402983-1 -LS630611 0057 or 0089 16606468-
'PEco Peachbottom 16106631 2 BW6408C9 0037,0039 16401372 Vermont Yankee Vermont Yankee 16401372 2
45708 0128 Pooled-Equipment Inventory Company-16402983 castings in all cylinders.:
Uppersi. 0032, 0057, 0056, 0057, 0057, 0056.-0056, 0056, 0056, 0033, 0057, 0056 Lowers- 0029 {all 12 cylinders)'
q
__.._.,_._.._,_....,_s
w Page ? of 2 Entergy Operations, Inc.
Iowa Electric Company 2
Arkansas Power Compan/
3277 Daec Road William B.
Converse Palo, Iowa 52324 Operations Assessment Superinten-Attenticat Dave Wilson
- dent Plant Superintendent Huolear Operations Route 3_Dox 137G
, Russellville, Ak.
72801 Philadelphia Electric Company -
Philadelphia Electric Company.
Peachbottomi 965 Chesterbrook Boulevard f.3D-3 R. D. - # -- 1 Box.208
- Wayne, PA.
19087-5691.
Delta,- PA.-
17314 Attention Mr. Hufnagel, Attention:
Kenneth Powers, Branen Head, Plant Manager Dalance of Plant Systems cci Bill McFarland, Systems Eng.
i Alabama _ Power Company Vermont Yankee Nuclear Power Corp
- J.
E.
Galirigton Governor Hunt Road General Manager, Nuclear Support Vernon, VT.
05354
.P.- 0.
box 1295 Attentions Rick McCullough Birminghom, AL.
35201 Experience Assessment Coordinator
' Detroit Edison
?ooled' Equipment Inventory Co.
-6400 N.
Dixie Highway c/> Southern El?ctrical
- Hewport,
.HI.
48166 International Attention:
Jerry Bragg, P.O.. Box 2625-Supervisor,.
Birmingham, AL 35202
-Procurement QA Attentions-Q.A. Manager GPU 11 Upper ?ond-Road Parsippany. Nen Jersey 07054.
Attentions.
Charles Paczalt Hgr-Nuclear Assurance
=.
hire
.i.
mi