ML20206P362
| ML20206P362 | |
| Person / Time | |
|---|---|
| Site: | Grand Gulf  |
| Issue date: | 12/12/1996 |
| From: | Schauki N WYLE LABORATORIES |
| To: | Thomas Scarbrough NRC |
| Shared Package | |
| ML20206P301 | List: |
| References | |
| 50-416-96-03, 50-416-96-3, GL-89-10, NUDOCS 9905180218 | |
| Download: ML20206P362 (11) | |
Text
7 FC.-12-96 Thu 4.li rm
- 44 w 12.DEZ.1996 18:C'/
SIKLs) Crr NDM4 tR.703 P.2/9 l
FAX Wyle Laboratories Huntsville, Alabama 35807 (205) 837 4411 Fax: (205)7210144 TO:
Tom Scarbrougli NRC FAX (301) 415 2444
Subject:
OL 8910 Closure Meeting u Grand Gulf andInspectionReport No.
50 415/96-03 I
DearTom; Referring to the Generic Letter 3910 Closure Inspection Report No. 50 415/96 03 of Grand Gulf's Program, the inspectors stated an open question about identifying major damage which would lead to unpredictable behavior by the Siemens Methodology. We be[ieved that by sending the fax about the comparative analysis on two of the GI-87 test valve on Feb.15, to have solved the issue. (See Attachment) Furthermore on our telephone conversation during the closure meeting on Feb.14th,1996, we informed your inspectors about other considered items to usure coverage of known condidons / design features which would lead to valve failuru. ' An example wu given on dise deflection on the seat at high energy loads which would lead to major plasti 5 cation of the body seat thus leading to functional fallure. Such behavior has been identised by peparate effects tests, large scale gate valve testing u well as by Finite Elemerd Model Analysis at Siemens in Germany. Anotheritem was the tihed disk seat contact at high loads. The calculation of disk stability and disc seat contact are one of the irnportant analysis models in the grouping methodology to cover this behiivior.
Please s!!ow me some statements about "Siemens Grouping Methodology" to usure clarifying this issue because the veri 5 cation of the operability and predictability is donc by performing dynamic tests on selected covering prototypes. The method does not verify functionalbility and predictability solelyby analysis.
The major goal of the Grouping Methodology is to identify s family of MOVs with similar dynamic and stress behavior as well as to define the test candidates which require test results from dynamic tests at " Design Basis Conditions."
~
~ ~ ' '
" ~ ~ ~ ~ ~ ' '
'~
~ ' ' '
i 9905180218 59d513"~~'~~
PDR ADOCK 05000416 P
8,oa.,
...v 12.OCZ.1996 18I07 SIMMJ CrT NDrH fR.703 P.3/9 l
Tests preformed on worst case candfdstes will show any unpredictable
- behavior, The test conditions and pressure build up are compared to the analysis models to verifycoverage.
The test signatures and inspection results are the basis to show that the e
valves are performing predictable thus a!!owing spplicadon of test results to the MOV family.
We did not feel during development of the Model that the analysis alone should cover the Amctionability verification. 'Ihis is due to the fact that a considerable number of valves have stress exceeding the yield strength and thus an elastoplastic calculation would be required, Due to the largs variety of designs and interaction of the parts, the analysis was not considered to be an efficient tool for answering the funedonability question.
Iffou have any Airther questions, please call me (205) 837-4411 c=t 552.
sese negards, i
1 s.
4
)
Dr.N uld
/7 p[
yleLaboratories 7800 Highway 20 W.
Huntsville Alabams 335806 O
e w
e e
. 90.
gog ogg p go agamagw e este.
e top stemmene menn.e. hm8
..D e a mee a6e e e..
e.eee mumeneemmemumm aw w ee
.*****=u'8
- '**8'#**'8
e
v w coee m ee m em
'te.rcz.i m is:er smu err tem re.7es e.4 v i
\\
Entergy, Grand Gulf Atteistion: JeffWright(601)437 2146 cut. 6229 i
Subject:
Analysis Mathodologyndentificados ofmdor damage.
Please ind attached as agreed upon dudng our meeting on Feb.14th,1996, reganilag bench marking of the analysis methodology on the DEL tested MOV No.1 (AD/900#,6"). As a comparison we analyzed MOV No. 4, (AD/900#,IO*) and MOV No.5 (Powe0/900#,10"). the mults of the analysis and conclusions are as foUows: : Compadson between PoweH 900# and AD 900# flergate valves. At first glance the two designs seem to be close and comparable.
- : Shows the dynamic behavior'of the Powell gara valve with a stable transfer of the disk kom ib guides to the seat ring. This indicates na acceptable behavior. : Shows the dynamic behavior of the AD gate valve with an arfy tilted contact between the disc and seat. This early tDred contact at high anergy loads with a large tihing angle indicates sousing and h % galling to be expected. The inspection of the intamals revealed that puts of the hardfacing were shcared off and gauing of the kliding surthces identified.
.. : Shows a compadson of the stresses between AD and Powell gate valves. Most nmarbble are the average body / disc guide stress of AD (Design No. 2) widca j
,n sceeed the 40 rapectively 50 KSI. This Wir=* a nudor damage of the guides to be espected. Again the Inspection of the guida vwilled the damage. Powou, Design No, I, had low average stress and therefbre only local plasdfcation of edges are indicated. This was also vedfied by the inspection.
' Attachment 5: Shows a relative compadsen between the 9 /10" AD gate valves to the W Powell gate valves. Tha dats shows that the d' AD is the weakest design. The 10" AD is snarginally betterin the disc seat bearing stress. The average body / disc guide are by hetors hisher than the Powe8 gute valve.
As a conclusloo, of the analysis a failure of the 6' AD, a poor performance of 10"AD and a Air perfbtraance of the Powell are aste valve during blowdown conditions is to be expected. These conclusions confonn with the blowdown test:
prfbrmed byINEL.
Ifyou have any questions, please caU me at (205) 837 4411, art. 352.
Beat Regards.
i Dr. Nab'
~
s
_________s
M C-12-95 YMU 4:18 r0 tru r u av. -
- w. e.e~e.
......e sr.ecz.sssa tones sym; orr ren4 rn.res e,sa fl l
~
SIEMENS DISK GEOMETRIES i
i POWFI I
}
l
+
g.,
-l1,$.
I t
e e
4 A
e e
t e
m.
ANCHOR-DARLING
+
Q y
,,,e
- ,e Anp.
e
'h-
.7 0
9 S e eees es e enee se s eemse e** epee ee
- 994 taesemer 9 5 4 se ge me ecos ese esen, e m e
- e b emameneenNW_
.m Weyee o
e e m eese t e m
4 9
+
u
~~
DEC-12-96 THU 4:19 PM Tru fAA W-12.DCZ.1996 18:00 S!KR) 0FF M fft.783 P.6/9 l
D.
O o
z
=
~
's i
3
,y cocas smE53 j
_ _ _._ _ _ _ _!_ _ _lohnoo Ad. s wwwi8 O
Ainadvis ossa o
O m
J m
s o
v W
C e
w
(
g W M
(
O g
O
-o g
m y)
J o
a o
- .j
- l l'
i 0
m s
g
)
O O
en
,O g, i i
- o '
i O
\\
Q l
4
- # o,%g i
i i
i
-W
-o
>J O
O O
O O
O O
p C
CD O
T
,N
'.XVN io %
C gegygggp ygg
__g-_
.sD gG
.4...O U MOO 0
4
p';;acn= e %*
~;
E
,e-
! 8!
NRta E8 5Ec k h
,I W
)
P D
0 N
\\
0 1
W o.
by0. $ *o.
9 x
0 O
9 O0 D
,#y3 W
0
>eaEg e $o 8
O l
l 1;!Il; l
L 6"
e m
%:w. -
4,.
0 pE B
7
.hM*s;A i
E SUE n
(
0 D
F W::
__N
't cT S e-E 0
F 0
E 9
p' 6 K
(
EO K
mST
. T O
i G
s O
N-- Q:'
a-s
~
R a
0 T
R L
S I
b.
oo
.a a$z~
5 T
R f
o S
A-
~
0 4 %
~
D-G R-N O-0 I
3 Aq[
S H-C-
p O
o o
N-sg 0
L A-Y A
g.e 2
~
C o
- o 5
0
.y
.x E
1 s
V B
L 1
0 A
0 Q
0 g
0 0
8 V
0 g"
6 f
2 1
i E
A<g,o f TA G
tI!
n
!8
- ,, i : t(.
I ii 1 !i ':.;
~~ Vts-[2-$bi$U 4.Is TM trw rn4 4w, 12.DCI.1996 19209 SI M ) Off tD14 PR.703 P.8/9 l
I Co 9
8 8
?
8 R
S I
O l'
3 7
q-
$ j-o s
a.
y, 3
.a 4-
. z Nc
\\
b 8g s~
-j}
C T
4-8 a
1
-4 l
l c4
~
e ya e
E-a m
I
&8 w
,~
O 4
a q-e
.9 b.
8 I
t-
.. l p
l s
O O
O O
l S
o w
1
... www e.we====* ** -
D
tf l)I l1
=
.: \\-
i i pE 5 a5 n
kI e
E s
0 D
0 0
o 5
5 0
5 c
o 3
2 2
1 s
s
~
.,~
e
=
sW e u rG e
t c
S. is evD A
e sd eie u G
r tS.y ed v o 0
AB 1
l mc le a s g
ei w
tD n
S:
i o
nn r
l nP Do a
it o
m n c D
es e el s
kf r
o Se o
i r
pD r
f U
h as c
r m
n pe ec A
t s f
0i mi 8D f
E N:
d Dg o
n g
ni Cn red en r
i t
e l
s r
pB 6
s a
U sD g
t a
er ke n
cS i
r o
o /c r
l t
h L
s a
S c
iD D
n r
A
,g en o
vi Ar h
a c
e B
n
\\
A l
as E
c e ad e
t.
l uG
-g a n vAr i
ae B
~
i
~
d M
M M
0 0
0 5
1 Mt i
.Qks-y e
m
~
!ii1: j.t
,*I
.'.I l,! -
IlI
- , I,,
Ot' -l z-$o i tiu 4.zu rm itw
., u.
a i B R g m' E M-W B M W~1.I...J.~ % C_
I"*E-
.,,_T~"? T..
r ADAPT A DETERMINISTIC APPROACIITO PERIODIC TESTING (FOR VALVES)
Problem:
Generic Letter 96-05 requires that utilities ensure that MOV operability is maintained throughout the life of the plant. In this age of deregulation and competition, utilities have more limited resoumes to spend on maintaining plant safety and operability. Full flow dynamic testing of all GL 89-10 program MOVs unduly burdens these resources and may, in some cases, reduce plant safety by degrading critical MOVs.
Solution:
I l
Wyle has performed an evaluation of the long term effecu of flow media and stress on nuclear i
power plant gate valve performance. The results of this study show which valves are highly stressed and/or likely to have corrosion problems. From this, a logical action plan can be developed that considers the subject valves' margins, testability, etc.
Benefits:
Plant resources can be better utilized through an improved understanding of strengths and weaknesses in the performance of GL 89-10 program valves. This approach also reduces the chance of degrading valve parts during testing. Finally, Wyle's approach results in a more cost-effective means of meeting GL 96-05 requirements over the life cycle of the plant.
e 2
Attachment to GNRO 99/00042 Appendix B Excerpt from Position paper to address NRC concems regarding GGNS grouping program Prepared for Entergy Operations by Wyle Laboratories February 22,1996 i
This appendix is referred to in response to Question 1 of RAl t
Note: This position paper refers to Siemens report 5221/94/025. This document is not attached because it contains proprietary information, but is available on site.
E
e Attachment to GNRO 99/00042 NRC Concem #3 How does the performance analysis address the potential for unpredictable behavior under blowdown conditions, i.e., are there established threshold criteria that defines when unpredictable behavior will be expected. Unpredictable behavior in this context is meant to describe the condition where metal shearing and gouging takes place on seats and/or guides.
Response
The potential for unpredictable behavior was addressed by a combination of analysis, engineering judgement (based on Siemens extensive experience with valves), and test results. A Similarity Analysis was performed for each valve, in which the stresses in the functional components and kinematic behavior across the entire stroke were determined. In the Similarity Analysis reports, the following were calculated / documented and used as a basis to predict the performance of GGNS's GL 89-10 gate valves:
Pressure build up across entire stroke Transition of the sliding surfaces from disk / body guides to disk / body seats.
Contact geometry (tilted vs. stable)
Overlapping seat area e
Materials of functional components (hardfacing and base materials) e These calculated stresses, combined with kinematic performance criteria such as the disk tilt angle, were used to determine the relative loading between valves in a given family. Valves with representative loading conditions were recommended for testing.
The test results were then reviewed to determine the predictability of the valve / family, in addition to the detailed engineering review, Siemens extensive experience was used to determine which parameters were most influential for a valve's predictability. For example, the disk / seat contact in a tilted condition is considered to be a critical item, especially under high differential pressure and flow conditions. Edge contact and high loads lead to scissoring effects that cause plastification of the disk / body seat. This causes unpredictable behavior at blowdown conditions with early guide to seat transition.
Another issue of concem which has been considered is the disk deflection on the seat leading to high stress intensities. In case of blowdown this can lead to valve malfunction. Fortunately, none of the high loaded gate valves in GGNS's GL 89-10 program experienced such deflections. In case of plastification, an engineering evaluation is performed to identify the affects on the valve's performance characteristics. For example, it was predicted that the lower edges of the disk guides of the 10 inch and 14 inch Powell gate valves would plastically deform due to the high local bearing stress without any major affect on the closure of the valve. This was confirmed by the post-test inspection of the intemals after the design basis tests were
. performed. See report no. 5221/94/025 pages 11,13,15, C-3, and C-4.
l;
<