ML20125C208
| ML20125C208 | |
| Person / Time | |
|---|---|
| Issue date: | 11/01/1979 |
| From: | Graves C Office of Nuclear Reactor Regulation |
| To: | Ross D NRC - TMI-2 BULLETINS & ORDERS TASK FORCE |
| References | |
| REF-GTECI-A-09, REF-GTECI-SY, TASK-A-09, TASK-A-9, TASK-OR NUDOCS 8001080003 | |
| Download: ML20125C208 (15) | |
Text
..
unonsuns o'
F
- "mi un utuin MONY COTYSION
[
V. ' u U N G I ON, D.C.?U%S D
November 1, 1979
- nZ':D'M FOR:
D. F. Ross, Jr., Director, Bulletins & Orders Task Force i NM:
C. C. Ciaves, Roactor Systems Branch, DSS N CT:
PRLS$URI7ER PU'lLR-OPLRAIED RELIEF VALVLS As you requested, a brief study was made of the flou o pacity of pressurizer p..ir-operated relief valves.
As I understood the probl 1, the objectives J shis study u ro (a) to resolve some apparent discerpincies in tabular preantations of valve capacities and uass veloci ties prepared by the staff, (b) check vendors use of valve manufacturer data in calculating valve flow ra tos during transients, and (c) express,ny judg::ent of the meri ts of any valve test results and analysis methods.
All of these objectives were r,ut
.et.
In particular, I..as not able to get suf ficient information from
" atii f ouse and Cabcock & Wilcox during the time available.
Hmcaver, I
' elieve tha t you vill find the information presented in the enclosure useful to your effurt.
(Ti.ie A rged for study, 31 ! curs.)
The folic..-ing cenclusions resul t fi cm this
'udy:
1.
Tos ts on the H m.i ty of 1 particular valves uscd as FORVs wi th dry, saturatcd st, 1 a t Li e inlet are:
cither a t very low pressures (30 to 50 psig) with cr.a valve size a.
as in the ca se of Dersser valves, or b.
nonexistrot, as in the case of Copes-Vulcan valvos, 2.
For the case of dry, saturated ste.m at the inlet, the uncertainty in valve capacity at, say, 2400 psig, c'ue to lack of test data may not be eve < sively 1;rge brcause of the calcula tion i 'thnds and gerieral iofc.ation avail ble on c! 9 cd flow of sie n.
Mc..aer, i t is recom-t nded that the s taf f require confit atury tests with stoam at representative inlet pressures.
3.
I am not aware of any test data for these valves with subcooled water, sa tura ted water, or t.:o-phase flow at the inlet.
There would be large uncertainties in predicted valve capacity for these conditions.
4.
The valve manufacturers, Combustion Engineering and probably Babcock &
Wilcox and Westinghouse, do not use the 0.9 conservative factor in calculating PORV flow as required for safety valves in NB-7000.
The" would not be required to use this factor or to have certified valve tests if no crcJit is teken for PORVs in overpressure report (NB-7000).
80010s00o 3 90010345
D. F. Poss, Jr. No w.ber 1, 19/9 Dresser and Copes-Vulcan do not use the same procedure to obtain the valve name plate capacity and would get dif ferent name plate values for U:e see valve at the same inlet conditions.
Copes-Vulcan uses fc ulas in the Instrument Society of Anerica Standard ISA-SP39.3,
" Control Valve Sizing Equations for Compressible Fluids." Dresser uses the Napier formula with a flow coef ficient obtaf ried from tests (see ALME NS-7000).
Because of this, the reactor vendor's method for calculating valve capacity should be checked (see C-E method in cncicaure).
6.
The tabular data for valve capaciLy for C-E and WesLinghouse in the P40 draft report are incorrect.
The valves are for minimun capacity required by the vendor and are not predicted capacities which, at least for C-E, are much higher (see enclosure).
f,d',Jr df""
Charles C. Graves Reactor Sj]trns Branch
Enclosure:
Valve Irfor,9 tion Dgg3 r
cc:
R.P. Dec.ise J
JWJUA b
T. Speis T. Novak Z. Rosztnczy S. Israel P. Norian F. Cherny A. That'a ni M. StolzenLorg 90010546
UKl05URE l
1.0 Comp 3rf.m of Pethods for Calculating Valve Flow with Saturated Ste m at Inlet A.
E a.t ropic Flew of Cogressible fluid wi th Constant ic The thoked flow mass velocity at the throat of a nozzle for a pcr'"ct gas which undergoes a frictionless expansion in accordance with c relation pv" - cons t mt '.;here c =c
= constant is given by C v (1) 2gc po ( 1 ) f 2 )2/(c-1) h, lbm
=
I A ft'-sec vo
, Y fl ('
'tre pg is the styation pressure lbf/f t and va is the stagnation 2
3' s pec i f ic. oh ne 'l' n/f t.
For a real fluid which is assned to follcw would be obtained from a table of
- En ;/ = c: st ant r ela tion, vc pr ; et cs.
For the ivmtropic
.* m i n of sa'urated steen under cu.' mf'!
a l u rp i i "
R. > c o aid be a large ch5nge in c li th decec3se in prouure si'{c the f' id is in '.he quality region.
lic., /er, i f, is asso d constant a t, say, the value for dry, satura ted steam, equation (1) could be used.
It is noted that Ice! pres-sure (2 100 ;.sia) ' s ts..ith cozzlc s :.nd natura ted sic a at the inlet indicate partial expansiun withuut forration of water (see standard texts md Wilson line).
l B.
- lsj-.c:r for ula_
The ':? pier for ula used in the 5"E ccde sec tion NS-7000 is W lbm = Sl.5A P (2) t t 7 2
where A is the throat area in in and P is the inlet presn.re in t
psia.
Note that the inlet pressure used in a test '..ould most likely be the static pressure as obtain;d from a wall tap.
90010347
. 2_
b
~
In NS-7000 a coef fic.ir nt of discharqe K d=
U f
t is obtained fro:n testing, where 3)
K W-sured d
t In the 1978 so.er adderA ",a modification to the Napice fr;rnula for 1500 psia LP 3200 psia is gi en as I'# ^t N
ki t hr
- 0. l ? ^'P -I nlo (5)
..here K
=
'y O. e d -!O61 The cu i,,c ti. n fo rtor F it.s E1.i.icn (4) e,: c. e rli nith the g
p e fii.i'ns of
.;; c i i l "a
(
")
j.1, "ote that for.ofety valves,a
.9 ccrcerentie r f ac tor ' ald be applied.
W
- 0. 9 V'd t
reted s.:ul:t irn O cn the valve is ra ted for rd P
- 11 r -
s,
'J to 1
'e.
a' r stcam.
90010$48 C.
ISA rarmula D
D D 9 i
od
., } e.
I 3
The Instru nent Society of America Stai Jard ISA-Sp39.3, " Control Valve Sizing Equations for Compressible Fluids," centains a o.iber of formulas and correction factors for sizing valves.
For the present study, the folle.;ing ec;uatiun w;s obtained fro;n the standard for choked ficw with dry, saturated steam:
35.68 F C p.- x (7)
W lb
=
hF PV
- T v
9 F
is called a p,;i: g 9: mtry fat. tor which corrects for the effect of rcJ im s ur c4 pan fers a ttwhed to the valve body.
In a call to f'~
Ud dir ec tly to the
' ' s '. -
se I
'.;.3s told iL it
n
.s.ce 4
1.0.
piping; hence, I set F
=
p For F = 1.0 and an x
= 0.72 for the Copes-Vulcan valves, this equation p
T
's, uld gico a value for C of 46.5 to get a flow rate of 210,000 lb/hr of y
'ry, sa t.e ted steam a t C;00 psia.
The 210,000 lb/hr value is a minimum re ;uired valve capacity specified by 'lestir,ghouse for 2000 psia set p r.. s : are for cn"e plants.
90010349 D.
Cppa ri son of Va rious "viel Pe sul t_s_ for_ Dry, Sa tzro te_d. S t;nm l
Figure 1 was originally prescnted in Appendix D of Reference 1 tchich I prepared in 1974.
It illustrates the ef fect of changes in inlet stagnation pressure on the predicted throat.. ass.eloc ity under choled flow c' mii*' ns.
It is evident that there is fairly clusa ep u o n t be t',,cen the lie M, "v dy,
and t'apier for mula results a t lo..,:.r pr e:uros.
":,,e. e r, a t, s ay, P 'd,, ;i a,
)
there are relatively large dif ferences.
Table 1 gives a tabular cc., ;rison of the resolts of all models considered in this study.
Figure 2 (from Refei er.ce 2) gives a comparison of prodict. d and eg.er mental resul ts for a 1911 ( v bj r.a fety-relief valve.
The valve 'r.odel nm ice is 11,D 22. T ': 6 - b l M.
The Crosby ca t alcg indicates that this valve has a thriat acca of 0.110 in?
'L te thit LLe vclve ca;.acity is generally
.eisen of, igui e 2 ord T able 1, rate the lei c r t ' i n ;,, f i c. ' :j.
r, at good agi t.:. ant h :t.een the 'os t
'2 fa co a nodels at the higher pressures.
90010350 e
5_
2.0 Di cssar n,d Copes-Vulc an V'Ive In forn tion 0
0 l
, A.
Dres;cr._ Power-0perat.cd Relief Valves-Dresser Elecrcmatic Relief Valves, h 'el No v :r 31533VX o-315 33VX-30, are us.J in,'ost CE and B&W plants.
The t ao models dif fer only with respect to the pilot valve region and, Lnce shuuld have the same capacity for a 3: <rn orifice size and inlet pressure.
v !e by i n g L :. ' J l a r An initial check of the Dresser valve capacities was e
values givcn in Icmser catalogs dated April 19/3 n!"-
' c 11/4 The y 1;. r resulting values for the coef ficicot of dischar.je, Kd' "OfU 3 PP' than those obtained fr u.a CE.
I'r. Williaa Ca sey, ".o ger, Eng n:cring Depr f nt, at Ormser '/as cont.,c.ted to rsolve this di f ference.
He call:d Er;k and s ta t J f 51t .e
/alues in tho c v:cr ca'_clejs i.ece incorrect, tha t the catalogs were no lejer being used, :.nd that the Kd of 0.95 used by CE was correct for all orifia
.. eters of the 315:^!X ad 310 3.x-30 val ves.
'"i e i s c e va l v e ro naplate capacity is ' ned on this K, the 'ldpier Iv: :Jl a a n'i the,et prPT: sre (Ul 3CCUMuldtion).
I asked for information on the tests ; sed to deta.aine this value for K.
He d
did not have specific information on the test but believed that only one orifice diL eter was tested..ith low prcssare steam during the valve developu nt,,wjram.
The test recalts
- .ie nut
.;r ti at the Natior,al Ec;rd of Boiler and Pr,:'
ce Vessel Inspections since the valve is not certifiec' for capacity in av.ar ':nce with NB-7000 of the Ai:'.E Code.
I called Mr. David Ray of the National Coard to obtain data on Dresser arid Cc,mVulca PORV's.
He did not locw of any tests on the Copes-Vulcan valve, but " j '., e da ta on Dresser Electro.r.atic Relief Valve (fiodel 15?: 'X(V) with a 1.8125" orifice din ter dnd 2 l/2" inlet diaueter).
The valve, which is mc'.. hat similar to t he Vl',VX valve, had a Kd of 0.87.
The valve was tested in 1968 at four inlet pr-uJres renging from about 34 to 49 psig.
90010351
b B.
Cor!S-Vul ca n Payter-Cy ra t ed Rel ie f_yal ves Copes--Volcan Type D-100 diaphragm-operated ccntrol valves are used as f 0RVs for rost "estinghouse plants.
Copes-Vulcan personnel at the main plant cr.]ir.^ering depart nt were contacted to obtain information on these valves.
These Cr,p:s-Vulcan valves.re not certified for capacity in accordance with NS-7000 of the A'ME G"fe.
The valve capacity is calculatod using fer ulas in the instrument Sutiety of f..cica !
Starobrd ISA-SP39.3 (April 1973).
"o tests of this valve,;ere run Lith s w."1.
}
l T o valve parc..aters (X and C ) and tuo s team properties (density, sl eci fic heat j
7 y
i ratio) are used in the calculations.
X is a dimensionless pressure ratio (picsn.cc 1
T hop across value, p -p, divided by inlet pressure, p)).
From tests at canct.nt j
2
)
p, p2 is dece ; sri.ntil the flow rate boce:.es independant of p2 (chot -J TI:w).
j
= W.4).
The valve sizing m Munt, Th< i.e uie ratio at this point is F XT (Fg C, is a... r,are, in part, of trin si7e (i.e., gup:,rtional to flow arna).
The y
ITA stwbu d pmroits use of '. ster tests to doterr:ine C.
This was d;ne for the y
Copes-Vulcan valves.
It was noted that the ISA standard has t,een adopted as an f NSI sim.,dord for control valve sizing.
E te also that the proc ures are static prCOsuces.
90010552
-)-
3.0 1cactor Vendor Information D
D
'A.
Coqustion Engineerinq u
Table 2 pruvides values for PORVs of plants with a CE ?.'SSS.
I is noted that thrce values of valve relief capacity are given.
The smallest values cre Lhe f
miniccum s eguired values specified by CE (these values were incorrectly 91.;n as attual l
vol.e ta;>acities in the t able of the M0 draf t r::,vt on CE).
The valve aplete i c;uti tios at the wt ci. :ure (i.e. 05 acco<ala tion)..c re calculated by Crescer using the Napier fon'.sla and the discharge coefficient, K,
f 0.95 obtained fr:m the d
i t.1 pressere to'... i =. h 3 f, an.
The valve na.acplate capacitics are given by:
lb W
= 51.5 K AP hr d
'.,% c e K
,A o
e aima in s
and P (ik u!.h d is N gi d
Values of 'he orifice dir neter and area are supplied in the table.
!ac t l'g,\\ r ;
' n an a f M dve od We ana.M& h ud d % a The u t.
c eritical flow ;;.r.!el for t he mss v : loci ty G in ib/hr - f t to calculate valve flow during a transient.
CE used the follo..irg sodels for G:
(0.10CCP - 1000) 1*
G = b l ',^4 (0.2/92P - 1061)
The f 1c in pi enthe:,es is a coii ection to the h'apier foioula us:d at (luc 3 psia.,hich was devel 4 ed to give approximately the value for G as that obtained from the Ha ogeneous Equilibrium sa,:
!'odel.
1; is given in the Sum.,cr 1978 Addcoda to the As"E Code.
2.
For t'.so pMse flow:
G from 'S.dy Model 3.
For subcooled:
G frca ";r.ry-Faushe N del as.nodificd by CE 90010353 D.
'.lr s t i ro;heus e The caly infor,.ation obtained sis that valve < anacities such as 210,000 lb/hr for f; orth Anna and 109,00') lb/hr for Cinna a re '.-lestini;i.cose -specified nini:. um c a ::c i t ies.
!!:nce, the t".0 ta!..lar velors fo r !lestinghouse i elief valves 3'1ould be corrected.
C.
B3}l I asked fur saeci fic infornation from Ef.W.
Tr,ey have not s'mplied this infro e ition.
DTD D ~T X [ l do o l' _11 g k i
90010354 1
I
. 9_
l I
~
1.
Appendix 0 of "Evalua tion of 1.0CA Hydrodynaatics," Regul.1 tory Sta ff:
Technical Ccview, USAEC, fiovember 1974.
I 2.
T h o:.4p so n, t., a nd E v. t o n, O. E., J r.,
M.
cn 1_ :ntr pic flow of Dry Sa tura ted Staa a Through Pressure Relief '!al /es," June 1979.
90010355 l
i 1
1
-e s
= -. -
m-----
' \\
l
)
s
./
D*"D "D ~T ] f
.cc c
I i
.,n99
/
/
//
MOODY 7',n,,,
!.'70 0 L
/ !
liOI.'.DC E f. E O U S
/
t.O U! L IB a l U'.1
/
/
f L C'V l.'.0 D E L
//
CDy3 l
/
\\
7t m _ t:AMt:n
/
r,O R fJ.U LA t4
/
n
/' /
I u
c E
.9
- ?: 2 o
/
/
/
2033 l
i 1000 l
I O/
I - ----- l - - - -
b- -
I- - -- -
I
)
2000 0
400 000 1200 1 coa 2003 2,100 UPSTar AM Hir MUHE, M23 Figuic D 1 ' Ct.mparkon of Theoretic.it V Jues for Cheked flow.
f.'ea',s Wloci:ics f or Dry Sa or..;cd Stt.ua at Intet 90010356'.
l
\\
t I~ :0 o
j
?203
( V/u) rncx j
o E l CO -
WN
/
o
- CO -
/
o l eDO -
/
g, g.
_1 R A'.* S IE NT
/
FLOW
/
4 y t oC)- y
/ o
- P' ?] -
/ c-
'D)**D
- D ~3~l 6 8, p.13 r-o Ice
,l i :'s3
/
o
- m u
i.j l 3Oo..
/
ec
' l 200 -.
/
oo
!(
!.] f l C D -
/ l f icca
.s. 'nl I
c 3 C 33 -
, l a.tl 3
T. '. S T.,. :Sll t.T S
/
o i.00
- CO I
t cea I
d' Z.Q - -
i o
400-le 6
e i
- i e,
,, -y a
i i i. i e i
i 05 1.0 g93 g,g
':!.,.9,
.N l
1 i
figure 2.
Comparison of Ircdicted and Experi::.cntal Capacity 2 Orifice).
of Small Ciocby W ive (0.110 in 90010357
Table 1:
Effect of Pr.ncore on Chnind Flow Mass Velocity for Dry Saturatcd Steam C
G M ody_
G HEM GSIA canstant"
/
G Moody 2
psia lb/sec - ft G Napier G Napier E Napier G flapier 50 107 1.036 1.020 100 210 1.017 1.0 200 413 1.0
.983 1,0 1.031 400 819
.993
.971
.989 1.070 6JO 1229
.9J4 909
.984 1.017 900 ICi6 1.0
.972
.387 1.022 1000 2074 1.007
.977
.997 1.032 1200 2513 1.016
.M3 1.006 1.042 1100 2365 1.028
.992 1.012 1.051 1600 3434 1.042 1.002 1.037 1.075, 1800 3922 1.058 1.103 1.057 1.096 2030
'035 1.076 1.027 1.034 1.130 2200 4979 1.099 1.004 1.130 1.165 2400 5565 1.126 1.064 1.164 1.2.4
+"ormalized to G Napier at 200 psia.
t 200 psia If formalized to G e
constant G
would agree with G c to o m,in 1/2i ISA constant D
lD *D
~
't
~ j f won wJ Al n
90010358
7 6Sf01006 v.<
C)
O O
O c-O a:.t LO O
U O
Ln Ln CO in D
M Ch CC CO
.I M
Ln e-i O
Ln w
e-4 e--'t e--f
+1 v')
C)
C)
.V O
CD O
C C)
M' D
O to o
Ln Ln CC Ln rr)
CD CU CO O
ci LO rs O
LD C
C ra re e-r 1
<e-.
l ro d.
x>
L' i e- +
9 $
Db C
W 3
O O
O r-CY O
O O
O O
.c O
t.'
Ln N
Ci r
n.
r--
CD OT U
O O
e to tT*
U O
O e-*
e-+
r C
e-4 e4 3
e e
e L
M v
O C'
Lt.
C!
L C)
L
~1 C
ec V
r-V V
Ch
- r-C' S.
G N
e e
m C'
LU C)
O O
Ci O
Ci C
C O
et Ln O
4 '
O O
O Ln in cc L%
cr 4'
C e
O res C,
C.
- 1 s
C^
C C.
J>
4' w
C' 'a e
e e-LO e-a Ci Ln U-
+
li j
D r- -
e-s e--<
e-4 C
V
~ru 3
Ci 4'
In
+ c-F li.
+' e i r
O V
eG V
U C'
C e
t O
F C
m I
e-Ct N
C)
O e-n C) h r
v t.
D C>
M LC:
C) v
.n to
- r C) ru O
CO LD N
C' L
C-r--
M
+
N C%
C' O
L C*
It
.C n
- r=
rn Q
M
.C O
(
r ro e--
Ln et O
t, CD p +'
4-Q>
F-rc e a e-e e-4 O
et e
L--
CL ET C) r-O m r.
L e
i 1
e r0 +> X i
C U
7=
C.
t
,4 t,
i N
c, N
l e-t C)
N Ce 7
Er e-c)
C>
C-Lc 9
O M
Ln'^
N C1 U
r LM O
LO Ln c) 4 rr) e (L
ci r,
C.i s
(
> 4 ro CO 03 4'
I r--
e-Ln e-*
O Ln w
C e
to e e-t r-t r-a
_C -
L
'c i s'
v C.
L e
L C
L C
4 '
C O
C 11 C'
f D
U r
Li C
4 l
M N
C) h O
i
- O
.O
+>
.O C'
LT Cr r.
L)
L C)
G) e--
to O.
r' r
t'.
e-
+)
e--
D U
c' a
r- - 4 C
CL C)
C_ >
r.
rd
- r-0 e-C.
s t^
4*
t' U
r-
+$
0 r- -
a L.
C D
ca U
.r-E~
CL C-c-
a O
O.
C~
r-U ru r-a' L*
c e
r C)
O U')
L.
s 4
et Z M e
.; C
[^
L r
4-C1 e---
Cq r0
- c-M N C#
r0
.C 3
t 15 3
ru Il ec O
U
+8
+
r Cr e-Cs u
.e-L.
e <
c'a c>
q L^
O O >,
L C) 3 h r e
c CJ
+2
+> N C-C, c
c C
Cr
+6 e-I)
+'
4 '
4*
4 4
ru er - Lt vi ro
['
U t
Cl L
e
+'
U e-C.
(>
O C,
c-r-
CL M ru U U
90 CL
- e-rU Ei fL r-
.C.
U C1 e
'r-to h
U C) to eu C U
.O r-e f-O U i
l L.
.c-ro ru lE O
C)
F.
O i
l
_e n _