ML20127G212
| ML20127G212 | |
| Person / Time | |
|---|---|
| Site: | Comanche Peak  |
| Issue date: | 10/31/1975 |
| From: | ALLIS-CHALMERS POWER SYSTEMS, INC. |
| To: | |
| Shared Package | |
| ML20127G205 | List: |
| References | |
| ER-504, NUDOCS 9211160372 | |
| Download: ML20127G212 (112) | |
Text
_ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ - . .
l ENCLOS N: 3 TO TXX-92503 Allis-Chalmers Powers Systems, Inc.
Engineering Report No, ER-504
" PROBABILITY OF TURBINE HISSILES f o, 1800 r/ min Nuclear Steam Turbine Generators with 44-inch Last Stage Blades " October 1975 9211160372 921110 5 PDR ADOCK 0S00 P
ALLIG=CHALMERS POWOR SYSTEMS,INC. C f h
- CONTENTS i
i i 1. INTRODUCTION ............. .............................. I
- 2. HISTORICAL DATA ......................................... .. 3 3 TU RB I N E D I S K I N TE G R I TY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9 3.1 esign ............................................... 9 3.2 Safety Analysis of LP Disks .... . . ............... 11 3.3 Manufacture and Quality Assurance ................... 20 3.4 LP Rotor Assembly ....... .. ....................... .23 3.5 Balance ar.d Overspeed Test ....... .................. 31
- 4. PROBABILITY OF FAILURE DUE TO OVERSPEED .................... 33 4.1 Introduction ............... ........................ 33 4.2 Reliability Block Diagram ........................... 35 4.2.1 Electro-Hydraulic Control ....................... ... 35 4.2.2 Mechani ca l-Hyd raul i c Con t rol ............ ...... .. 37 ,
4.2.3 Ove rs peea T ri p Sys tem . . . . . . . . . . . . . ........ ...... 37 4.2.4 Stop and Control Valves .............................,38 4.3 Failure Rates ................................. .....40 4.3.1 HP and IP Stop and Control Valves, E l emen t s A l - 4 t o D l - 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 7 4.3.2 Follow-Up Pistons, l E l emen ts E l-6, F 1-6 and H 1 and H2 . . . . . . . . . . . . . . . . . 60 !
4.3.3 Elect ro-Hydraulic Converters wi thout Colls, 1
Elements G1 and G2
.................................61 4.3.4 Mechanical-Hydraull e Con trol (MHC), Element i .......62 i 4.3.5 Admission Controls wi th Coils, Elements K1 and K2 . . . 64 i i l 4.3.6 EHC Speed Control and Speed Meast';ng Device, l l E l eme n t s L , M a n d N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 5 I
i 4.3.7 Power Supply, Elements 0 and P ...................... 65 l 4.3.8 C han ge-Ove r Dev i ce , E l emen t R . . . . . . . . . . . . . . . . . . . . . . . 66 l r i l 4.3.9 Main Trip Valve, Elements 51 and 52 ................. 70 i 4.3.10 Overspeed Trip Test and Resetting Device, Element T .71
- 4. 3.11 Overspeed Tri p Bol ts , E lemen ts UI and U2 . . . . . . . . . . . . 75 !
4.3.12 Trip Release, Elements VI and V2 ....................77 l
l t j
i n
ALLIS=CHALMERS POWEF4 SYSTEMS. INC G
( CONTENTS
~
l 4.4 Overspeed Fai lure P robabi l i ty Cal cul at i on . . . . . . . . . . . 77 l 4.5 C o mmo n Mo d e F a i l u r e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 7 4.5.1 Normal External Power P lant Envi ronment .............88 i l
4.5.2 Operation and Maintenance Errors ......... ... ..... 48
'.I 4.5.3 External Events .....................................91 4.5.4 Conclusion ..........................................S1 L I S T O F RE F E RE N C E S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........S2 i i
l l
l t
i i
i i
i i
l h
1
Au.so-CHAL.MGEMO POWKM SY8' REMS,IfvC. O E R- 504 10/75 -l-r
- 1. INTRODUCTION ],
l This report provides information on the probability of occurrence of turbine i I
missiles from burst-type failure of low pressure (LP) blade disks of 1800 r/ min ;
curbines designed for nuclear power plant applications. Two distinct cases are i covered: 1) LP disk failure at or near design speed (<l20% of rated speed), and I
- 2) LP disk burst due to an excessive overspeed incident, l For the fi rs t case, in f orma t ion is provided on the turbine disk integrity i including details of design, manufacture and quality assurance which provide very high reliability apinst disk failure in the design speed range, in addi tion ,
I it should be noted that even if a disk burst occurs in the design speea range, the disk fragments would be contained by the turbine casings. The details of l this analysis are given in A-CPSI Engineering Report AER-503 (1) which shows that I
burst speeds in excess of about 160% of rated would be required to yield external i turbine disk missiles. Since an LP turbine disk missile can only penetrate !
the turbine casings at a turbine generator speed in excess of about 160%,
r the probability of significant turbine disk missiles within the design speed .
s I
range of 120% is considered zero.
l I
For the second case, LP turbine disk missiles with the highest energy level are defined for a 130 to 185% speed range, where the average tangential disk stress is equal to 85% of the disk material ultimate strength. The I probabiii ty of turbine missiles for the s case is assumed to be equal to the l
probability of an overspeed incident of greater than 120% of rated speed. This is determined by means of a reliabili ty analysis of the turbine valves, speed control and overspeed tri p sys tems. The results of reliability analysis
( /
1
Au_is-osAa. mans mwsa sysroms, we C E R- 504 10/75 ,
are values of 1.6 x 10'7 per unit year for a 6-flow turbine generator and
~
2.1 x 10 per unit year for a 4-f1w turbine generator. These values are intended for use as "P j
" in the equation P =y P xP xP where P = overall 2 3 probabili ty of turbine missile damagr, P = striking probability, and 2
P = damage probability, 3
information needed to calculate P2and P is 3
given in A-CPSI Engineering Report DER-503 i
i r
l !
! I i
i l
< J
-i
AL. LMB =CHAL. MEMO POWKM BYSTEMB, NC.
ER-504 10/75 ~3-( ^
- 2. HISTORICAL DATA Historical data on actual cases of turbine-generator rotor f ailures is of !
general interest in connection wi th the question of probabili ty of turbine missiles. However, as discussed below, such data is not di rectly applicable, for mode rn turbines and should not be used for predicting future probabilities t
of missiles. i r
l A study published as Reference No. (2) includes data on turbine and generator >
rotor f ai lures cove ring approxima tely 70,000 uni t years of operating experience
, of uni ts larger than 50 Mw from 1950 to 1972. This study reported a total I
of 14 failures which are as listed in Table 1. Essentially all of these f ailures are not applicable to the turbine missile probabili ty considered l i
I herein for the following reasons: i
. i 1 !
! 1. Most of the failures were primarily due to high nil-ductility l transi tion temperatures and low f racture toughness, presence i
, of hydrogen flakes or non-metallic inclusions and relatively 1 !
t l
undeveloped quali ty assurance proceduras, all of which l
i l were characteristic of steel melting practice and forging i technology prior to the mid-1950's. Since then , major imp rove-ments including introduction of vacuum degassing for all alloy i l
1 I
steel grades used for turbine and generator forgings and appli- !
cation oi sophisticated ultrasonic testing techniques have greatly reduced the possibili ty of failure due to these causes.
l q J t
_ . - . - _ _.-.m. . ~ - - - . - - - . - -- _.-- . - - - - -- - - - - - . - - - . - - _ . - . - - . - - - . _ . - - - - . ~, -
L i
_ _ _ - _ _ . _ _ - - ~ . - - - ~ ~ - - -
f m i P -
Table i Failures of Steam Turbine-Generator Rotors at or Near Operating Speeds di o
- Units Larger than 50 Ne from 1950 to 1972 e i Manufacturer & Year . Suspected -
Sire. PWe other identification Failed Type of Failure 'Cause of Failure Comen t & . i u
- 1. 63 Siemens 1951 Low pressure turbine Brittle fracture Missiles-factory test .;
rotor burst i
. 2. 100' GI (Tanners Creek #1) 1953 Low pressure turbine High temp. rupture No external missiles 1st stage disk broke 3 100 GE (Arizona Pub. Service) 1954 Generator rotor burst Brittle fracture Missiles-factory test
- 4. 168 GE (Cromby #1,Phila.Elec.) 1954 Generator rotor burst Brittle. fracture No external missiles through repair 5 100 ' Charles A. Parsons 1954 Generator retaining Br . *le fracture External missiles b
! ring burst thru vent holes i 6. 100' Charles A. Parsons 1954 Generator retaining Brittle fracture Limited external e l
ring burst thru vent holes missiles 7 ' g:
- 7. 350 A-C (Commonwealth Edison) 1954 Turbine spindle burst Brittle fracture External missiles 1
- 8. 153 GE (PGEE'Pittsburg #1) 1956 Generator rotor burst Brittle failure No missises-9 87 .Hinkley Pt. A #5 1969 Disks failed Brittle failure Missiles '!
- 10. '87 Hinkley Pt.'A #4 & #6 -1970 Disks failed in pit Brittle failure Missiles-factory test 1 ;
1970 Mitsubishi (ENESA Spain)'
. II. 7 1970- Low pressure turbine Flawed forging (7) Missiles-factory test C Vestinghouse design. rotor burst N 3
- 12. ISO GE (Cutler #6 FP&L) 1969 Generator field winding 'Out of-step with No energetic missiles 9 assembly failure system i 13 105 'GE (Essex #1-PSEEG) 1972 -Generator field failure Abrupt braking Medium energy missiles-l of generator shaft coupling
.g 0 14. 7 Mitsubishi (Kainan) 1972 Rotor failure during (7)' Exte nal missiles 9' t
'Vestinghouse design overspeed test in plant 5
/
/
1
)- .
Au.m-wt.unne #=owan eversum, we.G ER 504 10/75 *S-3
(
- 2. Seven of 14 failures in Table I were of generator rotors which are not directly applicable to LP disk f ailure probabill cy.
f I
3 Four of 14 failures in Table 1 occurred in factory test pits !
which are not directly applicable to f ailures in power plants. .
i
- 4. Several failures in Table I produced no significant missiles. '
and are therefore not directly applicable to f ailure probabili ty
, for purposes of turbine missile analysis.
Historical data from A-CPSl's parent companies is includeo in Table 1 and l i
I i
the overall study in Reference U). This speelfic experience, updated through the end of 1974 and expanded to inclLde units larger than 10 % !
is given in Tables 2 and 3.
l i
Based on the data ir Table 2 for KWU, Siement and AEG units larger than '
10 Mw from 1950 through 1974, there was only one rotor failure during a l.
factory test in a total of 8,688 unit-years (26,187 rotor years) of l opera tli., e::pe l ence. The f ailure v:a a brit.tle f racture of an LP turbine rotor in 1951. The cause of this failure wan primarily due to excessively high
\'
hydrogen content of the forging. I I
i l
i 1
~
AU.tO-CHAL.MiERS POWCR GYSTCMS,8fMC. =A E R* 504 10/75 .
, 3 Table 2 Rotor Failures in Kraf twerk Union, Siecens and AEG Steam Turbine-Generators larger than 10 Mw put in Operation since 1950 Cumulative Cumulative Number of No, of Units No. of Rotors Unit-Years Rotor-Years Rotor ,
Year Put in Operation Put in Operation of Service of Service Failures i
1950 10 28 5 14 0 1951 9 26 20 55 1 (1) 1952 21 52 50 135 0 1953 19 53 100 268 0 1954 32 90 175 672 0 1955 10 57 276 750 0 1956 47 121 411 1117 0 .
41 124 1606 0 l 1957 590 1958 49 150 814 2232 0 1959 24 91 1974 2979 0 1960 29 92 1361 3817 0 1961 26 90 1675 4745 0 1962 30 91 2017 5766 0 1963 41 131 2395 6897 0 1964 33 112 2704 8149 0 1965 26 95 3254 9505 0 1966 39 108 3741 10,962 0 1967 38 118 4246 12,532 0 1968 33 97 4797 14,210 0 1969 36 118 5385 15.995 0 1970 34 112 6002 17,895 0 l 1971 21 81 6650 19,892 0 1972 13 38 7315 21,948 0 1973 15 41 7994 24,044 0 1974 15 53 8688 26,187 0 Totals 701 2169 8688 26,187 1 l
NOTES:
l l
i
- 2. Includes all turbine and generator rotors, but not exciter rotors. l i
I 4
ALLans-CHAs.MetMe MetN m>1978EMO. SNCS ER-504 10/75 -
- 7- - ,
Table 3 Rotor Failures in Aills-Chairers Steam Turbine-Generator 9 Larger than 10 N put in Operation since 1950 Cumulative Cumulative Number of No. of Units fio, of Rotors Unit-Years Rotor-Years Rotor Year Put in Operation Put in Operation of Service of Service Failures 1950 8 22 4 11 0 1951 8 18 16 42 0 1952 5 12 34 88 0 1953 11 25 60 152 0 1954 7 23 95 240 1(I) 1955 8 24 138 352 0 1956 9 12 189 482 0 1957 6 lh 248 625 0
, 1958 20 61 320 805 0 l
1959 8 23 406 1027 0 l 1960 11 37 501 1279 0 1961 10 32 607 1566 0
, 1962 5 14 720 1876 0 1963 5 17 838 2201 0 1964 6 20 962 2545 0 1965(2) i 3 1089 2900 0 1966 - -
1217 3257 0 1967 - -
1345 3614 0 1968 - -
!!e73 3971 0 1969 - -
1601 4328 0 1970 - -
1729 4685 0 1971 - -
1857 5042 0 1972 - -
1985 5399 0 1973 - -
2113 5756 0 1974 - -
2241 6113 0 I
j i Totals 128 357 2241 6113 1 i
NOTES:
- l. Brittli fracture of an LP turbine rotor in a power plant.
- 2. Allis-thalmers stopped taking orders for stearn turbine generators in 1962 and put last unit in operaticn in 1965, but re entered the business together with Kraf twerk Union forming Allis-Chalmers Power Sys tems, Inc.,
in 1970, 1
1 I
L j
e> > -- -co sat usuem mm eviavarMan, ava ^
ER-504 10/75 *d~
( r ,
As shown in Table 3 for Allis-Chalmers uni ts larger than 10 Mw from 1950 throu 1974, there was one rotor failure in a total of 2241 unit years (6113 rotor-years) of operating experience.
The f ailure was a bri ttle f rac ture of an LP turbine rotor in a power plant during an oversgeed-trip test, and is discussed in detail In Refe rence (3) .
This fallure also involved high hydrogen content in the forging,
)
The combined experience of all parent companies i of A-CPSI from 1950 to 1974 is ;
therefore two f ailures and a total of 10,929 unit years (32,300 rotor years) ;
of esperience.
^
l As discussed in connection with Table 1 the factory test pi t f ailure of Slemens and the failure of an Allis-Chalmers rotor involving high Hy content in the forging are not directly appIlCable to the turbine missile question, H owe ve r . It is significant that there have been no turbine rotor or disk failures in a power plant for units built by Siemens and KWU who are t he p r sources of design and quality assurance of units provided by A-CPSI .
. ?
e J
au_se-cH44.Marne earn myw rome.#vcb ER-504 10/75 f
- 3. TURBINE OlSK INTEGRlTY The following is information on those aspects of design, manufacture and quality assurance of 1800 rpm LP turbine diska and rotors which provide a very high degree of safety agains t burs t-typa failure of the disks.
3' I E*_5 Lil The configuration o' an 1800 r/ min LP turbine with 44-in i last stage blades is shown in Drawing TY 5.01.
Each two flow LP turbine rotor is made from a stepped shaf t with a total of 10 shrunk-on blade disks arranged in symmetrical groups of 5 The material for dlsks 1, 2, 4 and 5 has the l
l German standard designation of 26 NICrMoV145 which is a 3.5% N1 allov steel I
similar to ASTM A-471. Disk 3 is made f rom a similar alloy except with i
I 2.9% N1 which is designated 26 NICrMoVil5 26 NICrMcV145 26 NICrMoVil5 ;
Nominal Chemical Composi tion in t: 3.5 NI, 1.50 Cr 2.9 NI, 1.50 Cr '
0.26 C, 0.30 Mn 0.26 C, 0.30 Mn i 1 i
ond max 0.15 V and max 0.15 V i
i Kechanical Properties (max value for tenslie strength, all others min values) j At 68'F (20*C): 26 NICrMoV145 26 NICrMoVil5 Tensile Strength: Ksl 156 149 0.2% Of fset Yleid Strength, Ksl 128-135 114-121 Elongation (L/d = 5), % 15 16 Reduction of Area, % 40 50 !
I Notch impact Strength at -4F, Ft-lb 35 35 (Average from 3 Charpy-V specimens) '
NDT-Temp., 'F max
-76 -58 i
4---- .2_-- - -A - - , a-A- - - - _ - - --- --u, a-. u- - - -a,.L- - *_.- 2 6 4 4 m" ' - -
Y I I _ I I i ~
%: x ..: . ?~j - f j l I.
i i
. *e
- t T n-- -., ,.-,._.. --
^
- === 2 . w .. .o M V
, g,
=
i$ ,!
i II -k i
AI w !!
,- [] : f 't N 't !! P '
g a, ;
, . , t. t> <>
1, ,
f ,
. m I
, i
- :< i,
! , o-
- ~ .n. a .
. e, s ik Jh g db JL >
-l g _
. . x .. _.
, .., .; a-l .
\f, ,
I
(
.. N,/
- I {s
. c- J .hk r 1
.- - is r*
h Nbs$
4 h h0*
- / ~
'l I S
- .-I c, - n !<) !:g:1 > .
1' y 4.
Nll I ,s
, =s ;,t
<' .,: t**s W i.;;5;. :> *I j.; ;
s s
s s i h y -i {
- r- i , . ~j
.yt:.v I v P3 L E(
..y ,
8
,4 tf n_ j.,sE E I . '[ " b N' ..
( 6 1
k 2 ss i < 3 II;i
.- s
- 8 g WQ k
.y I .E ~ 3.,
l g
/ N ,, W, ! h t
%l,I jf
l/r
.;,up>
lse
~ :A >
M I M' E gs w : w -
a n.
F e* i .!.
!! .Y :M .~ 5 "
D'- !
{a;*, -m
,, _ i,-
.ge - if .t m e.e- 35c ..;. . - -w . . . 4 I
m _ ^; g 2 < . g ., p.
- +g g
.i ;,};l, 3- g
- f..J w -u v4 - -. , 14 i* LJ > l
"' ' } - of BE s
- 55r l ! C , ,
s si ? bb; '
- I Mb -
5:t ^
c.
/ .
. ^
I 1.Y *u'
>-d pa , ..
. . . , .. .. ... ... ,.- .- .. ,.. ~. ., .. .~.. . .
lQ; % 1
.ip g!
1 3 w. .. . ..:a 4.c. m , .
~ ---
ER 504 10/75 ALues.CHALMMEMes PowsM sev1ss7m
~ll' Ma WVC' 9 The compressive residual stress level of s the heat-treated disk forging;, as _.
measured by the ring core process (
see Reference 4) 4 Residual tensile stresses are not pet shall not exceed 11,500 psi.
mltted.
The dimensions in the last last s tage region, and the dimensions of the stage disk are given in Drawings MA 4.04a a d MA 4 ..
. 05 respec t i ve ly.
TFe average tangential s tress in th e last stage disk is plotted as a function of speed in Figure MA 4 16 It .
speeds up can be seen that at to 120% of ra ted, the maximum disk st than one hal f of ress at :he shrink fit is less the burs t s trength of I
the material. Thus, disk failure this specd range could only occur if in a major error the material is seriously defe:tive or if is made in design or manufacturing .
3.2 Sa fe t y Analysis of LP Disks The safety analysis of each disk design is based on the principles of Linear i Elas tic Fracture MechanicsFor (LEFM) .
the purpose of analysis of the inner portion of the disk (see Figur e MA 4.17) , it I
5 mm diame ter equivalent is assumed that a flaw of
' flaw size (as determined by the Krautkraeme r l
' ultrasonic tes t method) exis ts at the worst possible (hlghest location. stressed)
In addition, it is assumed that '
area has the " worst case flaw egeometry" as d fithe flaw with this st ned by LEFM theory.
a long shallow elliptical crack having a depth This is
- to-leng th ra tio of <0. 2. l The hypothetical growth of the assum e d
worst case-flow for an assuned number of abou' is then calculated f
. 4400* full s tress cycles over the expected In the case howe ve r , .% of dius, approximately 4400 cyclas approximately 2600 cycles.o'.neral design ofype turbines of this t(s tart-ups) are assumed; is based on
i t
,1-_ i I 'o ! 7. u e 1 & F m
i i
il
( 3_ o. g.) i
! ! t 250" ! 1 250*
Axlal Exit ,n-
- w n. .
Angle i t
, INNER CASING 4 1'.3749 %
-/
o
.l i E[Vl$lCN -
I r A .
dijo- O[ }1 n
. _0 _ s . li - f #1 I 2 ' S 1' 9 --
a- Wp3r,2"' 4 , rv l
y
.Dl ! ', . [ i
, -1 C! c i ) J' i
i I rv c,
\ l- / u di 0 \a;i :
y l
?
i l / /! 7 i
e i /i '
- . _3 ,,
s if, *-
,/ !
i ; g h 2p,fi,, _ o A
/ i,,
l o a
*1 !
i 1340"
/,
@J'[' $
i i t 9
=
wf \
, J. ' ci
.s
/[' / ,
/ )
j g 059" " g'9'J } '2
/
I .Itk'y i e Ys gg'y%1! s . r I : , i l , j s m~ m 1 k '
' r N i
,' m a OUTER CASING !
$ l 9
t I i 'm I I
' I
/ l i
8
/ ;
t-
)
I lo I L-N,
/
o u -
!i y !$5 e . ,,
d m ll f '. n I 1:: r i < 1, a ;u ..
- e t.5 . .l *.
u g G e i ~n la I } j!};{j 91.1" -4 To Axial !b i! !,
<; 3ej Center of i I
- LP Rotor *-2'-0.8 0 3' =
g+2 -6 -y .
. t h: srit s
t<
- b , dJ g W q ,
i m l 1 M 4 l N l l _ 1_. . l L_ >
- _ ._.L - L' Last Stage Region au,..c~.tuen.
m* a . v a r n e. .~o
,[ .
of an 1800 r/ min Turbine ; i + with 44" LSB's MA 4.04 a '
'J
I f i e tar)
~ #90 ~ H2
- ff%iui9V 0I i e e,c>,
1 bI I C! l 5 , dI ! 01 l 1 4 1 e no
- 4 to a e30 r
2 p , r20 - T 2 I 1c . 4 o I t
.r 4
.,_' to'
!7 . t 8 sto
/ ?Ae
. ? -, i .
- 7 r ;
's
'. A I'
~'
- 5. ;. ,. i i ,
+
Iitf }i 12 :, ,1 1 .a . i. l Iier2..' I D ! f f '!! I 1 a <: .? ca+#4 R i
. .- - > r y . .. .- _. L - . _ 3 sp _ -- __ ,
A Last Stage Disc ./,\ ALA NS =CH AL Me me # s ', NOTE: All dimensions of an 1800 r/ min Turbine ******""'"*""V are in millimeters. 7' - wlth 44 " LSB's
't. MA 4.05
1i!:! ,-l; I,!l, i;i !;f,' ,,!,. *i.;1 >:tiificLut!jlLIiiy + [It![5 !i' I
.' ;i l > l , ,"
n i t m 0 d n/ 0 ner , 0 E m n t, h0 4 i ec8 m n
. t a2 / i at 3 r m
_ me iD = V /
- 0 r
_ x m"d0 _ oed u^e6 , rd e a . , I,
- eea3 e e ' s * '
m~ e3 '8e,' ,a , d _ pae .i " e pl p ABS a" M x o x S e p
=
r _ p u :._w3z
- r (
,, p r\ Y v , ., = , A o
/
_ / r t - _
)
t 0 0 i
, 0
_ '- u 3 n - L - C l 5 e a 8 d c - i e _ r 0 e e p t = S u t a d 1iJ t en k a t i w i s ( am R/
,t D n r i f
< f m o0 1
o / r (>
.o,
- 1
*e ', s e h 02,.0 t 2 2 0 7
4 g 1 1 d n 0 n 3 ei , 0 e 3 en 2 r p/ t t S r S a
- 0 ee d0 .
e t e 0, l s t P i r aI s u R n B e T k _ m m u i D s _ i a x d e m M u
.- s s
e = s 0 A 0 em ,. t
. i 0
1 _ ., u o
. w.
% .a...
/
6 t _ -,. ..n _ 5 t u ai .. _ oes c ., _ n. s ...., o. r i
.v,
_ a se o.sr - mor .c..,.. _ _ o _ t4 mo... .es , u m .c * . _ - a m v *, 4 . ie.. . ew n o,,o _- mt i ,, v. _ e . - . m mt a.n O 0 0 0 e4 n.a S I 0 5 1
"c sf ooo a _o c . ,neLuw _Q._ucemc*, mmb0>r< e"o e e 2<a,eee amsos3n m~ 4.n,meM <* mu* " 4 S
> rr g .0I* teza T r men wn=ea- rm ac,7~3e'o *r 3gI* 3 3
- emta-
- E n7 rea 3c7 rwo
- i l' 3 ; .ii , f - * ,'1- ; I> ~ $
! [l: ,'P : it'
,6ti-6 i 1 ' ;l l !i !:! , ! ;i j (i ,t! i t!}trl
~;
~
~
^
.k
' t
' d -
er
~
ee pt
- sf ^ a ~ d e <. " t e . _
al t c
' r hos _ i e t
~ "Q Er 5
?,
0 n 2id 1 pe t __ l s a
\Y3/ oq, ._ _. e r gp e _ t 3 ^ -
o~ _ ons f i r - 6 k e " M~ ncv r oo n~ t el y t er L ~'~ ~ C ._ ah o pt t s~ e sd a c _ - '~ snf _ eu r oe t rh " S at _ - v 3
- - !- t 0
6 - 3
- 1 f
o e {t
- m z t - o - t G Il a' 4i t
r - l e n o > - a i h tmA
- w n
- i a i p vf l f p o, s A
/n/ F d eg xecn m .a i ufk
- i -
- 4 f sr c s uo A sl
( - +eu
+i si n..
o-
~. *& .
. s i.
s ta . a, -. .. c.
.m n .e
. ~ .-
no f i _ . .s o.
.,s .
. l -l '
f w a .o n i /,
) m, /'} u-
_ r e s. n ..,.
. t l t f r ok L
. d ps -
mu 4 ~ e i gvZ1-mrd - ue sne snh A .a t
/ -
2 so,an,n oa oa _ .E !0I>rIa1' wo,+on< r p2n, <w.@ n3 e <e*nI9 l o- - r-o o mre * "
; - * ' " [,
ALue-CHALMERS POWCR SYSTEMS. NC. 0 ER-504 10/75 ^ ( m life of the unit. The design requiremer'. .s that the calculated crack depth at the end of the unit's lifetim must remain below 50% of the critical crack deptn, a , for 120% of rated speed. The crack growth rate is calculated by the following general relationship: pa dN = CSK' (Cycles ) shere: a = crack depth (mm) N = number of cycles SK = cyclic stress intensity range at crack tip (kp/nm ) as = applied stress range (kp/mm2 ) C = a constant m - t5e fatigue crack growth exprenent of the disk material Q = crack shape parameter g(a)= crack geom-try factor (iG ) ! Fatigue crack growth data for di f ferent rotor and disk materials are obtained l in extensive laboratory neasurements as discussed in References (5) and (6). j t l i A few examples are Illustrated in Figure MA 4.15a. For the safety analysis of t disks, the upper boundary curves of the fatigue crack growth rate scatterbands i are used in the cair.ulations. ! The cyclle stress intensity range, AK, at the tip of the crack is calculated by ; i LEFM methods. The AK-range is linearly proportional to the applied stress
! range, ao, and depends on the size and configuration of the assumed . rack, a.
Therefore, SK may be represented r aK = 43*o(a) where for an internal crack g(a) is defined by /a x n/Q and for a surface crack g(a) is equal to f e'l . 21 x a x n/Q. The crack shape parameter Q takes into consideration the ; influence of the actual crack shap. on the cyclic stress intensity range. Tae ;
\ )
l' . N. C r uev 3..N,ve ,- 35.,9 crua, l ,f; * .y- , , e , c ,i )
- u. ,
,e Tgj P
/ *V .
e R 'C4 .'t'Fj Aas '. , gc.c __. c; ; 5-$ino-l , <*,s. o
,6, 3*
s
. . s a
( m 1 J ,) . . ,= Yo ! .$' l R[Vl$l0N 3:: -
'l . h.'
9 9 __.2;> : N ;y
!Gi i
.g.,;
+ ;
C e' , *: 0s! 4 i c . . r- - , - ,
., .. - : t :.a .:) 4. .
- i e -
g} l
. .ag- 52 j, y . . 4.
, ..: ,3s7 ts . . . ,: r m
-- , os: e 59s d.J av .
. +
4tr
)
._,/_.__.._ "
."v.;,ota.f
~
9a-
. - - . - . -- - - 4 Aa s
/ . - - - -/,6 - - . - . ..
~
;; t; L. , iE '. C 10 0 . 087 S Xi 30 $0 100 . 'DO /1 c - J Y' SPECIMENS C il- 2 *CL 1T - 3 T / ROOMf E MP ( AIR , / f = t0-160 eZ
.,.w, . . . : .
- n../ .anvr 1
.n i z a c-
,. . . a zan .
- ic E5l b h p/mm 3d 6m 3 5 NiCrMoV e02 79 5- 97 7 m p/mm2 j
,og70) _,
15 0 8 E e
/ 2 Nicr Mov ,
a / 30 2 : 5'- 6 3 h p-ra ' e, , a .
/
j ~, 40(; * +
,' l(1960 )
- i. 50 0 8 e
./ !
I
'" "D" .'
3 NiMoV l 5 '_i.
. +
s 4 , ,' ' e02 60hp/nrn2
. 3 ,ysc I. e d
, n.
./
A L
- g ' , ,
# ?!
- 200 .
E4 ~.ij 53 50 o NOTT l if isi y ;g -
; .... .: -_ + L.
y,.. .. - .)*w,ss.ii dt auses
- o f.
1 l G-2 **
- el
** KOU
<< pas
-:-2
!- 2 - 0 Temoeroture in 0.
l . .180 .120 -60 :0 60 .i20 5 E. ?55) a! v fi -240 12 0 to .12 0 240 '$ l 1 n Kn,o, a r s n ., e r O n.. _ . i U n... ..nt.
- t. , n_ a.t. ., _ L.d. s . l. - . . . . . . . . . .A v um.cas r, ur e....
en . ,. . .
.c GHAFT3 AL -D1;C3 AMD FOR GENERATOR-ROTORS.
l SOURCE: Forg.ings from Fatigue Crack Growth Rate utgantic Ingot w.th i and Fracture Toughness "'a m a
.'.ow a n , .' *r e en ,vc
*aa >
140,, Diameter and of 2 to 3.5% NI Steels 881,000 lbs. Weight
~~
Part 2 MA 4.15 a
ER-504 ALLIO*CHALMBRO POWER BYSTGMO. INC.'S 10/75 (' stress range Is calculated for the start up and shutdown loodina conditionc. - of the disk in question. in order to predic t the cri t ical crack depth, a ,, the fracture toughness, K IC, f the disk material must be used in the calculation. In the case of a long-shallow elliptical crack inside of the disk, the critical crack depth is given by: K2 iC,q "cr " "# ( mm ) t2 where: K = lc fracture toughness of rotor material t2 = maximum tangential stress at bore surface The maximum s t ress , o g , which is used for the calculation of the critical crack depti, as described above, is derived from peak load service stress and internal residual stress of the disk forging. t Due to the high quality manufacture and quality assurance of the disks , it can i generally be assumed that the disk bore region is free of flaws. However, the safety analysis assumes during turbine operation a possible formation of a long shallow semi-elliptical crack wi th a/2c = 0.2 Intersecting the surface of the small ! i locking pin holes as shown in figure MA 4.17 on page 15 and also ! i Figure NH 4.02, page 28 ! The critical flaw size for this case is calculated by: l g2 cr " n q2 t3
- J
I ! l Au.so CHAL. MEMO **OWEM SYmTNMB, HMC.S ER-504 10/75 r 3 i ( \ The value of e 5 a n e ress pa M m m un e c ng pin holes t3 which is influenced by plastic deformation during the first (factory) over- . l speed test of the rotor. The plastic deformation results in a cor pressive stress which reduces the value of e the calcW ated C astic stuss t3 valuu as Illustrated in Figure MA 4.17, page 15 . I l The safety analysis starts by assuming the critical flaw size. Calculating backwards with this critical flow size-must result in an initial fim size of
?5 mm equivalent diameter. For this calculation about 4400 start-up cycles are used. This Is conservative because a very long operating time would be i 1
\
required to produce such a crack. lj The fracture toughness data used in this calculation is based upon extensive tests as discussed in References (5) and (6). With tests performed in accordance with ASTM E-399, f racture toughness K is measured as a function of lC f temperature. A test data example is shown in Figure MA 4.15a, pace 17 For the
! safety analysis of disks and shaf ts, the lower boundary curves of the K lC i
l terrperature data scatterbands are used in the calculations, i For correlation purposes, additional tests are performed to determine the f racture appearance transl tion temperature (FATT), and the nil-ducti11 ty transition temperature (NDTT) per ASTM E-208. On this basis , the NDTT In the -. most critical region of the . disk (the bore) is speci fied at <-58'F or -76*F _ (depending on disk material) which is far below the minimum operational ' temperature of the rotor. i N j l '.
Au.m-onAL.Mano r>owan evnernMo.m4 > 10/75 , - ER 504 - it is well known that inhorogenei t ler, such as segregation s t ringers exi s t in forgings, and extentive work has been done to study this subject on turbine and generator rotors up to *.he largest monobloc rotor forgings ever produced, as described in References (5), (6) and (7). The studies include microf racto graphic examinations in the zone of maximum inhomogeneity which occurs ' between about one quarter to one-half radius of the rough machined rotor forg- ; ing. Research to date indicates that the anavoidable minor allow segregations and microporosities in modern forgings have no affect on the mechanical properties l l important for the safety analysis, fatigue crack growth rate and fracture toughness.
- 3. 3 Manufacture and Quality Assurance l
I Each LP disk is suNjected to a comprehensive and coordinated program of design, manufacture and quality assurance to ensure its reliabili ty through- i i out the life of the turbine generator unit. The basic principles, materials, i research and developrent and other aspects of this subject are discussed in l detail in References (4) through (7). Fol l awi ng i s a b rie f s umma ry o f t he I l key points applicable to LP turbine disks of the type described herein: 4 !
, The disk forgings are produced from vacuum-degassed alloy steel and are heat i
treated for an optimum combination of high fracture toughr.ess (throughout I the disk volure) and calibrated compressive residual stresses at the hub bore surface. Each disk forging is examined by ultrasonic testing as follows:
~
In the rough as-forged condition before heat treatment i l - Premachined with contours c: 'or to heat treatment for mechanical properties
- Af te r hea t- t reatter t for mt;hanical properties J
~
A ALLIS=CHALMERS POWCR SYST* CMS, tNC & ER.504 10/75 '2I*
< T The ultrasonic testing equipment and technigtes are in accordance with DIN Standards $4 120 and 54 122, and can detect and reasure flaws as stall as one em (0.04 inch) in equivalent diameter.
The evaluation of ultrasonic
- nspection is based upon reported indications as follows:
- 1. Isola *.ed single indications of an equivalent defect site of 0.2 Ir ches (5 nm), as per DGs diagram, and larger.
- 2. All isolated Indications causing a decrease of nore than 10% of the back reflection-.
3 All indications of the linear type or area type, as well as clustered indications, regardless of the size of the single indications in the cluster area. I
- 4. All indications of defects located with a 2 inch (50 nm),
l zone surrou.1 ding the axial center bore. *
- As indicated in Figure QA 4.Olb, material samples are taken from each LP- turbine disk f orging near the hub bore surf ace. The purpose of these samples is to get representative test material for the determination of the ni l-ductili ty transit ion temperature (NDTT) as a fracture toughness criterion and of the 0.2% offset yield strength as a strength criterion. The results I
of these mechanical tests in combination with the UT and residual stress
\
measuremont .esults are decisive for the acceptance of the forging. l Before machining out the hub, the followl..g test results of the fully ( j
i i l l Residual Stress f Measuring Points Ultrasonic Tests ! (l001 Volure) i l j
\ ME ASURIN G
- PLANEI REVISION i /1 N I
.i **~
REFERENCE POINT
/- ','?A V k, i gg
__ l J- is 5 0 7% t' / , l (sE RIAL NUMBER) b asi,g r;, ' 120' , j I C I g t, _ _ _L l _ __ . . _ C i I \ ~ !
\ {
/ '[
( \ s ME ASU RIN G l O d tj . PLANE 11 j 3 1 2 { (HE AT TRE ATED CONTO9 , ROUGH MACHtNED CONTOUR .
. ,_ )
/
l l (Sample Position, see Order Drawirig) l Material Sample for NOTT and 0.2% Y.S. Tests and Determination of Chemical Composition M iub Bore .
'l , . .
1 4 w j sfi w' 4 t E
,, .' g .
j gg lij l',N' ,-l . 's ,.
- i n Isa l 3 j :.ia
' .' ' .'r -
( ./,; '$
~
i h5 Y=k ] ,-l ' ' ' >.
?b h hi ).e
. ' :; ?. : f .i .*
m.; a .~
~ .
- , 6 ', p
- ,' ci- :5 . w ,
'.. r \. ' 11 ,
Y i 5; mu - ___ a-I Etched Cross Section Through a Small Disk Forging i l 4 W ^
-, M
! i4aterial Testing of /\ LP Turbine Disks , _ , , J-l
.ow.a.v reuss e b be
ALLtO*CHALMERS POWOR SYSTOMS NC. ? ' Ek 404 10/75 i heat-treated forging are obtained and scrutinized:
- 1. Resultt of ultrasonic test, covering 1003 of the disk volure, and with documentation of all indications as required by the appropriate material purchasing speclfication.
- 2. Actual chemi;al composition of the forging material at loca-tions illustrated in Figure QA 4.Olb on page 22, 3 Tensile and drop weight (NDTT) test results at location i
indicated in Figure QA 4.Olb on page 22. Af ter machining to the dimensions of the order drawing, the bore of the fo rg ing is magnetic particle-tested, ! in addition, residual stresses are i measured by the Ring Core Method described in Reference (4) and ll.lus trated by figures QA 4.02 and QA 4.03. By careful control c,f heat t rea tfren t , I i desirable residual stress characteristics can ba built into the disks. These characteristics are verified by measurements at point I to 6 in Figure QA 4.0lb ; en page 22 before and af ter machining for the f!rst of a batch of similar disk l f o rc i n gt, . Subsequent disks of t'le same batch are checked only at point 1. 3.4 LP Rotor Assembly i Figure NH 4.01 shows the assembly line for low-speed LP turbine rotors ; j which includes shrink fitting the disks and couplings on the shaft, as l well as inserting the blades into the grooves of the shrunk on disks. To I date, more than 200 disks have been shrunk on LP turbine rotors using the tecnniques descrioed herein. t ,
. _ _ _ _ _ _ _ _ J
l ME ASURING L AYOUT ME ASUREMENT RESULTS l STRAIN G AUGES l (2
\ t tkU ;;~~~
, ot(Z) - 02III ! w= - 5
- REVISION i T ,n,,. ,3. ,en,r g A 5 i C ) 0 MILLING DEPTH Z -= l i _a \ME ASURING POINT / i I U u----_ > ,
l 9 i ! l l N l // j l .mi
) ., Nj g la-
- E 4, c / ** D e
; n i! '
E*e . A 521
.! 7 j ;. RELAXATION FUNCTIONStK : K2 (0 = 14 mm. t = 2 mm)
&.f1.2l
?
i * -t ' J g" gf gjj BASIC EQU ATIONS FOR STRAINS c,; ob; Oc =a APPLICAll0N R ANGE
/ 6
- g i g I -
{ j f? I*, Os .b.c
- I t 's,b,c,l I K1 K2 ,E,41 g 0.3 ;! '
4 7
*$l', , Kg . !
$ 01 i
fp 2}$ ;E 3 WH E RE: K g ,K2 * '2 (Z,0,t, STR AIN G AUGE GE OMETRY) E m k2 i\ f f
~\.
0.1 - E = Mfl0VLUS OF ELASilCITY
$_ 0 -
-- N - _s - ,
u = POISSON'S RATIO E -0'1 E I ! 5 a 0 5 10 15 20 mm E MILLING DEPTH Z t Ring Core Method for Determination ,u,,, ,, ,, , \ of Residual Stresses mam mverma m QA 4.02
. . . . - - - - . - . . . - - - - - - . . _ . . . . - . - . . . - . . ~ . . _ . _ _ - . - . - . - - . -
l -_.
.- 3 m i e
Y> %' .Y e- ; i l
\ , 3
! . i1 l l l . b.. i REVISION 1 .
/ % Wl 3-8, ,
1 0 .' b t C - i s d e Q1 _ 6_e*I w. e ,
- =* '
1 l R r l < i , 6
; 0_. i 4, cov.nw.m_r,.,m 3 e . ..
\f' / f
%* %__'s_ x 0":,: . .s'
! . , .:.:+: 2 i i M: $ Ns N .2:Wi!^: l \
/ L:
.oT : .:
i i )/ ' t ..
+N N\
< oR\*:):$. - .
l jk 1:. l ' N . 3,; $9 l.\ (: > og s ! t; s 0 l k\ Cf .;,'s' .\ ', s 3 ! \\ h g (N , ;;j { ' l n \\ / t.. x' . ' ': : 1 1- A\ f? l .!! oa- N ~ l oR !' y : xw. , :
,[';"A 5, _
oT
,1 W;. N ULWMM83 is
;* : 4 b!
* * .s - 4 6 4 2 0 2 i fj,h RESIDU AL STRESS, kp/mm2 f.- 1.422 = kii) 5 s
3 St.
~ ,1 -4 o
,n -
<* es3, 83 #' oy-j$)~$
o' RESIDUAL STRESS BEFORE MACHINING $ g .6 - # s A ~f j (TURBINE OtSK PROFILE .) lll a* [ w
*f' i
!# ,ii # .4 f N o RESIOUAL ;"RESS AFTER MACHINING g N, j !
, \
G E2< i (TURBINE DISK PROFILE ) aT 0A gi* f .
\
0
/ '
\
l r Determination of Residual 4w.-o- w a= E Stresses in an LP Turbine Olsk
" "'" " "* "" " " ' " [
4 j QA 4.03 L_________________________._________.___.._.___.___..______.__ ._ _ __
o
-g%I
~
\ .
# \
[ 8 \ s 4
\
g
/ ' ,
v yg W \ \
\- g " ,
~ \
\
'\ "
b' '
.f
\ JN t*~
\ '
' ~
\ -
x \
^
4 *
$N
\
~
\(.;
\ R, l !l . h_ j lI s ' [fA , S. ' (O) n g, lpr
,t ,. ,
h
- j kI
.s b m, .-
N' M.;'O
/
lq t Ye \ kk5. N' t .1 . 1
,t '
j% *Fi
'g j 751
- s
^
')
\
^
gW .' \ f;'s Production of LP Turbine g*llr$"* ' Rotors with Shrunk-on Disks NM 4.01
-w---, .-y ,-,.-e-- e, -,w,v--wg-m-a,.,,-----,-,-..n.,vnm--,-- _ - , _ - _ . . . - -
.__._ _.~._ _ .=.___.___-__.._.._____e______ -
i Alum-oHAtmarme paow=m everces.wc O l l ER-504 10/75 l C 1l ! 1 The shrinking process is perforraed with the shaf t In the horizontal i l t position, and with rotation of the rotor during the process, which provides the foltowing benefits: i l t , I 1. Defined heat transfer f rom disk to rotor, even during the initial
- I
- stage of the shrinking process when the clearance I$ $tiIi greater than zero because the disk always rests with its weight on the shaft. '
- 2. Exact positioning of the disk by the axial compressing device j during the initla stage of the shrinking process when the f clearance is still greater than zero and turning gear is in l operation.
3 Easy correction of rotor .run-out by stopping the turning gear l! j i i with the shaft in the correct position during the Initial stage j of the shrinking process for a calculable period of time, t l 4, Axlal and radial run-out checks during the entire shrinking i i process wlth shaft in Its operating positlon. l' l l' i i A shrinking stand and heating oven are the rnain facilltles required- for j shrinking disks on shafts, we shrinking stand (Figure NM 4.02) has two I rolling _ supports, one with a _ turning gear _. and the other wi th helght adjustment. Before turning the rotor, the helght of the second support has to be adjusted in such a way that the rotor rests on the arm bracket, which is rnounted either on the -lef t or the right rolling support, l depending on whether disks are to be shrunk on the generator or turbine ! end of the shaft. A hydraulle support device is used to prop the rotor- ,'
~}
i
+ e w w *e -- w + eer,- -oe w= m w ee n ver ---w- w+-w+=rr**,-w w- e - W ,-wervv'*
i l . Ai m 3,acket 7 Retainmg Oracket ) u l .
~ ~ ~
~ ^
. i 1
- ~
[ - -, i n
- S,N,,D .,
( l^ f". }
~
Railing Suppor t d bl
, Rotirng Suppor t i_
pgg 4'th furnina 1 [ _ (He%95t Adiuttable) I-t
/Geae
[OI%ID' t - Hydrdutic d " A oal Compinong
?
bI I I 1 ss ppor t --- - --
,s
] Device l
s s s s sss ,s s s sss s s s s s s s ss u s s s su u sss wx s sssss cl ! dI l Stand f r Shrinking- n Olsks with Horizontal Rotating Process t) l
?
I lypical Sequence of Shrinking 8 7 6 1 2 1 4 9 9 10 3l
-Y 8 A g b O WS VW O M
~f My 4 d 4 @i4I W[
d y N g ti 11
'i d
s f !1 ;a O o 4 p h y p} hbEt&dQ d i t
+l il I;
l l L Y . I
! : l 'j m **.
' l i - -__ y ! ; !
. i i i i j j 'l t-:
5 TYs)M 5, ' P t? W il % yt b w y l
- j R
4'hNrM p**1 =UNOUT CHICKSm'"' - ~
7 g ;F
'/4 i f 7 RADIAL OtRECTION
- ; 4 " % AXIAL OtRECriog g i .
- j
+
{?
> 9 LOCKi4G PINS AT Two tiAL7 Ri4Gs SMRINK RING FOR SECURING TNi Two
!f I Eg TW4 CIRCUMFf Ritect POR AKlAL LOCKING i.1 l A NALF RINGS OF tnt LAST DISK 3t I _-
.'s 2 $ - 5 5k _-- . - ,- /' / p/
i; 27 0 ;/\4 j = t,/ n
. _ r _ A4,xw/ .
. /N/ .N+w, 4
. . , ,. 2
$5 i55 / j 'O ~n
!//lI// Of f All A
//// Of f All e a
Shrinking of Disks on LP Turbine Shaft au.
,-ow...
.c~ w.
.r.w. m +
A[\ i
.- _ _ . _ _ _ _ _ _ _ _ . _ _ _ . _ _ ._ _ 0
T au.m-osA4.marne mwarn everous,wea 4 i E R- 50'4 10/75 3 i g l 1 3 while one of the rolling supports is removed to allow a hot disk to be
- slipped over the shaft end. A retalning bracket at the other rolling l 4
I support prevents the shaft from tilting. A compressing device pushes : 1 } ! j each disk against the appropriate axlal posi tioning shoulder of the i shaft whlle the disk cools down to ensure a shrink fit at the correct i 4 axial position. i i 3 ! i i ! [ The oven Is of the hood type with an electrical haating system which i 1 l l transfers heat uniformly and accurately controls the disk temperature. { i ;
! Before being put into the oven, the disk 15 mounted on a loading rack ;
1 i f and leveled so that it lies in an exactly horizontal plane. After i ! i j heating to about 700 F, it is plckad up by the crane, moved to the ! i l j ! shrinking stand and slipped over free end of the shaft. Then, the- j j i . j , rolling support is remounted at the free end, and the hydraulle support { i 1 at the middle of 'he shaf t is withdrawn. Next, the disk is noved into 1 j lts final position and pressed against the appropriate axial shoulder l } by the compressing device. l ! I i ( The shrinking process really starts when the disk is in contact with the shaft and the crane hook !s disconnected. A large heat transfer takes place between the hot disk and the cold shaf t at the top of _the shaft. The shaf t is turned alternately in- both directions to provide a uniform. shrinking process, checked by run-out measurements. Af ter about 45 minutes, the entire shrink-fit becomes zero, as Indicated i i
.by equal run-out results for shaft and disk. While the disk cools !
q J
AL.LIG*CHAL.MERS POWEER SYSTUMB. INC 0} ER 40k 10/75 _ l i cown to about 120 F, run-out checks are performed wi th the shaf t being rotated at about 1.5 rpm by the turning gear. This procedure for shrinking on a disk takes about one day. i Af ter the disk and shaf t have cooled of f, five short locking pins are installed at one disk side between disk and shaft, and two hal f rings are assembled to axially lock the disk itself and the five pins (see Figure NM 4.02, page 28). The shrink fit, which resul ts f rom the disk bore being about 0.1 inch smaller in dia-me te r than the shaft. Is sufficient to hold the disk firmly in position at speeds . in excess of 120% of rated speed. The five locking pins are only an additional protection against circumferential displacement of the disk. The two half rings , for axial locking are caulked into the shaft groove so they cannot drop out while the next disk is being moved into place where it holds them in position against cen t ri f ugal force. The two half rings for the last dlsk are secured by a locking ring which is shrunk over the half rings. A stop in the outer dianeter I of the half rings holds this locking ring in position even if there is no shrink force. Numerous run-out checks are made to ensure that all disks are properly shrunk i i on the shaf t as indicated in Figure NM 4.02, page 28 The maximum permissible l l radial run-out at any disk ci rcumference is about 3 mils which is about the same { i as the allowable run-out of the shaft at its center. $1nce the run-out check in i axial di rection determines the correct fit of the disks against the axial shaft shoulders, they are requi red to be twice as accurate, nanely within 1.5 mils.
ALLse-CHALMene POWER SYSTEMA, tNc. C' E R- 504 10/79 g , i 3.5 Baiance and ove rspeed Tes t i Before delivery, each completed bladed LP rotor Is balanced. and subjected to an overspeed test at 120% of rated speed f or two minutes a t a minimum temper-ature of 59'F (see Figure TT 4.01). All other turbine and generator rotors are also subjected to a 120% overspeed test. I l l l l l l i Y - J s
r 5 fl _ I 1. ms , dr~'I 1.1 oj
~, . s~
.e m e-se i n, b;hamWQ& tDJ: W a%ud 6 J = m
,,,j--..,.
l REvl, S10N 7 [ AA.h ['~pME'Ag;g'.^ r.If*; NI :t
*F7 O ;f "r"7"V4 -
d e 7 7! C '. a' I. LI h, b~ I d Qq] k & i _ _ [_ L CE3 4 N a O L@O f Balancing Planes of an 1800 rpm Turbine for LWR Applications F ACTORY & POWE R PLANT
]. Q FACTORY [ POWER PLANT 4
s
- 9 j o.eeef 4..
, N r. - . . s i g
~
D . , h.e ur s ti. s y vu ..<. (, . ss ?>j g> ! H! !i m ai
!4 1,! it y -
s iy t 0-il 16 - ev 3 56 I is o< ;si a '5 _
- .f n.
Balancing and Overspeed .w. .c~.t u. \ Testing of Turbine Rotors * *a"*""M*"* TT 4.01 _-_____--__._______m. - -
..-a._,._m.,,_.-..-.,,_m-.- ,m ,m. . - - - . _.,.,_,,m_ _ . , _ _ _,%,,
. . w, , ,
^
i AL.LJen.caHALMEMe MEM OYmTEMo, ENC. O - E R- 504 10/75 _ s
- 4. PROBABILITY OF FAILURE DUE TO OVERSPEED 4 i
4I introduction i This section contains an analysis and calculatic:n of the probability , s I o f a > 12 ^t' s pee d e ve n t . Although such an event would not : necessarily result in an LP rotor failure, it #5 conservatively 1 I assumed for purposes of tnis analysis that the f ailure probsbili ty l
, is equai to the overspeed probability. I The design and operation of the overspeed prevention features of an
/
j 1800 r/ min turbine-generator for Light Water Reactor (LWR) applications are . described in A-CPSl's previously issued report *E415 (8). The reader is , 1 referred to report #'6'S for basic descriptive information which will not l.. l be repeated in this report. However, fc.r convenient reference, a cops j of the speed control system schematic diagran NC3.06b is included in this + report. l The methodology of the failure probability analysis i ad on IEEE Std 352-1972 (ANSI-N41.4) entitled " General Principles of jh j i i Reliability Analysis oi Nuclear Power Generating Station Protection Sys tems." (9) The s tudy cons i s ted c f the fol lowing bas ic s teps : 1 i analysis and preparaticn of the reliability block diagram for a >l20% speed event. Establishment of candom failure rates for individual elements of the sys tem. l {\ Analysi s and preparation of fat.l t trees leading to a i
>l20% speed event.
(- J
mu
!$$ HOI kk
-m,-,
* ' + + +
L ,' h_ ~ a
;t;-
er t !s-:
,E5537 n; 4
l I
,,C,; Yr /.
a
=_ ),
,, prr"X ,
l-
~
~4 (I C i[;g gtif gg s
[ _ j . . ... .- l~- l i;Itikm- f ,3 r :; r -- f g _ .i
- m. .'_ . _5- c ., .
s v,7. - _
~
. h h f>=
.,- >= -h ,5 I - g- -% ,,M
-/
r f- , u g Z$$$$ EE m:;.lo l' C R* hi h 2 _u ~) # _ _ -_a'I I
!)f"!
V I
?
![__
y u g 5. -
.~
y'
- 312 ~, 0 ? 'J %
{, ~~ ~ 7
, , d$- .
l I [7 - - - ' Q fjkI55III
.~ : : : : : : : ] % j Y fI N .- - :
C-- : P-.----
.3, 3333hhhh <!. /~ I W
,.1, bS z : h s. :E m
)-i _f h 5h h f;I
- v : :t
_/ }N ._ _ b . _ . _ ~ I
$h2:h3 ';[
I i 3- r g: l 61
} ..r; ll ,y- ~~-
i a j$ _ L 1 ji,.I
, ><li
- m. een *** *
- e M i
- ucy- l _ -
3 FN w; e l __ / x
,K \ l
/
w, - I ~. j P W"~]L i -[- f i c %i ; uw .-_vJ M; . w , N ce m a is O a s/
. - - - a .-.. ..
m-. a . . -. - m _ co,.. .-. _- - _ _ _ , y~
- i!H ,
p= , u 9 Q0- ]%% 'Wh L_
===$
in g:: _ mf
- Tn ,
Ehh E
/
,e l -
A
\ $
1, . . [
' J i j, A [ [. 4 fg [ l !
p , jo , - i1 y< J==~~r s@ g . _.: r a
.> l a
- y. - - ,} l L m
AL.LJ8-CHAL.MarRn POWKM 8419YEMB. NYC. ) ER-504 10/75 ~-- Preparation of a computer program and calculation of the >l20% speed probability. Qualitative analysis of common mode failure possibilities.
- Review and preparation o' conclusions and reconnendations.
Throughout the study conservative assumptions and methods were used consistent with safety analysis such that the final result may be considered a safe estimate of the probability of >l20% speed failure of an LP rotor. 4.2 Reliability Block Diagram i The Reliabili ty Block Diagram NC 5.01a shows the overspeed prevention system of a 4-flow turbine including the Electro-Hydraulic Control (FHC), the Mechanical Hydraulic Speed Control (MHC), and the Overspeed Trip. Only one of these three systems is required to prevent a >l20% speed event. All three syttems form a one-out-of-three* signal to the control l salves. Each system has built-in redundancies of certain components con tributing to high reiiability. I I i 4.2.1 Electro-Hydraulic Control i The EHC receives power f rom one a.c. power source, and ei ther one or As used in this report, the term one - ,* -of- three means prope r operation of any one of three system elements is sufficient to prevent a >120% speed incident, and similarly for all terms x-out-of y. L j
4. Q
;r' m3Ae 1 ,,
-,.; R --(.4)2
~. T.. w l 0
a
'J ]!l11" 2!J j ,.
4u dlJ j
- i-6 QL-e a
e, l Il d f'-2 tif -
, !il I =
p 2ll i$O $<. I T r,5) *l $?<j!
;2 17 ~
<3 I
i A O2il 1 e~T,o w s & V Di r-ii#--:iE en s,fr' e , ;ip -, w x , r, <
! i *$5I M
l
. , Q, ;
5 l 7 in --4 If$g_< 5E L$ 5 -A E af r7;
-p .c , ta f g- 3::;3
;5 Q
a y o 1 l d' T
-h i
?- # +
O - O - O I! I h,ltf #- wI.I.l n!. si I!- e!;j I!. lte!.l irq-ll t!. l a .c t
.r ,;c n gu ;; 3rY ir i,
_ _ *b M _ M, - M
; I I i
,.' .T lI i, !!l.
H. i j
. ill. '
- Y i t I
l- i ll!ij! t .sW!iliF il?ij "
- l ell rili*
i l
!!! j il! l i
./ i
!if k
( hs- ii i; T r n ii Fi L,u'
',i i a!'.-- fil II.rj!! {M Q..
.: I i:I
'I j .iJ.L 'I ,' + n;=W[li; ii -
Iip 3 fiyi j} !!p8 I;li: gp li lIII
.i' .i !.!
'i 'I' ZK g '! ig!
- 's i r . ..
b $ $.. eq Ii d
=e jj
- I I
- =. ::==.~.c.~. .:
'* "r'J."W"."lJa?M'.
Ii ao um I
^
A1.LAs-cw-dA1.MatMe PowatM sysrarwas. avc.- E R- 504 10/75 / T two independent d.c. power sources depending on the reliability and characteristics of the primary a.c. source. The internal EHC power i supply is formed by a one out-of- two supply from the a.c /d.c. converter ! and the d.C./d.c. Converter via diodes to provide a continuous power supply for the EHC. i l The speed nuasuring device is a two channel system w.th automatic detection and alarm for f ailure of the primary channel and automatic 4 switch-over to the back-up channel. The steam admission control
; wi th electro-hydraulic converters i s coi l t as a two channel system with an internal s'4pervisory subsystem, if a channel fails, it ,
l will be swi tched of f and an alarm given while operation continues f l l' on the remaining channel, in regard to overspeed prevention, the follow-up pis tons which f orm the control fluid pressure signal for the f control valves have a redundancy of one-out-of-three taken twice. I 4.2.2 Mechanical-Hydraulic Cont rol The MHC is a one channel system positioning two independent follow up l l pistons to form two separate control fluid pressure signals for the I two EH converters. l 4.2.3 Overspeed Trip System The overspeed trip signal for the stop and control valves is initiated j by two independent trip bolt and releasing devices in a one-out-of-two ! system. Also, the proper function of only one of the two main trip valves is sufficient to close the stop and control valves. The ove r-I 1 ( j
acua-cnALMene poven everama wo. 2 ER-504 10/75 ,r , speed trip test anc resetting device and the changeover device operate only during te.st of ths a system. During normal operation and a real tri p of the turbi".e. these dg vices are inactive in their safe posi t ion. 4.2.4 Stop and Control talv_s To prevent overspeed, the cont rol valves and ext rac tion valves receive a one-out-of- th ree c los i ng s i gna l from the EHC, MHC a'd trip system. Four HP control valves are arranged one in each of four HP admission pipes to the turbine. An independent HP stop volve is located immediately upstream of each HP control valve. The four HP stop valves and four HP control valves are in series formi..g a one-out-of-two system in ear.h of the four admission lines. No interconnection between the four separate HP stop valves and control valves is required for throttle controlled 4 operation wi th full-arc admission of the HP turbine. However, two inter-connections of a pair of two HP stop and control valves is establi-hed by the heating steam supply pipes to the two MSRs as shown in Fig. E 5.l;9 i The valve arrangement wi thout any in terconnection would be the most reliable j i system because one stop and control valve in one admission line has to fail to drive the turbine-gene rator to a > l20% speed in a very short time. Having an interconnection, especially between all fou valves, would lead to a higher overspeed probability because a failure of one of the stop valves and one of the control valves which are interconnected produces a
>l20% speed event. For the failure probability, the cons.ervative i l i
approach was taken, that the heating steam shut-off valves between t e two i i pairs of stop and control valves are fully open during a load rejection l 1 or trip, desoite the fact that they get a closina signal in case of a I unit trip. l N J
J jf d f MAIN STEAM SUPPLY jf jf i f 4 HP STOP VALVES i i i ! REVISION 1
/7 5U) mW IO !
bI : i Ci i HEATING STEAM SHUT-OFF VALVES - -- - a i Y / i E N MSR HEATING STEAM SUPPLY / ' I n 4 F HP CONTROL VALVES j r 1 6 1 4 ,i 5I 1s
- , i
i .
;5 i- 5iith;
-- HP 3 : -
1 E3 II E TURBINE f !A $il
== .1; i *!i,-
I 9
!<~'l .
d :s E } a2ir
!. !g5 eIk
! E l 7, it s i:
$ t4 29<
2! d HP STOP AND CONTROL a u .. .c~ 6 ~. m. .A/\. l VALVE ARRANGEMENT m. . . .. r. . = p i LWR APPLICATIONS' r .:
; E 5.119
ALLIS-CHALMER8 POWER SYSTEMS. ItMC. C ER-504 10/75 -
! i Each admi ssion pipe to an LP turbine is equipped wi th a but terfly-type stop valve and a butterfly-type control valve. A closing of only one of the two valves in each line is sufficient to avoid an overspeed event because there t is no inter-connection between these admissions.
The Schematic Reliability Block Diagram E5.120 shows the same overspeed prevention system as in NC5.01a, page 36, excapt in terms of symbols for tre l system elements from A to V and the component numbers I to 36 which are used in the computer progran to calculate failure probability. This diagram also gives a visual impression of the very high redundancy of the signal to the control valves, and the importance of a high reliability o' the stop and control valves and the trip system in regard to overspeed protection. i 4.3 Failure Rates l l The following is a summary of analysis of the system components and elements . 1 which was done to define the failure rates of each element. In general, the approach was to calculate the failure rates based on actual operating experience using a statistical confidence level of 95%. For example , l operating experience with stop valves similar to current design totals 5.3 x 107 valve-hours; and during this operation there were two failures which could lead to a >l20% speed incident. For a confidence level of l l l i 95% and two observed failures, the Thorndike chart (see Figure NC4.12)
! Indicater the failure rate should be calculated by assuming 6.25 failures; therefore," the failure rate, A= 6.25/5.3 x 107 = 0.117 failure per million I
hours. I m j
DVPC41%f 5'YtSt .aYI % Fit R 414 =P sf oe vatyts i rc s 13 26 9'A L P :, O' a t '. f 5 I 7 0 2 11 16 I Ci4 we togf a06 w AhtlI 70 4 19 12 014 t p ^ 0 % f RO L v a t vt 31 f C 4 M - l'i i ~~50tL O* ur Pisf 0%s 11' entit a r i as ni 616 5 0t LO A UP P $T ONS & 6 Baf fl A t ') 16 ft ) f at t Qe t.tF #1$ f 0%$ i 3 0 A f f t R v it a' 11
'as sOu.u tr a s m u saf fin i > ia m a [L(C TRO M9 C#autlC (C%vt Ril# ' a%01 'l il ,
= *; v=C f pt L O A vP riST 0 418 % 0 2 56 I k h] h . , . ,
Yf =4 %iC M v 0 a d ut t; f .' 4 f A OL C OU8ti f f e us,m n,.o. 4.. e : cc ,a u = ,. ~~ _I' fw. ipI ( E '.C % ? n g t ; cypt ['T t ,
<0 vw - m e vm a^c a v'" ov'a "
- au 2 g l % v a C % E T R$C a f T., H 8!% E i* 45 7 l
_ i i aC :: au ec ec , a%viafia i a%o i, ___ 83 ) P2 4t R SC JRC151 'O 3 __j
# *2%CI Ovi A DiviCf va t% f a ip >. a t y t i 2 % D }
ldl _it s [/ M , O v t as't t 0 Intr f f 57 4 43 atSif f t%G Of v'f E r g $27. SPE E D iei i 2, a s , s.n o .... .m i 2 % o ,
.a i .i v: .iuas,%covun iau 7 j
*s
- .A,s s
, 7',, .
/,- . ', x% - ..
l
\\\
W' V? y/%NNN'N '1'% 1 3;[2T Ti '/ ~' . - h u-2Q
%aa f i
,f
! E/ :/
Qq x E l l t ~ , - - :a j l : ts _x > iJ ~% % g o ti! - -legrz;rrl, e,
.~ e Ny ,
- - - + -
x m w-
., , i i !
5 $4i N 8
, '--+3_3 . n__
27A50 1 '. rfr '- \ D2v' , -
~r.
;D3r -
l . >--e f D4r g m
\
- : .si.e ; 4jesi..i i i r---+-,- ~ to , i la r i n i
K_w\:G2I- n- e iG1l> ,
~ ~k '
l
.s.
<. N 'U s 'Ht: ~f prl lH2l EdAN%
/
,' ~-e 1
l
=?
\L%
-\-
'-e l
4-- - , l 51 ;; . ,s n i 1 r--- - --[ O i \
; -4 int yi l' . ]
,'.= n2
.s%
)
i . t'_.
- - . -g ,9 0
r-t 4 ' Lx -
- ? ! N. /
rs: l -
/N x
- .- m ,, ;
, i Im % ,/
I l I A .? i 4 l \d/ qp - -,
-j I' gZ,
, 'i
.\
uti 1 M 'I i ' i,, j ra i,4 IQ L___ . _ .-_. .@ 1 I L__ ' _ . J SCHEMATIC RELI ABILITY BLOCK DIAGRAM WITH COMPONENT ^"'" *^4~*aa en a . . re e. ,~c d NUMBERS 4-FLOW TURBINE FOR LWR APPLICATIONS E 5.12 0
4 i d' i o. , e, .o, -o
. ~ c o
- g u .1 x n J.
-y y pp,a _
. ,-r c ----
-~ W _ A / y y yr g
-r--u r1 s x 2 v Q F A. r x<rxL n ---- -- -,
, . r-s , Ff/fd'fMX ff x -2 m m :=
, .. , . u -
.2 _1111 w s w -
. . _ _ _. ---r- -.
. ' /[ . 1 a ll ffffs !!r
. //
1_/ / / ff/M fiffs r s s s wr T s. o aw i REVISION 11 - vf/ / f ~ f / f; /I g > .. - t 'Y 1" / /
- > m eee /v r o e
/ e ** T J,._ _f ffff /1/ / / / f /
{
/ f /////// f_ f f f / / / / / / / s / ,\ / o fc ~
1-. 1 _ 71/ ssr ,1,77 t;
- -ss - -1,7.1
, l
-_ - - -s1- 1: 1 r . - - - I - - ,
-= r -
bl g - l /,Frss/r17 sss s ss s 1 s7 r 9s 1rs .1 r s 7 r. :s e r r 1 v. 7. 71 1
- e -
- t . ; 7 /-//////sss: 7 7 7 1 s s 1 1 1 .r , 1 s =
-i Ct ! l, ? : 4 o,i f r.s,y ,, / /,r,/,s,1,s,I,T,/ ,1; ,- ,_ 1 x 1 s . 1 .
,r ,r ,J ,J. ~ ,
fj i' ::= 2 %. Kff//// / /_
/~ f'f. 1 J-
- w:. - _. .: -:- : . - -+-.. .--.:=.m. /- / I
. : . . / A . .- *- $
t : ;
.e
,__ ,<,><- , = .
. %.u ,:- - . 2- - - -
5
; --:1 _r- 1.. . -
cl
,':--_ -1r' /-+/-/--M, J'. ! T. : ,'.
], s'i,' s : -c
- 1. i --
t - . o
= . - c , .1
- ~ -
. -=
- 3 -3 e 1: 1 r r c1 1 1 . _: 1 : 1: :1. .s 1 1 g 2 s _-
i *2 n
- 11 /I f 1 :s 1 .s s. s: r.s_ 1 . 1 - 1: '
1 . a " ' =
; ; t :1 s. 1 J
r_F fM / QJ -r 1:1 : rc :~. r 1 J l I E
; z r *
.n
? a .1 :1 s r 1 s s 7 . r. .s.
1
" - ? i "2 ~s G.
- - i :f - - .P-k ,&f-f ~f= ;4;Lf. . . - .*.];
^
l
- l
. g ,-
-1
_ ,. ._ r - i ~2
- .- n
} 4
,. ,-f+/~-/
,. ,6' ; ,F -
2 3 - I _I
,J
# : I : # F. _ E .
- :i. _i -1: 1.
I N 7- 1. 1 1: . 1 4 g
" .: 3 : .s 1 1 - i .. 1 - 1 r- r 7
1_ I
, :b -4 T- 53 ef .7 Ii
.1 I. TI s -r h:
, 1 I
. 1i I' 'E
- 1. : r:
E '7 _1 i I I 1 I a c
-e
% 2 i f f_ f F ! / / f .f t . 21 1
1 w l 5 : ~ 2
- fT / 'l / ! f. I'A1 l'TE R n i g '*
i f iE
- {* {' ;b! /!
I
/: /
T : / . A 7! / f T/
/i !i/
if t !I I < 3
~5
; i W' '
me io- >- g!
- i i
- *im / t i / 1 l I f '
1~Ai '
!I e
- *
- - = -r ,1 .
,1 :_ 1i ,1 : . ,1 : * '- il o
- D
** 1 : ,.7 .
.I i I- J i I E
- 1. A. I
-s
~ . v .e.
i I t _ :, / 1-
'I 1 - 1
/ .a 1
e 1 i J
- r. :
11i 1.- i o i
.O '$
' 2 *- : g r
'l r
~I J 'I I 1. t I .
s-i ga!
!; o A
;g u -
u I ' I - I i ,, 2
,7 . ,I
-.f -
,I i,f I-ir _,r I'i:
c
;; 2 ;
O
'I U1
...-f : - -
=-
- 0. , o g; i }i, i : ;: , : . :
u , - 1 :, ,": : En D *
. , . , e i IT ,2 d /* : ,7 .'
- o o E- o .m
- _ :: ,1'^ 1 i ,.': ,
fs ss ,: . . . , au L , s' 1 : 1' E ! K.
,., , n j
- - . _ i
+ i ~t 3 i E K F I J E . I V: I 1 . o 4O 9 i ~ :A g 34
~2 r2 1 i
1: r 1 1 r r i 'i_ 1 .
' *em .
- T- kC -
r 1 e . -i.5
=
n h a; oc os
*o e
o c:
- .( =i / -03 -
1
=
oc w- >e
!v,f 5 : ( 5I !) i ,N ,'
3M Ec [ * * .! -z 1c 3 - - . . - . - 5* 4 1-+.Ij - 1
,1 - 1 's 1. , ~ ,f J 4 m *I -
O3 5 1 E _ 1.
! E E _
i .. QO* o e c. 1 t ;% i! 7: 11
.1
's -
r I . ..
- o*
' ; i . .i: ; Eo wC
$r tj 3G{N '
- c. /
r . [ Ii . r [ i/ NW: . l s E 2' i is! 4 ~s'I ! I Ii il t , i -36 O ;I 3*j ! Ii i Io fG i i 1 i , fI ux i I g,}v ij 2 4 . ' ! 1 ! I i J t i ! i I 1 ! 3y lii.J ilaim i .ua _ j , z-s< z<<
-- i i illui ihj ljj pg; g; j y;g g jg ;; g o i a _
$ I a *
- 3 $ _ o 8 e e e e A ec ,e .. o o o o o o o ev . o o e o o ooaooo o o o o o o sloegep ssel Jo 3 40 e3usJansoo go Augqcqo2e ao geno, govepyucQ,seewnsuo;) .I i 1 l
l i 1 A L L #f9 *C&4 AL f444 r994 .
~~
THORNDIKE CHART
~ ~ ~~ ~ [ ,
4 e 4 NC 4.12
-- . _ , - - e._, . , . , . , . , , , ..- -. -- . _ ~ - - , , . , _ - . m..- . , , . - , ,.sm-. . ., .c .
ALLse.cHALMEne Powan avm!!ntaMe, oruc. O E R- 504 10/75 / The values for operating experience and failures given in the overspeed ) ; i failure rate chart are cunnulative f rom the starting dates shown l, through January 1 1975. The f ailure rates of the elements also may be considered to include all interconnections from element to element so that the calculated failure probability covers the entire overspeed prevention system. i The chart "Overspeed Failure Rates of the Stop and Control Valves and i l g Turbine Trip and Control System E lenents" snows the result of inves tiga-tion. The chart lists the operating experience, and the number of I failures which could lead to a >l20% speed event for each element I i of the system. The app'icable expertence i nforma t ion and the app i i cab le l'-
! i failures are included in the data chart. The mean t ime between failure (MTBF) and the failure rate 4 are calculated on a 95% confidence level i using the Thorndike Chart. Wi th this conservative 952 confidence level, i
i the MT3F and A for an element which had ze o historical failure are , 4 calculated with an assumed number of 2.95 failures. For one historical failure, 4.7 failures are assumed, and for an element with two historical i I failure, the MTBF and A have been calculated w*th an assumed number of < 6.25 failures. i i The mission times "t" are based on a bi weekly test of the trip system and the stop and cratrol valves, and a yearly inspection (e.g. during , reactor refueling) of the entire overspeed prevention system. Experience . indicates it can be safely assumed that no significant oear-out of our ' system elements will occur between yearly inspections so that purely random failure rates are valid. ( /
- - . . . - - . . _ . . - . - . . - . . . _ . . - - - . . .. _ . _ ~ . . . . - -. ~ .
ALLIB=CHALMEMS PC>WEM SYSTEMS. INC. O ER-504 10/75 , OVERSPEED FAILURE RATES OF THE STOP AND CONTROL VALVES AND THE TURBINE TRIP AND CONTROL SYSTEM ELEMENTS!II III No. of Cornpo. MTBF witt 3' ' A. Failure " ate 3' i Ele.ments, Component Name , Experience ,Compo , nent Failures' 2F 95% l t, Missionl4' < *im Sn
! Nur,iber Since : nents Hrs. Confidence ' Time Confidence I 4
7 Al 4 & 21 & 22 Stop Valves 1953 i $34 # N3x10 2 8.48x106 HR$ 336 dRS 0.ll7/106 HRS l 91 4 _ l j - . _ _ _ . . _ _ _ . _ . Cl.4 & 19 & 20 . Controt Valves ! 1965 506 1.85x107 1 3.93x106 HRS ; 336 HRS ' O.254';g6 sp3 l !' 01-4 l
~
, E14 & 5 & 6 & 13; Folic
- Up
! F14 & & 14 & 15 Pistons 1954 1360 7.2x107 0 24.4x10;HR$ - 8760 HRS 0.041 Ic6 HRS H'.H2
, Gl.G2 11 & 12 Electro. Hydr. >
1963 86 + 2.65x106 ' 0 0.90x106 HRS i B760 HRS 1.11/106 HRS Converters i !
. mtncut Coils 1 t
7 l I 2 Methonical 1954 427 1.68x10 1 , 3.57x106 HRS l 8760 HRS 0.23/106 H R$ l
.i Hyd. Control i
- _. 4 Kl.K2 7&8 Admission 1963 36 2.65x106 - 4 0.29x106HRSI B760 HRS ' 3.41'106 HRS i !
Controls , wim Coils ! L&M 4 EHC Speed 1963 36 2.65x106 0 0.90x106 HRS 8760 HRS 1.11 106 HRS
&N ,
Control & i
' Sceed Measur-ty Device
, 0&P 3 Power Supply 1963 86 ! 2.65x106 0 0.90x106HR$! 3760 HRS i 1.11/106 HR$
i l R Change.0ver , 1966 j 53 il.45x106 } 0 336 HRS ' 10 106g ggisi
, Device 0 0001123 106 HRS
- - ~
! SI,S2 , Vain Trip 1958 273 1.43x107 0 4.85x106 HRS 336 HRS 0 206/106 HRS i
Valves i T [byerspeed Trip j 1958 242 l.4x507 0 , t 336 HRS : 10106HRSi6i
; Test and '
1 l 3 l I and ' O.0335 ~106 HRS Resetting 1 l l ! l B760 HRS 0.000225 'l> HRS Device I ! !
! Ul,U2 i Overspead 1953 436 - 2.?Sv107 7 2.12x106 HRS i 336 HRS 0.471/106 HRS
! l Trip Bc .1 ,
i ! VI.V2 i 1958 486 2.7Bx107 0 9.42x106HRSI- 336 HRS 0.106/106 HRS
! ino Release {
Devices , t ( l'), ( 2) For predict:ng random failures of the elements which could lead to a> 20% overspeedevent nistorical fa$ ares in" control and 'or stop valve oDening direct On" are listed. Thts includes " control or sten valve stay open failures." too. (3) MTeF iMean Time Between Failures) and 1(Failure Rates) are calculated on a 95% confidence level usmg the " Cumulative Curves for Poisson Distribution lThorndike Chart)"
- 14) Mission tee is defmed by the bi weenly automat:c test of the trip system and stop and control valves and a yearly inspection of the overspeed control system.
. 5 ), l6 ) Tne failure rates of these e ssments wnien are only required durin: te . ting are defined as described in tnis report.
(7) rtistorical data througn January 1,1915. J
At.ue-CHALMGCMe POWGut e>wTEWo. MYC. ER-504 10/75 , -- The mission time "t" and f ailure rate "\" are used to calculate P the probability of failure with the following formula: P= le~ This equation can be expanded into an infinite serie4 as follows: s P= \t - + -+- f ...... 1 For At <0.3 the following (imation including the fourth power term i s used in the compute r p r:gr o.7. to calculate l , P=At (\t)2 + (At)1
- 21 ~ ~ ~ lit)"
I I 31 4' l l i
\
The failure probabili ty of all components : is assumed constant over the ! whole i li fe of the uni t based upon periodic testing and maintenance which continually checks and repairs or replaces components to maintain i the reliabili ty of the system. i .. 'ther words we are assuming there will I { be no significant wear-out ef fects. In this connection, it should be ! noted that in our study of operating experience and past falltees gcing back to 1958, there is no indication of any componcnt wear-out trends I or ef fects. i j f I t
)
e ALLas-oMALMarm man eversnes, swo, ^ FR-504 10/75 f The historical f ailure rate of the stop valves for example is based on the periodic testing of these valves which was actually performed in the pas t , and is an average A for the actual tes t interval. Mos t of the valves in the si.mple were not equipped with an automatic turbine tester (ATT), therefore, valve testing was not performed bi weekly (monthly test intervals can be assumed as average). Considering the importance of testing, and having the benefit of the ATT, valve testing of nuclear units is recommended bi weekly. As the following diagram indicates, we-took the onservative approach..and did not change tne hi s tori ca l failure rate due to more Frequent testing, but rather used the bi weekly test time as the mission time t of the stop valves. FA!wer ' 4A re- ; i
% 1; ,
A j , ' , I
%TA__ __ _ _ _ _ _ _ _
_________l________./_______.___f____ I e ; g , V,/ I I
; / *
'- .l I
Titilf tg336hr d l 1 t(67Ahr ---* 4 AVERAGE FAILURE RATES OF REDUNDANT SYSTEMS
- gemmmuna j The historical A = A g , based on a four week 'tes t interval. .With bi weekly test-interval, A-and t will decrease to A = A g and t= t2.
However, we actually used the larger A = A for'the bi weekly test t2. f i J r 1 ,_ _
. - - . - --- - ' - - ~ ~ ~ - " "~~
AL.ue-cHAL.Marne powen overnmen. ona O E R- 504 10/75 ( ~ Follo< lng are brlef discussions of the experience, failure and other information used to establish the failure rate of each sys tem elemen t. 4.3.1 HP and IP Stop and Lontrol Val'.es, E lements A l-4 to D l-4 The applicable historical data is as follows: Stop Valves Operation experience since 1958 Valves in operation 884 Valve hours 5.3 x 107 hrs Failures 2 l Control Valves Operation experience since 1965 i Valves in operation 506 Valve nours 1.85 x 107 nrs i Failures 1 i 1 The stop and contro! valve designs have been changed and improved during l l l the past fi f teen years ; therefore, arriving at a realistic historical 3 i failure rate applicable to our present designs requires a correct j i
+
i selection of experience. To select a representative sample of valves wi th previous, but similar design, w .nade the following assumptions: ; I
) a. For HP stop and control valves, the chaage from stem packings I
to sealing bushings is the starting point of our historical data. All HP stop and control valves with sealing bushings I built by Siemens and KWU are included in the data. l t j
an ues.cmsn.Marne raowan evnernMe, uvo.S ER-504 10/75 (
- b. For IP stop and control valves of fossil reheat turbines with the !
relative low steam pressure, the sealing design is not as important i a design criteria. Therefore, another criteria had to be chosen. As shown in Drawing NC 4.08, basically two diffe ent IP stop and control valves have been used in the past. The valve with a combined seat area for one of the two controi talve seats and fo r the si:op valve seat was the older design "B" and was replaced by l design "A" wi th separate seat areas for the stop and control valves. We started our datt with the introduction of this newer IP stop and control valve design "A" in 1965. All IP stcp and control valves 1
, for fossil reheat turbines of design "A" and the present design k
built by Siemens and KWU are in the data. 1 l l All presen t HP and IP stop and control valves for fossil and nuclear applications a~e of the same basic design and fab.-icated under the same quality control program. IP valves for fossil applications and HP valves 4 for LWR applications operate under similar condi tions; whereas, the HP I valves of foss" turbint.s operate under more severe condi tions , therefore, it i is a conservative approach to include all valves of present design in I our statistical data to calculate the valve failure rates. The Drawings TC 2.10b and TC 2. lib show our present HP and IP valve d 5 gns j for fossi! appl' cations. ! The Drawing NC 2.05c shows our presen t HP s top and contrci valve design for LWR applications which is described in the speed j con t roi report No. E415 i Some of the improvements of the present design are: 1 (
)
i 4 g -...~... y l *
-w A .s,
. * ,313' - f ,. 1 _ REVISION
/ J ,6 7d ?L ddal_ .j c--.I tj f0i ! erY % 4 J
- L, 3i
[b: l J L, C! ! l: n f, ( !, dt ,
>E ,
n I _
!,-~~ ,==w
- t .' d el- j
+ ;~ r
- ygame 4
\. y-2 '
4 j'
;l i}e
; ii _s -
ll ml L _.<p N.'=~- n: M W. ,em. -;$a. .,.x t
. - a -
w _.x .'
; tgy
- 1 _
ei s, LyMY[' i . f'3.s.___ ?n%s Y i bd . '
%EVIOUS HP STOP AND CONTROL VALVE DESIGN WITH THE STEAM STRAINER IN THE STOP VALVE CA$ LNG y --
i -_ ~ h' e
% FLm . 71
'6!=.;=ak '
JHh
~
i GD'WD a pn # ~ .., Utiir ,
-t w L
njg : --g
; .LY:- " "'
i- sjjj; vwi - g DT y* "' [' - '
- p. . ,. , .
p ,,. .
.9-
{ ;J -- s Q
.._ 'wp ml
,t--
M' i 4, b $ v n=*
*N fnl (E ; [4 D. jigy "4 -
,f( ' 7 7(i 2
2 :; ' [ Mda,Q3 { ?,
, riI }' +4 ! y _I.Fp g[
~,' // ; i ~
; 'f 4 (is Lgg i? E- _- - da --
GJ , v- y { j{i; \ g %sn LGu7 !L4( a ' O^ ~.j! PREVIOUS IP ST0p AND CONTROL PRS.VIOUS IP STOP AND CONTROL VALVE VALVE DESIGN P WITH COMBINED DESIGN A WITH DOUBLE SEAT CONTROL VALVE SEAT AREA FOR THE STOP AND CONTROL VALVE
/s-PREVIOUS STOP AND CONTROL VALVE DESIGN Z.72^/.Z". m I NC 4.08 4
I i
;7-]'-5 i
l l ! i I ! REVl$ ION !
/1 i
.t i
J k, /p . 4 i _ _ . h p y!,eje; I w, m
~
Ci ': , h di - e1 i l [ \S STOP VALV E { -
.,s .
's.
i f
- c._ , a ' '
) - p._y- _ L___L _ FROM ! f ;, ( BOILER C
==
,[
CONTROL VALVE WITil {' ;) 450 PSIG HYDR AULIC ACTU ATOR Ii _t TO HP ' TURBINE 33 Ii{
"7' ,
7-
!( il ' --9 l
fi CIl ;) 1F* * *, 6,' t =- s "
}j s, {p[]e, {h,d " -
) l 3H b.= n, i
1$ 3 A i ' i% siO
- k
- 41
- it 31; l l HP STOP AND ^u'=<'*'-a= l CONTROL VALVES FOR FOSSIL AND HTGR APPLICATIONS "'"""""""""'" [. l-TC 2.10b
Et I REVISION if] { . _p3 -._ i 0 !6 r Ft 7
. ._-.E A - . .
+
J ,l
- s 5 1,. ,
$' _m. mm-
'g l l a < -
di . t- ' I _ _0 l I H.j 4 l I? , . ' m. STOP V ALVE s
..- Y k
l .I ~ i t FROM - *- REHEATER - I
==__
~>
,, q CGNTROL VALVE WITH e N1. <
4F0 PSIG HYDR AULIC ACTUA'OR
\
g I> 4 l*~~._... ____ Nk I'h -- s4 TOIP , j }~: TURBI IE M X. .$ l e='
.3.
n - ;;. +
, D .E=-.
.u
[- _, . P '- 1l l 13 12 :1 5iid 53 - = 21 151 Z h = :1
;i : e=i a< Z$4
/ 4
!? STOP AND
^ " ' " ' "*'**"" " '-
man averawe ,~c 4 CONTROL VALVES FOR , FOSSIL AND HTGR APPLICATIONS l TC 2,11b
VALVE DRIVE
" VAPOR
~
SUCTION
- 14 PSI A
.. tt :: Cn_
00 .i - ' hh CLEAN SEAL STEAM 8 W ~ ~ SUPPLY 0.7 PSIG
- t. - ~
l R[VI$lON . - - . _ _ - ,9 f . i /i j E - O ' yi : I jl SEAL STE AM gj ; [ SUCTION = 7 PSI A tb1 ! ,,
., . , 'MJ~ ~~~ ]j [ij .
di i - # L ---- 0l ! y %= -y STE AM COMPARTMENT
, ii it in 1
CLEAN STEAM SEAL ON THE STOP j l L .,3 g AND THE CONTROL V ALVE STEMS FOR BWR APPLICATIONS Y l [ 11 g! a_ Z_"__~ Lh' " ' FROM REACTOR _
-i
,[ k*
-- I STOP VALVE
}E N ij k, 3 -
. q*
j i ' %,. s is 2 .-.
- $ ?i5 ,
h -Dl__ i TM 5f
- e 5;7 : : c TO HP l
- -4 l
I l- -- _ - - +Nz ET;J2: - - - . TURBINE !,', - t
~'
rify$;',
- t 1 M i! *.'n 7 J R __Ja*=f 21 i'%
na.ij 1- +[ CONTROL VALVE WITH
!I iil 450 PSIG HYORAULIC ACTUATOR H
TO REHEATER HP STOP AND CONTROL ^ " '" c * * " " " mw.'. ".v o rm ,~c i VALVE FOR BWR AND PWR APPLICATIONS NC 2.0Sc
'd i-i Au.e-oMALMurme powsm eveTOMe, avc ^
; ER-504 10/75 (
e t ! - Seperate steam strainers j !' i - Stop and all control valve cones guided (except HP stop valves d for 31800 psig main steam) ! - HP and IP stop valves with cones covered in open position i 1 Single seat control valves { Control .va'ses with two-step ampli fiers with 450 psig fluid i l pressure cetuator Two (2) overspeed-type failures are known for the above specified sample of-t t j j HP and IP stop valves, and one (1) failure for the sample of HP and IP j i j control valves. All three failures happ(ned on the same fossil fueled i turbine-generator unit in 1967. The first stop valve failure happened j l because riveted support rods in the steam strainer broke, and parts of l l these rods prevented closing of the stop valve (see strainer Drawing NC 4.09 l and previous HP stop and control valve Drawing NC 4.08, page 49). The steam i I strainer was replaced, however, the same type of failure happened again six weeks later. This time steam strainer parts also got into the control i a valve and blocked closure of the control valve too. These two stop valve-i t failures.and one control valve failure were due to failures of the steam ! I 4 i st*ainer and not failures of the valves themselves. The steam strainer } - - l was designed with riveted support rods and local stress concentrations led l I j to.a fracture of the rod ends. Since then the strainer design has been ! I completely changed, and in the current design heavy support l 1 i ! bars are used to form a rigid welded frame, instead of the riveted rods, i { Furthermore, the steam strainers arc no longer located within the stop valve
- casings, but rather in their own separate strainer casings (see Drawing NC h,09)
I-4 installed upstream of each stop valve in the main and reheat steam pipes. j- k 1 i , __ , ._ . - -
g . l 1
- [ ',
1 y ,
'A
.ac= -=- ==n l
Q
\
REVISC
,/now
- - - - = rk-
)
+ -
p 0! i \ bI ' # i f i c ! ' - g ] DETAll A ! e i' __ -_ i I PRtVIOUS STE AM STRAINER DESIGN LOCATED IN THE STOP VALVE CAS
'ld. . _ ._A , ___ % '
1
, __ j ._._ _ _ .. (
1 3 [ ,3 i ' e a#t 4
""*"j n - -i
/c-; ~; a) 's i
j-' [;i.- 3 '3- i I -
- !i E, . ki 2 3 l j {i ei iii /
- l2
~:
) :' "
t 06,.' $! ,
- $Mi!I : :
P 45 :
~
~
j ;g: 1
.s e g_. ; y 19 r ; ;
e 72 t 's : yr-qj n i : 3-) 1
' i
~ "
55 s. ' : :
- u. a v E"e.
t y E$$ , 5l** { :: U
$$ $E i E
-= \
,.):N a[
jf ;*,071 js, STRAINER BASKET WITH ! l 'l ,,i rg 3 52 1 WELDEDIN MEAVY :l ; sd z_ sg.
;)
SUPPORT PlLLARS ! 'i ae ; ' rt Si
!! .t El ev 1
. L ga_
6 Ai d 83 n. -n
!! } ! e-
-~
M: . M SEPA* nTE STEAM STRAINER LOCATED IN THE MAlh AND REHEAT UP STREAM 0F THE STOP VALVES aam c a nan. STEAM STRAINER DESIGN [M[\ NC 4.09 - e ar ._- ,. +,n-, w- , , , . , , - . , , - . , , - , , - , , - - - - . - - - - ,__---,,,.-----,,-n--.,,,,-.vmw....n-e-, ,, ,-,,w-,,,w
At L se-cHAL.Marna Pownwr everCMe.nc.6 ER-504 10/75 ,
< T These three f ailures are the only ones known for the selected sample of HP and 1:' s top and cont rol valves which could lead to a >l207 l speed incident. There are no reported failures due to bent or l
sticking valve stems preventing valve closure. l I i. Based on the foregoing, the failure rates and times for the stop and control valves are as follws: S tco Valves i Stop valve hours 5.3 x 107 hrs t Historical failures 2
! Assume, failures with 95% confidence 6.25 t, mission time 336 hrs (bi-weekly test) i I
A, failure rate 0.117/106 hrs I ! Con trol Valves i Control valve hours 1.85 x 107 hrs ) Historical failure 1 , , Assumed failures wi th 95% confidence 4.70 ! t, mission time 336 hrs (bi weekly test) ; I A, failure rate 0.254/?O6 hrs l l 1 l I Also included in the historical data of turbine valves are the butter ' fly-type '.P valves of nuclear units for LWR application. i The butterfly-type LP stop valves and LP control valves are built with the l t l same design criteria, as well as manufacturing and quality control procedures, as the HP stop and control valves. The Drawing E5.127a illustrates the stesm compartment and mechanical drive i arrangement which is exactly the same for the LP stop valves and LP control N J
- t jil j l{ ((
t I ty _. S
~
A R H ~D S c C.
=
R .* t u . Ptt - A P
=.*
A ~R L .D c Y :
=
_. f H
,jpl ,1 ,Il, J 4"*
D j{1h < 3 -iy . Nv Ai E _ _ .~~ H _< w" f GD C _ [' N . _ O t ILH _. ND _ 5.- 4 . [M t PC UE _ GN t SA ( . f
.L e3 C
OM _
'((
^
( S D E S EVf V [,lL q
.l.}l
-ifl VL L L A A p -
AV V VP MR r.li' EOt PT0 l
- m. .
Af E5 T 4
. YSf 4 7)
T K . SE . TP D N Y( L HO Ile t AN b -- o1 y $C O _ F I C M f S _ ROP j . a* _ 4 FBt
. g }
iN I H I} il% - s] s hw'd. l1, ,
. aLA g AE M S Yl O Sl I
U l 5 0
- 1" -
.M l4
- i1T c$S LA _. f I'
T ATS _ p AC .I o E L . E' - ml1f-P ' tt U - SP
. i A H
. MPA _.
~ .
t Ak S . A _
, + , E E NE . fH S
GV . ,ll T W_ N\' gyf~." u
~ I L S SA _ W B -
MH N H _
, .{ EV - A
./,h,' D L - aEC g AT s
[ O_ F EO - g T t g$S L .. - I . . - V L R C t ' i T - _ _ at ' AN _ _ l II I VO A EC A E D 3
.u -
P - _ y Y P - L i, 1 l t i , e g f' T L D - ._ (;,k . Y N - _ _ L F RS A TT1 _ 7 1 Ix. N. . E E TV TL MM
~ _
. . g' 'Q'- UA - A BV -
a~
~
' .I -
g - - LP - u s -
~- hj . AO .
, ( _ < A CT I S T .
~ .
\' N P -
[,
)
d - g E Df1 i - n-g I O F (I :l!n!j,t 1
!/ -
- - T -
t*
-y -
- ig
- . 9
- ,7 .
pT l g S g y U gA g H gX gE g M A f 1 _ T M nSU
, a S oW E
'i g
iH gX A/ n ,
. % N<
G t R= mL sA u E Aee oT E HN E n- -r L SR C *
- Rr 6 E '
A
. ,a w ;g t m - u568*
- E SS P
} 5 Pa 7 N AC T fO i
F A A Z agf M -: . AT HA I
) N L SC j
, 4 i' y I aT9(
yJrM l g ( O A DL I
-- I T T E
NP APA
- C E
S D - t I G
. : F m IN W R r O
/ ,L@n ii i RP
_. =//[ . .
/
/
N OM Tt C" A E _=
-- h.
y I RM 4 D 9 ,m~ ,1 B
#I y
;qh g ?
5 ~ - _ j.$:- _ g M = ,* I
! ;y vd )hsjp
_ g E tJ
- ~ ~ ~
I! ge,jj:Nh if
,/ 0 'C -
s
ALLIS=CHALMERS POWER SYSTSMS INC. C ER-504 10/75 T i valves. The butterfly flaps have a ci rcumferential clearance of about 1/20 l inch to avoid valve sticking. The maximum possible steam leakage through the value flap clearance is taken into account for safe turbine generator fuil-load rejections to Ouxillary load and no-load condition without tripping the 1 turbine. The butterfly-type valves have a symmetrical design with a syrret rical beari ng arrangement to eliminate axial thrust forces. The axial clearances and thmst bearing arrangement allow operation under all transient conditions without increased axial friction or excessive axial forces. Spherical journal bearings guarantee a minimum of rotational f rictioh. Each l flap journal bearing is designed with two separated bearing sliding surfaces, both equipped with low-f riction material; this allows proper operation even in case of sticking of one sliding surface in each bearing. The valve spring and steam forces are acting in closing direction. The valve flap bearings are located eccentrically to the valve housing, resulting in propor steam forces in the closing direction. During load operation of the turbine- , 1 generator, the LP stop and control vch !s are fully open. In the open-valve position, the hydraulic drive pulls the valve flap against ! an open ing posi t ion support , thus avoiding valve flap vibrations. The l mechanical valve drive like the valve flap itsel f is supported by two separate bearings. A full-flex gear-type coupling is provided to connect the flap shaft with the mechanical drive. i l l The hydraulic actuators are shown in Drawing HC 4.21 for the LP stop valves and l in NC 4.22 for the LP control valves. The hydraulic drive of the LP control valve is of the same design as the HP and IP control valve actuators with two- j step amplifiers. On the LP stop valves, the two-step amplifiers are replaced , L j k
d s pagt '. M
*-l :
L m iJ __ I. I
~' '-~
- REVISION i !l j f ] s ,; o; !
i ! OI i !, bi. ! i Cl lj i dI - . j ei ; i . _ . ..
- [
_ . - -4 .
. .4 i
s l .
' ._::/ . _
f fi
- 4,("y .:;:,
LLi p l 3
+ i - N
! M ? ',.
.' g -
L4= t '], , f.a;l;!L T. 7 i .. I t pH. I !q. 2; {d4 -l.ib. -.:
- l. -
n
- 2. x.
<g - , t .-oj
- m s 3
w :-
-y.
i II. -,
/.ft u.-. . ,
L, ! , ,. Al* L.- - 1 i I 4 la f 4
't l --~ _ . ,
- a. ; a ,
a :
)
I ! t i
- )! ,
V=- $ h# . _, l e i
- Ti;
' il 'I
>d
- ;c 1. i <
~, __
to - ! 53 i y ,n' . T lm 34 ;.: ,- , m. )
/
1c.lt 7*
;.i i
/ v, i
-.9;& 1 .- % . 2
\_' </[x \ / ,.
1 5-se t- ' y } I
~
y'
; [11ij tr- Adf - '; ;
il2-3: 3-: 33'e2 t. s.
~e:s. ~~ ~
t s< zsg i
- - . . g
- -----. -.- _. _b <
' - ~ . _ _ _ .. . . _ _ _ _. . _ ,
4 ACTUATOR OF A ^ '= c ** *' ~ e a= ./- i 1 BUTTERFLY TYPE LP STOP VALVE
~ ""*""***"'"
[ NC 4.21
eg-(;i;;4
~
i L JO '
<=._'
6- x o- . y g.
- q. , , ,
r
! PEW 510N I l
p, p . g . , , l
!al i i I
bi
~
CI i di - t e; l . _. [ - , le ,
.:~
, _ . . . ,/ s .. T p .. . . - J . . .
j; _ w . 5= -
.,i r
,s
?.+
+ -
s, ,, :
' }
WilCil) ,
.ncr .
~t 4 ,Lt. _Q i i " j
. : .j Ta 0' , sa { c .-
[ [M's . ; c4
\ \
+
\\
- =
U . s ! i 3E l
~
l! i 4 ~_ t l i
~ 35; *
.3 ,33
- s
$ 3 o 31 l'1 Um ;1 ll 33 .ij j{;s;!
_z 2,
";[
i
, hl . .-. .
is I F , s$*[ I 3
.: /$ -
~
< 1 _s i i, f 5 .s-3it ,r it iin g.wm ACTUATOR OF A a u = -c ~^' ~
- a= -
BUTTERFLY TYPE LP CONTROL VALVC
~"""""""""*.[..
NC, 4.22
ALLIS=CHALMERS POWER SYSTEMS, tt%C. O ER-504 10/75 (' D by a fast response control piston. This feature provides the LP stop valves wi th a highly reliable two posi tion (fully open - fully closed) function. I 1 Similar to the HP stop and control valves, the butterfly-type LP stop and control valves are provided with automatic test features for the bi weekly valve test. 4.3.2 Follow-Up Pistons, Elements E l-6, F l-6 and H I &H 2 The applicable historical data is: Operation experience since 1954 Pistons in operation 1360 Piston hours 7.2 x 106 hrs i Failures 0 l The follow-up pistons are simple adjustable drain pistons. During I the past 20 years there has been no failure of such a follow-up piston l in closing direction which would result in a faulty opening signal to the control valves. The f ail safe direction of the follow-up pistons is ; the drain opening or control valve closing direction. The failure rate and mission time for these pistons are as follows: l l Follow-up piston hours 7.2 x 106 hrs Historical failures O Assumed failures with 95% confidence 2.95 t, mission time 8760 hr (yearly inspection) A, failure rate 0.041/106 hrs
ALLI6-CHALMERS POWEM SYSTEMS. WC.5+ ER-504 10/7; -61 [ 4.3.) Elect ro-Hydraulic Cor,verters wi thout Coils , Elements G l f, G 2 The applicable historical data is: Operation experience since 1963 EH converters in operation 86 EH convertir hours 2.65 x 106 h rs failures 0 ~
$1nce introduction of the EHC in 1963, no failures in control valve opening direction are reported for the mechanical-hydraulic pbrt of the CH converter. The electrical part (coll) is included in the steam g admission control. Th:, converter is not a n w design because the hydraulic part was used as a hydraulic arr.gli fier fyr turbines wi th only Mechanical-Hydraulir. Control (MHC), and wi us the present EHC system.
the EH converter receives both the electrical input signal from the ; adraission control and the hydraulic signal f rom the MHC. The failure rate and mission time for the EH converter without coil are: EH converter hours 2.65 x 106 hrs Historical failures O Assumed failures with 95% confidence 2.95 t, mission time 8760 he (yearly inspection) A, failure rate 1.11/106 hrs l t n' - _ _ . - - . _ - _ - - - - - _
ALLtO*CHALMERS POWER SYt9TEMS. INC&* E R 8i04 10/7g -
! i i
k.3.4 Mechanical-H,vdraulic Control (MHC), Element I The applicable historical data is: Operation experle e since 1954 MHC's in operat ion 427 MHC hours 1.68 x 107 hrs Failures I One failure of the above MHC's in the control valve opening direction has been reported. The hydraulic speed transe,itter did not receive a sufficient oil supply from the turbine main shaft oil pump (see Drawing E5.121). The low oil flow supply resulte in a too low dischar ge pressure of the impeller, if, during this condition. a real overspeed event would heve occurred, the increase in oil pressure f rom the speed transmitter would not have been high enough to prevent n overspeed. The failura rate and mission time are: i MHC hours 1.60 x 107 hrs Hl.torical fallares l l Assumed failures with 95% confidence 4.7 { t, mission time 8760 hr (yearly inspection) i
! A, f ailure rate 0.28/106 hrs i l i
, i l l g
4 3 TRIP BOLT HP ROTOR / ;< , i ! [ Ep1Sl0N f Il lI ji , .t" lI Il l N h l k ll C1 i ll ll d! i il ll '
-l 1! l l'
i
\
II ) O CC ' K 2 l l I TEST Olt ? I CH AM BE R ,,, .x
/ /
'J f 1 PRECISION SPRINS
% \\\Y .
NW ELECTRIC SPE ED TR ANSMITTE R "e ITWO CHANNELS) c=7 '
- ; HYOR AULIC TEST / 2_' ' =c i -e_
^
- OF OVER SPEE0
/ :z- O:~Ja CC d' "
TRIP BOLTS / R OVER SPEED ' _f'. M
~
TRIP DEVICE 1 RESET *
.?,2~ <<s. , .. i.. -..
ji ' !' t HYORAULIC e f f f j f f f f ,T,U,RBINE GENER ATOR SHAFT , SPEED c dhlm,# u* 1! g [ . T R ANSMITTE R wh, kc-, tw.w-m_-a; um
=_~ATLF'-
1?*.i ' ov5a 5'5Eo
==-' W Tr,> ' ' -
}I 1* j ?: TRIP DEVICE 2 4 i!.~~ '
MA8N Olt RESET g}: j j f PUMP
.m _
4 ALL am CH AL Ma mes f* OVERSPErg m .....,, .,w y TRIF BOI.TS T A fLCtjlfdl
ALLt8eCHAL.MGR8 POWER BYSTHMS. UYC. ? ER-504 10/75 , t 4.3 5 Admission Controls (ath Coils, Elements K I & K2 The applicable historical data is: Operation experience since 1963 Admission contro?? In Oeration 86 Admission control hours 2.65 x '0E hrs Fai l ures k Three of the four f allv; 1s were open-ci rcui t failures of the plunger coli in the EH converter, which happened during turning gear operating caused by extremely high amplitude oscillations of the electrical EHC signal. This problem was eliminated for future units by filtering the electrical low speed signal. With the nevi detign of this component, failures of three spring type connections (which normally provide one-out-of three reliability) to the plunger coil rust fall to produce this event. l The fourth admission control failure was a broken electrical lead in the ; I l the EHC cabinet. All f our failures occurred in control valve opening j direction and could have led to an overspeed event. The failure rate I and mission time are-l l l ! Admission control hours l 2.65 x 106 hrs i l 1 l Historical failures 4 l l i i Assumed failures with 952 confidence 9.1 I i t , mi ss ion t ime 8760 hr (yearly inspection) i l A, failure rate 3.43/106 hrs ! l
\ .. >
ALLie cHALMens powcn sysrcMs.sNo.O ER-504 13/75 ( ' 4.3.6 EHC speed Cont rol and Speed Measuring Device, Elements L & M L N
! The applicable historical data is:
i I Operation experience since 1963
! Speed controls in operation 36 f ; 5 med control hour *. 2.65 x 106 nrs Failures 0
. These elerents consist of the EHC speed control with digi tal/ analog converter, the speed neasuring device including the disc with 120 pernanent magnets nounted on the HP turbine rotor and the two channels of the Enc speed transmitter, pulse converter and time supervisory s ub s y s '.em. There have been no reported failures in the control valve opening direction for all of these elements. The failure rate and mission time are: Speed control hours 2.65 x 106 hrs i Historical failures 0 f l Assumed f ailures with 95% confidence 2.95 t , mi c s l on t i me 8760 hrs (yearly inspection) A, failure rate 1.11/106 Srs i 4.3.7 Power Supply, Elenents 0 & P The applicable historical data 15: Operation experience since 1963 , Power supplies in operation 86 Power supply hours 2.65 x 106 hrs ( Failures 0
T o ALLeo*CHAWGRG >%~>WBR BYSTEMS. INC 5 ER.504 10/75 _ ( 1 l For a high reliability of the EHC, we require one prinary ac power source, and either one or two independent de power sources depending on the reliability and characteristic 4 of the primary source. There are separation diodes and an ac/dc as well as a dc/dc converter in the EHC cabinet for the internal power supply. The failure rate of the power sources cannot oc determined by A-CPSI sir,ce we do not generally desigo or furnish the power sources. However, in our total experience with EHC, we know of no failure in the internal supply, and we also know there bas never been a case of loss of all powir sources during operation of a turbine generator with EHC. Despite the limited operating experience we feel it is sufficiently conservative to use our historical data of zero f ailures for the power supply and power source and to calculate the failure rate with 951 confidence level: Power supply hours 2.65 x 106 hrs Historical failures C-Assumed fallare wi th 95% confidenr.e 2.95 ! t, mi ss ion t ime 8760 hrs (yearly inspection) A, failure rate 1.11/106 hrs l 1 4.3.8 Change Over Device, E l emen t R ( The applicable historical data is: ! i Operation experience since 1966 Change-3ver devices in operation 53 j Change-over device hours 1.45 x 106 hrs Failures 0 ( j
i l ALLNE*CHALMRRB POWER SYSTEMen. INC.S E R- 504 10/75 ( During normal turbine-generator operation this device is in its safe position, and also does not have to act in case of a turbine trip, it is only used during periodic testing of the trip systems. Thi s means the probability of an overspeed failure due to the :hange-over device is extrerely srull, and the historical data yields an unrealistically high failure rate for this device because of the relatively small experience t i en . The change-over device shown in Drawing NC 4.07, is mainly the change-over valve which conducts the trip fluid from the main trip valves to the turbine stop and control valves. During norr.al operation the solenoid valves, A, B and C are in their u.fe (not energized) position and the ; I lower chamber of the change-over valve received control fluia from i ! i solenoid valve A, while solenoid valves 8 and C are draining the chamber of the change-over valve. During the bi weekly turbine trip test with the Automatic Turbine Tester (ATT), the solenoid valves B and C are positioned to supply control fluid to the upper chamber and the solenoid valve A drains the l owe r chambe r. The change-over valves moves downward and the trip signal from the normal overspeed trip devices to the stop cnd control valves is interrupted. However , be fore thi s change-ove r is initiated, electrical trip circuits are connected to the solenoid valves 8 and C to trip the turbine in the j l event that a real overspeed should occur during testing. Part of the ATT l program is a test of these electrical overspeed trip circuits and a start j of the trip device telt program is only possible after a successful test of these electrical overspeed trip circul's. The electrical overspeed trip I l l ( )
e o a$ 5: k $ Ec *
".S .
I e I 1 g
- N l fVISlC,N '
j\}p l9 f $ iM H 4 y b'Jg &
/( /we. 'oi i L
1
, , n&
' S;\'C' A A 4>
bg igi LJ c\ \ r, r, j dl el rz i ! 9 ^ r NNsg j < i
/ / :C w 4[{
i i E 5" v i xsNN w NN si E5 ,
, ;.d -
L_J p _ y:ms os Eq i s m s - yk - m N 'M lP > z 55 m
%. 8 ,
3
}"z 2
. :E
"\ . 1 b1 E
gsxx' w - s s v , Ee_ is k' N"
'N : = E,UeE'
- i. g v =<5' M E h El -
;!g$
/ =>
x < <> e E e m. _ - E i dYhN1kN I i
'.N'\'N
!?g%
! ? 3 .3 t_J S
i17'1 hs et <z m EE5e
-vo e Es ssi, ibi lam n t- 5 C o ,E9".
-o
!! *}
- 5 i:;
Ga
;a 3
m5 s e ged2 5 5;E
!$ #!I "5 3 EE g4-5'
!: Jii -R" 52! <e2-i as $4 EE 3E Bo BoSE 2" _
25 s#E5 E [ hw o nze EN 5 %5 CHANGE OVER DEVICE au = c- c o. /' FOR TESTING THE TU3BINE , TRIP DEVICES NC 4.07
l 1 ALLIO-CHAL.MERS POWER SYSTFt 98, wC Q ER 504 10/7; T S is a one-out-of-two s!gnal to the one-out-of-two solenoid valves B and C. Af ter testing, the change-over valve Is reset in its normal position by opening the control fluid supply with solenoid v,sive A and draining the u rip e r tha-ber with solenoid valves B and C. Very high reliability for the resetting of the change over valve is provided by checking the position of the change over valve with a limit switch, and by two pressure switches sensing the control fluid pressure decrease behind the solenoid valves B and C, For practical purposes, the failure mode of this change-over device is i simultaneous failure of all three solenoid valves. The average failure rate of solcnold valves is approximately 10 per 106 hrs per Reference Ill). table D.I. This failure rete i s ve ry cor,serva t i ve for our case because
;Se solenoid valves need not act during a real turbine trip; they only have to stay in the safe (not energized) position.
l Since all three solenoid valves must fail, the failure rate of the ! i change-over device is calculated as a redundant system consisting of I three solenoid valves as follows: l l t, mission time 336 hrs (bi weekly test) A, failure rate of a solenoid valve 10/106 hrs ! I EFR, the ef fective fallbre rate of the i one-out-of-three solenoid valve system i used as a model for the change-over device for mission time of 336 hours 0.0001123/106 hrs l l l l I i N j
AL. LIB *CHALMetRB F.QwaR BYSTEMS. NCt C ER-504 10/75 - I This is calculated as follows. , P * (P . ) = failure probability of I-out-of-3 valves s, stem sol. va ve EFR
- t =
( j, vajve a t) sol. valve EFR (10/106 hr)3 (336 hr) 6 EFR .0001129/10 hr (e f fective failure rate for t = 336 hrs) EFR = .0001123/106 hr (exact computed value for the 336 hrs mission time) 4.3.9 Main Trip Valve, Elements S I and S 2 The applicable historical data is:
- Operation experience since 1958 ,
! i Main trip valves in operation 278 Main trip valve hours 1,43 x 107 hrs ! i Failures 0 i in case of a trip the main trip valves drain the trip fluid pressure initiating a clori19 of the stop and control valves. The main trip ; valves are arranged in a one-out-of-tm system and are spring loaded in draiaing direction. An auxiliary trip fluid pressare holds the main trip valves in closing posi tion. This pressure will be released i j if one of the trip signals actuates a trip device and the main trip i ! l valves move by spring force tc their draining position. Cor: s i de r i n g this simple reliable device, it is very conservative to use the historical datc with a confidence level of 95%: k ;
AL.Loe-CHALMERS DOWRR SYSTEMS. lNC. - . ER-504 10/75 -
,- ~
l ~i Main trip valves in operation 273
)
Main trip valve hours 1.43 x 107 hrs l l Historical failures 0 Assured f ailures wi th 951 confidence 2.95 l t, mission time 336 hrs (bi weekly test) 4, failure rate 0.206/106 h rs )a l 4.3.10 overspeed Trip Test and Resetting Device, Element T The applicable historical data is: Operation experlence since 1953 Overspeed trip test devices in operation 242 Overspeed trip test device hours 1.4 x 107 hrs Failures 0 I Like the Change-Ove r Device E lemen t R, the "Overspeed Trip Test and ' i Resetting Device" T is only used for testing, and for resetting the I trip devices af ter shutdown and tes ting (see Drawing E5.122a). During normal turbine-generator operation the overspeed trip test and ; resetting device is in its safe position and need not act la case of a real trip. i The manual overspeed trip test device and the manual overspeed trip l reset till be used only for local testing at the turbine. Operating Instructions call for a yearly test or test af ter ar inspection of the l L ;
~ ,
nM l ., M $. . .* 4' A< R o 6> u M .J M w a b
*j E
Zd2 . Oo k
*= (J
.J E
o { es . - 6eJ u.= Med, w M i .M., y
.._. h
. A.
-l / h i . l ,, a .
..._<-:S
-g- g~
"., e w
cc y
>=
< -n 1 3- =1 3 -
0"- >> > > o, o
> e'-N- w w b, !
b 7..._.'. - _,~~ yUm
- a cc 47t- h G
!Ci
~
a
.>ien-
+ L .;' ' n . -
o m a, T
~ w
' < M Q ,
' ~^r.: ' M >E =
o
' I 3 , .
x e1 i i 3e I
)
- z. x .
{ g rc.n er.:r-- ,ur 1_
~ m--- 21.;t.=2 .-
--*2- *
; =t. .~ . . Er : 1
+ * ,, - 2 =._r,c = : ===I .
, . 4
- m ,
scLr,' r "R*~ " J: -
~,,p , _ I,., O ~
l * ['
^
j d FM , ,, - 4 i
- f
, r-- x.~_ .e2 ;k~ 4 l n
, j-2.r x . 2.na s- > l
.r== r= J j
/
]r 4 6" s u M.=
_.i e. - ,i # 3 w o"' ,
, w
.( " ]1* d a-ks.. < *
~
Q a. b .,~b - b' . I
- a 80 o [ ., *- 4
- 4
<, w w%
ww w. 4.
,"A. s
.g ,' wcr ' ,
i- ;3 w
- c. g., ,
L g w-k x, u
= '2 w
6
< pu h wh ah .L
. t.
G '4 w m m S M e- > ,cc- >-. 'le; <
. ww>- swe.
1t-
. . m M ws- O~c .
o
- w. . ' ,I,. e 3- w mM .
6 _ - a. L a . w w sm, -
- ~*-
i r z u ._ e s Lll.o k ,0_ a. :: j s '
~
g w, a 4 ~ w l * ,q g :,. - ,< m 34, = -Q s g d* m
. LJ
** W 4
; , .s* k" -
s
,b; ga
's -
o a.
>~
,+ -
3 s D 3 '* E+-i r
't [
e4 *CN
.; % s e
-e ;; - w : w y s
- s ,
< cc z 2 , -a g gg s Os h2S f.;^1 t; e>H];
- s z_ ?s;, tz o wm em I
oo cc m .J
<w e
p 0_', J- p si s is e is; a " O-( 5, x 4 4 Mas;1 k; r!I
~,I$.
o E
>. h' W>o>W<aC
.#.EE3 -o h!
}!
---4 QJ H=-
T W =f f oo E .- t
!; 3 ja w m o
- + _. -
xw o
< .,. c 5
'I . W~-
g
- u. . .; a "L- > .(
w
.s . r r- Q
.J o
cc e o u s ALL,n *CH AL ME *tte OVERSPEED TRIP TEST c>w. a . v a r. . ,~c y AND RESETTING DEVICE f E 5.12 2 a
ALL 98=CHAL.MERS POWER SYST*EMdF. INC$ ER-504 10/75 *73 _ l t ip system only. The test 's required to check the adjustment of the j two pressure switches of the test oil supp', to the trip bolts. The setting of these pressure switches during automatic testing with the ATT indicate at rated speed the proper function of the two trip bolts. i A nalfunction of this manual trip test and resetting device in tile s top and cont rol valve openir g di rection is very unlikely because the pistons are spring-loaded to return to their safe position after tney were pushed down for testing. Testing is only yearly and locally at the i
; turbine. The two pistons are supplied with key-locking devices wi th which the pistons have to be ;ocked in their safe position except when required for test.
i For this extremely safe mar < t rip test and resetting device, we assume , the same effective failure rate level as for the Change-Over Device R l i i because all the above precautions against naltunctions make 9 e manual overspeed trip test and resetting de ice as safe as the change-over device. Since .here are two pistons, the ef fective failure rate of the manual l overspeed trip test and resetting device is two times the rate of the l l change-over device: I t , mi ss ion t ime 8760 hrs (yearly test) ! i i EFR, ef fective failure rate of the manual g l overspeed t<ip test and resetting device 0.000225/10 hrs l l 1 1 l i l ( )
ALLt8eCHAL.MMR8 POWCR SYSTCMS. NC. A* , ER-504 10/75 _ f ) i The second function of the overspeed trio test and resetting device is ! the re,'ote resetting of the trip devices including the main trip valves and overspeed trip release devices. These functions will be used I bi week!; during each ATT tes'. The fluid supply for the resetting ; comes either from the starting and load limit device as auxiliary start up fluid, or as control 'luid. Due to the design of the starting ; i and load limit device, the auxiliary start-up fluid line can or.ly be ! pressurized when no pressure to the control valves is wailable. The control fluid is available whenever the control fluid pumps are running; therefore, it is reasonable to consider only the control fluid supply because the failure probability dua to on accidental control fluid supply > l is many times larger than due to an accidental supply of auxiliary start up fluid. l A nalfunction of the two reset solenoid valves could prevent an overspeed trip; however, in their safe (not energized) position the control fluid
; cannot block a real trip. During reset the ATT opens the solenoid valves 4
fe,r the centrol fluid and closes them automatically. During the test l and reset time the electrical overspeed trip described under " Element R" ! is active and would trip the turbine in a real overspeed event. The i l , two reset solenoid valves will be closed in sequence by closing first the I i second supply valve (lef t valve in Drawing ES.122a, page 72), and checking the pressure 3ecay between this solenoid valve and the main trin valves. This also indicates that no auxiliary start up fluid pressure is built up. Ther the first (right) valve will be closed and the pressure decay between the I two valves will be checked. With this valve and pressure switch arrangement, it is nearly certain that the valves wil' be reset to their ( )
AU 18*CHAL MERS POWER SYSTRMS, WC, Q,,' k E R- 504__I.c./ 7 5 r sa fe ( iot ene rgi zed) oos i tion , end cannot orevent a ceal overspeed t rip. To tale a conservative approach, we used a f ailure rate of 10/106 hrs for aart solenoid valve regardless of the fact that for a real trip the reset solenoid valves remain in their safe (not energ!;ed) pos i t i on . The l failure rate and mission time are as follows: l t, mission t!me 336 hrr. (bi weekly test) l 1 Failure rate of a solenoid valve 10/10 hrs ; i
., failure rate of the rn.4ote resetting ,
device "; th one-out-of-two solenold 6 valves 0.0335/10 hrs 4.3.11 Overspeed Trip Bolts, Elements U 1 & U 2 The applicable historical data 1s: !
, i i
Oper. tlon experience since 1958 l Trip bolts in operation 486 Trip bolt hours 2.78 x 107 hrs j , Failures 7 i The operating experience ircludes all the trip bolts which are automatically or at least manually testable at rated speed. One of the ! seven failures was caused by ventilation suspended particles from an j abnormally di rty envi ronment which entered the f ront bearing housing and orevented proper operation of the tri p bol ts. The remaining six failures i i were all due to fretting corrosion, which resulted in a sticking of the i trip bolts. l l r l l e ,
allee.cHAs Meine powacn overemo met & . ER-504 10/75 ' 7/" b i The trip bolt design is shown in Drawing E5.121, page 63. I Under certain ccnditions scch as large saaf t vibrations and aggressive air or oil, there ' can be a tendency to form tretting corrosion. i Tne best method to detect sticking of the t r i p bo l t s due to f re t t in g co r rof l on or d i r t is the i regular test of the i trip bolts which includes with the ATT an automatic check that the oil pressure level at which the trip ' solts move into tne trip position is correct. All seven t rip bolts failures were found durinc. manual testing, and there are no reported feilures of this type on a turbine equipped with ATT. Although we have no proof, we believe that the failed trip bolts were not tested as ! frequently as every two weeks, as is i recommended for units equipped wi th the ATT. ! During the past 15 years our ' large turbine generators were equipped wi th two separate trip bolts and separate trip release devices, and there is no case in which both nverspeed ' t rips failed at the some time or Juring the same test. I i j Regardless of the fact that bi weekly testing shou;d prevent future trip l bolt { i failures which happened in the past, we will take the conservative approach that a failure in the future is as likely as failures in the past . Therefore, the failure rate for the trip bol ts is: I Trip bolt hours 4 i l
- l. 2 78 x 107 hrs i Historical failures )
7 4
! Assumed failures with 95%
i confidence level 13.1 t, mission time I 336 hrs (bi weekly test) . A, failure rate t i
, 0.471/106 hrs I l
A LLse.cs.ALMacMe PowCM r*9 v.araMe. fNCO ER-504 10/79 ( 4 i ' 4.3.12 Trip Release Device, flements V I & V 2 4 The applicable historical data is: i Operation experience since 1958 Trip release devices in operation 486 i Trip release device hours 2.78 x 107 hrs Failures 0 The I trip release device is a simple hydraulic piston and spring system converting the movement of the trip bolt into a hydraulic signal. No
, failure of )
)
these devices has been reported, which results in the i following failure rate- ! i I , i Trip release device hours l 2.78 x 107 hrs be.torical failures O i Assuced failures with 95% confidence level 2.95 t, mission time
' 336 brs (bi weekly) l
. A, failure ra te 0.106/106 hrs 4.4 r Overspeed Failure Probability Calculation ! i i For purposes of computer calculation of the overspeed failure
! probability, the overspeed prevention sys tem was arranged in'0 system components 1 to 36 as shown in the reliability block diagram NC 5 02a .
which is equivalent to diagrams NC 5.0la, page 36 and E 5.120 page 41. l It can be seen that some of the components I to 36 are fortaed from several elements (A to V). Fo r examp l e , the trip device componen t No. I consists of eight elements including the change-over device R, two main trip ! I o l J
,h i, ,.
1 C. e (i 5 s fi w
/
! b
/ . iii:fZ!g n O' V E i
V j.!;85 z$ ,-w v v
\ g.4 -2+Ee
;tge i,
- : *g5%g t
.Y3.-, r-
-- =
na-a= 35
!be ll=
.!u Ii':
i !!': !!': !!' ' !!' ' $ "kI8 g,.. n.- [ _l } l l's l J } l * i G::l
-.1 te i
~
l'
,J - -
v v v
. i 8';r I Ills i
812: !!;t I f- it 18
- i .
IB
..III....iRRIC-4
!, ,i!
- s .:
jg a-
- -< a- -a
-) w.
.I s a ,- .
e .e ,R- e ,a s asl. i a t t s g .e 3
> r~g:g
~g e:r-rh ~
'] 12)\
IIkEEEEEEE!!. u ~r .e e. E.e..
.. E..
E. E E - : . - v
...g
. {, .
.~. _s .. ~-g:- . .
~
v et
. .... [
T ; t R R I ..E. R .8 8 81" .
- s. .- si.
,gi n,. g.,
.... p ..
---------g-
- ' p:
- it i:
i'*' ! c r a g ...... e r. .:'.: . 5 imi.:: g . . ~~. . :. ::. :. :.: g; ,, 4
.t ------.---. .
it- 'g 1 l g: , , .!. a= i jo ,g. . . :.a8aaaa m. l
; r. :* "::: A i i lji, li g'g y5, :----E . w d
a
- a a-i 3 s I*n'a 88 .t .te=a=j;: ,
1
..,i
. -.m a si
! i sul
,g g a:
,i El A
MSw .. . n......:::.r-.w.:..:.:.:~.:~.t:n sr.F: l yy [ I , .
. . .. ~. .. ... .. .. ,. . -~
. . .s
Au so-CHAL MaERea POWEM SwSTEMB,1942,. C ER-504 10/75 ' ( l valves $ I and S 2, the overspeed trip test and resetting device. T, te overspeed t rip bolts U l and U 2 and the trip releasing devices
! V i and V 2.
l I l
> The computer program calculatet- the system failure probability on the basis of the following commonly kncun principles of rellahllity analysis:
I,
.i
- l. Probabllity of failure: ,
P = l-e- t ( for exponetit At >0.3) t)' )3 )4 P = At - . , (for exponent it 10 3) where: i A = failure rate, failures per 10 hr t = mission time, hr
- 2. Series elements I (conservative equation because it assumes l ;
P=PI+P2 + ... that events are mutually exclusive) 3 Parallel (redundant) elements P = P, r. P2 * l Theire is no single failure of any component in the system which could lead to a .,120% speed event. However, there are mul*iple failure possiM lities which could lead to this event. The computer program evaluates all possible failure combinations
% J
Au.so-cnAL.Mesne powan everaMe.swS ER-504 10/75 ( listed in Olagram NC 5.02a, page 78 The highest order of failure is one ) ;
! quintuple simultaneous failure which could k ad to an overspeed event.
j l j The list of multiple failures at the bottom of diagram NC 5.02a, page 78, shows clearly the
!i!'
importance of the components I, anc' 19 to 36, which are the trip device, and the stop and control valves. ll Double failures of these 1 components could lead to an overspeed even t. The computer calculations show f I that the double failure combinations have the fol sowing influence on the overal1 fallure probabi)Ity results: 4-Flow LP Turbine 6-Flow LP Turbine Double Failure Combinations influence % influence % HP Stop Valves and HP Control Valves 13.22 17.65 HP Stop and Control Valves Cross-Connection 13.22 17,65 LP Stop Valves and LP Control Valves 13.22 26.48 Trlp Device (1) and HP Control Valves 4.45 / 5.96 Trip Device (l) and LP Control Valves 4.45/ i 8.96 Reverse Reheater Evaporation through HP Control Valves 28.71 - Total influence 77.27 76.70 The remaining 22.73% of the 4-flow and 23 30% of the 6-flow LP turbine of the overall failure probabili ty numbers are influenced by triple, quadruple and quintuple failure combinations including failures of the extraction system. l The failure probability of the overspeed prevention system was :alculated for two dif ferent modes of operation; first, for turbine generator load operation, and, seconc:ly, for operation of the uni t in the speed i k
- J
Au.so-cHAL.Mano powen ovantmMo.wo. C ER-504 10/75 ( ) l control mode wnen the generator is not connected to the electrical system. ! The fault-tree diagrams NC 5.03a and NC 5.04a show the failure paths causing an overspeed event. All exclusive fallure paths leading to a >l20% speed event have been taken into account, t he re f o re , the computed total f ailure i probabill considers all component fallure combinations which could lead to this overspeed event. The normal load operation node is analyzed in diagram NC 5 03a. The unit . is on line and as long as the generator is connected to the electric system, , there is no failure of the overspeed p. .vention system which could lead to an overspeed event. However the overspeed pre"ention system must be i continuously ready, and in an event of generator disconnection (load i rejection) it must act to avoid an overspeed by closing the control and/or stop valves. l The second operation mode is shown in diagram NC 5.Oka. This mode erists primarily during start-up and synchronizing of the unit, but also could j exi s t during shu tdown, in this mode, the turbine generator speed is controlled by the electro-hydraulle speed control, wi th the mechanical-l i hydraulic control as back up. A failure of the control system could lead l to an overspeed event by accidentally opening or holding open the control l l valves. In this case, the speed control system causes an overspeed, and
; i f the overspeed prevention system with the already failed speed control l
l cannot stop the speed increase, a >l20% speed f ailure co.:d happen. The fault tree for this mode of operation has one branch which initi .tes the overspeed event and a second branch which prevents the overspeed event. I
%-- )
-k- ~~
ll
,t.
a J oi 3~ .1 1 3 I$ ll 5. E$"5~$E
=.ii i l
[i!E{ff
- ,: --* ili l
q C . ' ' -J '. 3 13 L.
!(
j j j l
- - ~
*fgl.
S: l f << i l i , D I ij E'I
"$ w
.D () {
~
+
p-
~'-'
'H ao0h z
$3y!
; l
! lE!
5 I ! A A , , m:n b . i h kb bbbbbbd6bb n ew n e r e r r rue u$ t e b 4' -Q r d !N!
,pf!
l ,1;f!;I,[it .!! ,j:t! t .:i
.l:t!! s!:!! ,gt ii ,iti . g!!i j,n!!:i o li i a :I! q:1 q:!!
t a:t11
. 11! .t! !! !! ' 11
[ h h h a h h h h' h h!h h h h
+
hli Tl ii! n T , W i
.is . >
n ,' o. n e: ( n+ . . . - . , 1 r 1 c_ a g j 1.Q,,pgx
,. i ,
,x 7
i t < l :! !
- _ ._ J b_ . _ . _ !! ! i I r--
i 2 r' 4 1 A r.
') .
1 l+ ) W . I L._ . t2 - - . . - bb [' !.! i .i !
^
b'
--, F l , R S rH J S i! it!
i.Fi!!}!!!!lilll[4 r ry ttyt mm:!!!!ililhi
/ . . r ri .+1l F
- v n t. m evro11
- v. Q ja.gi . g v .s e n
A w n [l Y M h- : 1 l I w ; d 1
-a ,._ i i- ,i A rw -1g u 1
i . i. i!. . A 44!i u [4 -
%. ' 4 $.. i.ji:
.!l . .
-a 0 0 0 0 0 6 6 03 6 6 6 g3
.ii ygg .q i.
hm .' r= . . ~ , - . . . . , . . . . i :17."::* fi".7"..*."4" ** T :
$ \ o.mu 1
y e. I ] '* "f4,""W""Z".TJL'.
4 i s. I a i
- 51 ll:::: >+;: : - -
ll f-3 0 i!
- 1 i 1 i II i
isi:1
- .::: =1 pj _I
.i ;t
.I 3 !
L ..!!!! I IIM. i li 'ij. l::!!&a33
- ! . .l k s II;.ig.g.
Q !!!i.l u illi!!!!ll!!!"ljil.i li s!:,r! ne 4 6 !!iliiiiin;il:i.iiii
- tiliiinuni i
i Ini p<as sr up .T $$E 'E
~ ~ -
o boo q',i l,
-I:nnui;:.;.:.a.:.::.
a..d nos s I ::!s5 _g:e
- 5. n 2 a3; ;ig i,s:: y.; -pe
._e. . .
rii_,...ss!!; i, s_s
- ' ~'
a g I sii ' lii!
'o..
ill iij
= ,
l i, f
!. t 1:
i ! h 2 4lil i t nii .il i'Njjj iij il A,> oo o boo a A v a 4n i mL i I i rLL,I S a l ll: :l= l!i !!i !l ll: !) l12If l'l Toooo o0000 9 , 9 i O h 9 ,
, r,
,i:
~
iji ili rii ii,' ip ll. j ? j!i ,1: 1 ? j!= Too oTo ooo ooo
!*'- A ,
- 4. ..3+..,
; TA.E.T Ul O** E*, .
Wh ~*f " " . \ \ ,~'~.n = = ::::.
AL_1.ms-cHAL.MEMas PowarM sysTarMs. Wo. C ER-504 10/75 f The overspeed prevention part of the fault tree is basically the same as i shown for the load rejection event shown in diagram NC 5.03a or page 82, wi th : the trip failure and speed control failure as final results. The >l20% speed failure probability was calculated based on the two fau trees wi th a computer program wri tten by KW, The first results calc'Jiated with this program were doublechecked by Interatom* Independently and with a different computer program. The interatom results showed full agreement with the failure probability numbers calculated by KW. i The calculated results for the load operation mode are 0.8040 x 10 -7 per { unit year for a 4-flow turbin< and 0.6025 x 10 -7 per uni t year for a 6-flow turbine. These are the probabilities of a failure when the speed control system i s requi red to prevent an overspeed following load rejection. This calculation includes the conservative assumption that over a 40 year lifetime of the unit there will be 4r load rejections (one per year) recuiring proper operation of the overspeed prevention
,i system. This does not include transient load disturbances which do not require operation of the overspeed prevention system.
Also, it is important to note that normal turbine trip or shutdown i does not require functioning of the overspeed prevention system because tne generator will .:nly be disconnected af ter the turbine trip valves
- Interatom is a wholly owned subsidiary of KWU engaged in research, development and studies in the nuclear power field.
AU JO=CHAL.MEMS POWEM 8YSTEMB,ItYC D ER-504 10/75 85- ~ (
)
I
; are closed. This function is normally performed by a reverse power relay which provides a very high degree of assurance that the turbine valves are shut before disconnecting the generator. The renote possibility that the unit will be improperly disconnected can reasonably be included in the l above 40 events over the 40 year's Ilfetime.
l Tne calculated results for the speed-control operation mode are 0.8216 x 10 ~7 per unit year for a 4-flm turbine and 0.6158 x 10~7 per unit year for a I 6- flow turbine. These are the probabilities of a >l20% speed f ailure based on the conservative assumption that the turbine generator is operating disconnected f rom the network more than 336 hours per year. With this a s s ump t i e.1, any failure of the speed control system at any time could cause an overspeed event. I f the speed control would not prevent the runaway of I the turbine-generator to a >120% speed failure, in case the unit operates less than 336 hours per year speed controlled and not connected to the i system, this probability would be decreased. ! Despite the fact that in actual practice the turbine-generator runs mostly i
! wi th the generator synchronized, and that before operation with disconnected {
t
! generator (start-up), a testing of the overspeed trip device and the stop l and cont ol valves can be assu med, we took the conservative assumption that the turbine-generator operates f.)r more than 336 hours per year disconnected 3 and over the entire year connected to the network, and added up the two calculated results to arrive at a total of 1.6256 x 10 -7 per year for the probability of a >120% speed failure of a 4-flow turbine and 1.2183 x 10 ~7 per year of a 6-flow turbine.
( >,
l l 1
.u_,.-- -, -.,,.- - m I ; ER-504 10/75 #
T i l The different results for the turbine generators with 4* flow LP turbines i a and 6-f tw LP turbines are caused by the following two deviations of 1
- these uni ts
3 l a.) The 6-ficw units with three Instead of two LP turbines have }1 two more LP admissions with two additional LP stop and two 1 addi tional LP control valves. The LP turbine admissions with 4 i a stop and control valve in series are for a 4-flow LP turbine I l a "4 times 1-out-of-2 system" and for a 6-flow LP turbine a i I l j "6 times 1-out-of-2 system." 4 i ! l t j b.) For the 4-flow units with a much smallar moment of inertla, ! l there is a probability to produce a speed slightly above - i i } l :120% in case two HP control valves fall and allow a reverse stream of steam out of the two MSR's reheater tubes and end i chambers into the HP turbine. Even such an event would not i
- j produce an overspeed leading to a t arbine generator missile.
- i
+ i Ve took the conservative approach and adLed this event as a
>l20% speed probability for the 4-flow turbine.
- l. >
I i i . l_ j i ! The preceding calculations do-not . include the turbine extraction system f l because this system is designed and furnished by others. As shown' I in the reliability block diagram NC 5.01a, page 36, our equipment provides the i
- pest tive closing signal fe,r the controlled extraction valves formed ou, t
of the highly reliable hydraulle signal from the EHC speed control.. ! the MHC speed control and the trip syst m , However, it Is'the responst-j bility of others to design the extraction system'and select the valves, A careful study and design of this system and use of high quality, t ! rellable valves' are recocmended, especially for extracti.ons receiving-( + J l
l At.us-csauwrue mwCM mvmTOMas.uvc. O ER-504 10/75 steam from secondary sources and extractions containing large arounts j cf stored steam energy which could drive the generator to >l20% speed. I { Periodic testing of valves and good maintenance practices are ilso i j recommended for the extraction system. l l If these recommendations are followed, we believe the extraction system will contribute only a reasonable additional probability of failure leading to overspeed. We suggest that the extraction system should be designed to contribute not more than 50% increase to the >1201 soeed failure probability of the turbine-generator to control a load rejection event, l.e. 50% of 4.020 x 10-8 er unit year for a 4-flow turbine
~0 and 3.013 x 10 per unit year for a 6-flow turbine.
j Adding these values to the previous totals of 1.6256 x 10 -7 and 1.2183 x 10 -7 yields the final results of 2.0276 x 10- 7 = 2.1 x 10' per unit year for the 4-flow turbine and 1.5196 x 10-7 = 1.6 x 10 ~7 per unit year for the 6-flow
\
turbine as the probabilities of >l20% speed failures including the extractic
! system. ,
s f 4.5 Common Mode Failure The foregoing analysis is for random failures in which components are i assumed to be subject to f ailure as a function of time at a rate following I l the exponential distribution in accordance with the commonly knwn theory of { l l I reliability analysis, in addition, it is appropriate to safety analysis ! I to consider the so-called " common-mode" failures which may be defined : as simultaneous failures of more than one component due to some common ! 4 design error, improper operation or maintenance, or influence of some external condi tion. A pertinent example is the possibility of failure of Y ,
)l l
AL.ue-cnauwene nowan evarramen,m O ER-504 10/75 I ( )
! many components in the speed control system due to a fire in the i electrical cabinet hou s i n g. Due to the nature of this type of failure, an obj ec t ive and accurate quan t i tat ive es timate of cor.. mon-mode f ailure l probability is not possible. However, we have done a qualitative analysis t
l which is summarized below under the headings of: l 1 l 1. Normal external power plant environment i 2. Operation and maintenance errors 4 3 External events.
! l I
4 The Common Mode Failure Chart gives the results of this study. l l 4.5 1 No rma l E x t e rna l Powe r P l an t Env i ronmen t ! I The only significant influence of the normal external power plant i' envi ronment a re wa te r and elec t rical interference which could adversely ( affect the electrical part of the speed control. Because the electrical i speed control is only a cart of the overall speed control system, and is hydraulically connected to the electro-hydraulic converters ir. 1 parallel with the overspeed trip system and the mechanical-hydraulic l control in a one-out-of-three system, it can be assumed that a common
; f ailure due to normal external power plant environment will not change the basic failure probability value due to random failures.
4.5.2 Operation and Maintenance Errors The second column of the chart shows the possible common mode failures (
, J
D au.se-cs::u.mxne powen everame,ance N e 5 3 l
'-~ 20N MODE FAILURE CliART'l' i
{ I N0fMAL OFbil0N & i EXTERNAL Mti :f EN ANCE EXTEP;4L 385 eND EVENTS
=
-nq ENVIRONMENT ERRORS EVENT 5
.,. FAILURES , g m w w E
- s"l 5a ' s 0$ w h:v
,e e
a U, e 5c$e d
! w5 ;55 E g a st ;s 8I a
l s.-5
= '2 S
s>,a e dy w_ M m g _,on: o 55! t ! S w a@
- z. ' OP & COMROL VALVE 3 x'
I ! i
' 7;EED CONTROL SYSTEM '
!x x x; i X 'x x x
= ! , ! !
OVERSPEED TRIP SYSTEM i I . x'x X X
. . . ie
- Ddent:ai for r;mmen Voce Failure h C;mmon Mcde Failure' 'ead to a > U0*. speed event 4
I E4
= ,
1 t , August 4.1975 Te !
\ -
. . . )
"A M
alt _is-cHALMaras powara sysrarMs. wo. C ER-504 10/75 (
! due to operation and maintenance errors. in this area many different m
l common node failures are feasible. We believe the worst case would be careless maintenance of the entire system of stop and control valves, speed control and overspeed trip. In this connection, it should oe noted that there are no reported cases of common mode failures due to misoperation or improper maintenance in all of our historical data. We F.lieve that with current and future units wi th n dern operation and maintenance practices , and considering the excellent testability of the stop and control valves and overspeed trip system, the probability of such common f ailures 3hould approach zero. Automatic testing of the stop and control valves, as well as the overspeed tr!p system af ter maintenance work and before each start-up clearly will reveal any maintenance error or miscalibration. Improper testing and operator errors are eliminated by the very secure t design of the ATT and the local overspeed trip test and resetting device as described previously under Elements R and T, par graphs 4.3.8 and 4.3.10. Even if an operation error af fects the speed centrol of both the EHC and the NHC, the uni t is still protected by the trip system which is not subject to operator error. An operator error or improper testing of the stop and control valves is not possible, and as described previously, is very unlikely for the overspeed trip syst: . Therefore, we beiieve the probability of cocinon mods fallure due to operation and i
)
AU IO*CHAL. MEMO POWER OMETEMS. INC ' ER-504 10/75 ~91- , ( m, j maintenance errors will not significantly increase the calculated I i random failure probability.
, 4.5.3 External Events l External events could influence the electrical portion of the speed l control system due ;c fire or flood, but as in the case of environner.tal i
l effects this should not be counti3 as a common '~ilure of the complete overspeed prevention system. l t i i l For all other external events and components it can be safely assumed i I that if the system fails, it will fail in the safe (s top and/or i
! control valve closing) direction.
< I l i l l 4.5.4 tonclusion i Although we know that the failure probability due to common mode j failures is not zero, we believe that common mod- failures af'ecting i the whole overspe:u prevention system are suf ficiently unlikely that they should not change our very conservative value of failure l probability calculated for sindom failures. , i f1 \ ! l l l l : i l i i i i t
\ )
I 1 l j i =a a =-cw4aa mmme snowsm erworums,wo. j ER-504 10/75 L r ' I 3 LIST OF REFERENCES 1 1 ] j i (1) Aills-Lhalmers Power Systems, Inc. j Report No. ER-503, July 1975 " Turbine Missile Analysis for 1800 r/ min j Steam Turbine-Generators with 44" LSBs 1
- for Nuclear Power Plant Applications."
! (2) S. H-. Bush
" Probability of Damage to Nuclear "
j Components Due to Turbine Fallu e",
- Nuclear. Safety Vol. 14 No. 3 May-June 1973 t
(3) H. D. Emmert
" Investigation of Large Steam Turbine i Spindle Failure", Presented'at 1955 l Annual Meeting of ASME as Paper No.
55-A-172. 1 (4) H. Wolf and D. Sauer New Erperimental Technique to Determine Residual Stresses in large Turbine-Generator Components", Paper-Presented '
- at the 1974 Amer _Ican. Power Conference,
. Chictgo, Illincis. i (5) R. Schinn and U. Schieferstein " Forgings from Gigantic ingot with 140" l- ! Diameter and 881,000 lbs. Veight", l Part 2, " Operational Stresses and. l Acceptance Criteria", Paper Presented l at the 1972 International Forgemasters l Meeting, Cherry Hill, New Jersey. - l (6) U. Schieferstein and W. Wiemann " Influence of Material-Inhomogenel. ties l [ on Some Application: Properties of Steel l for Heavy - Forgings", Paper Presented . at the 1975 International Forgemasters. Meeting, Paris, April 20-25, 1975.. I (7) R. Schinn } Quali ty Control of large Castings' and ! r Forgings for Steam Turbine-Generators", l I 1971 American Power Conference, Proceed-i. Ings, Vol. 33. (8) Allis-Chalmers Ps ter Systems, Inc. " Speed Contro l of -1800 r/ min Steam [' - Engineering Report #E415, Noventer 1974 Turbine-Generators for ~ Light Water Reactor Appilcations." l 4 k l 1)
,., , + - - - _ - - . - . . . r -.. . - . _ .m .,,_-.,.w,.. .. ,..e,~ ,,1...,.~~ m ._m # ,.m .c.,_v..,w.,+.,v-,,,.,
E R- 504 10/75 (9) IEEE Std 352-1972/ ANSI N41,4 IEEE Trial-Use Guide:
" General Principles for Reliability Analysis of Nuclear Power Generating Station Systems."
(10) Igor Bazovsky "Rei? ability Theory and Practice." (11) B. J. Garrick, et al "HN 190 Rellbility Analysis of Neclear Power Plant Frotective Sys tems." I l I i i l
. i I
i l I I l l I L l
)
l i i l ( ENCLOSURE 4 TO TXX-92503 Supplement 6 to NUREG 0797, bafety Evaluation Report related to operation of CPSES Units 1 and 2, November 1984 Table 10.1 (page 10-9) I l-l-
V Table 10.1 Turbine system reliability criteria Probability, yr 1 Favorably Unfavo ably oriented oriented turbine turbine Required license action (1) P i< 10 4 Pi < 10 5 This is the general, minimum reli-ability requirement for loading the turbine and bringing the system on , line. (2) 10 4 < Pi < 10 3 10 5 < P < 10 4 3 If this condition is reached during operation, the turbine may be kept in service until the next scheduled outage, at which time t*' licensee is to take action to reduce P to 1 meet the appropriate criterion in Item (1) atove before returning the turbine to service. (3) 10 3 < P i < 10 2 10 4 < P < 10 3 i If this condition is reached during ' operution, the turbine is to be isolated from the steam supply within 60 days, at which time the licensee is to take action to reduce P t to T.det the appt3-priate criterica in Item (1) above before returning the turbine to servica. (4) 10 2 < P 3 10 3 < P 1 If this condition is reached at any time during operation, the turbine is to be isoltted from the steam supply within 6 days, at which time the licensee is to take action to reduce P to 3 meet the appropriate criterion in Item (1) above before returning the turbine to service. Comanche Peak SSER 6 10-9
1 i a 4 i t 4 i 1 i ENCLOSURE 5 TO TXX-92503 i 1 10CFR 7751 " Federal Register Vol. 51, No. 44, Rules and Regulations", March 6, 1986, page 7751 ] i i
+
?. 4 i 4 f 4 i P T i i i I l i e i f f i 6 l f i i c e
- , , ,- ,r...-,-..-.. - .,
} r m . . kc o te: i Lt. 31. .'. 4 , it.. s.24:.. .\farcn 6.1944 / Rules and riegulanas mi
) it.i nt i !.n;'2 or im a tutmg (ni) Fur a nuclear power reactor, a elsewhere in the nuclear industry or m j f.ahiv wdl hkly im Immel tu mvolve chance resultinit from a nuclear reactor other industries. and does not invnive a j wmf a.unt haeards unmulcrutions if enre relo.uling. il nu fuel unsemhhes sigmficant incrimse m the g rolwhihty or nta ianoo of the faulity in accordunca siumIn.untly dillerent from those found consequences of an accident previously witti the prupuseil amendnient mvolves prniuusly acceptuble to the NRC for a evaluated or create the possibahty of a
- une ur nmre of the following- previous core at the facility in questlott new or different kind of accident from til A wmficant rolasation of the are involved. This assumes tnat no ar.y accident previously evaluated
- and
, irNna used in establish safety hmits. sigm'icant changes are made to tha (2) The repaired or replacement j in) A sigmhcant relaxation of the acceptance entens for the technical component or system does not result in
- t.n. 3 for Imnung safety s) stem settmgs specihcations, that the analytical a sigmficant change in its safety
.a huntmg condowns for nperation. methods used to demonstrate function or a significant reduction m any i Imt A sigmtaunt relasaimn m hmitmg conformance with the technical safety limit (or limitmg condition of l inndmuns for operation not specific ations and regulations are not operation) associated with the i nomp nied ti u,mpensatory changes, sigmfic atly changed. and that NRC has enmponent or system.
i .undan.nt ur asimns that memtam a previously found such methods (x) Ac expansion of the storage
< omnwnsurate les el of safety (such as r.ccept.ble. capacity of a s),ent fuel pool when all of
.dlun mu . plant tu operate at full power (iv) A relief granted upon the following are satisfied:
dmu.. a peud m wni:h one or more demonstration of acceptaule operation (1)The storage expansion method ! ufii> % ieme a.e not q;.r.blel from an operating testnction that was consists of etther replacing eustu"yt l In ) Hent w a . . .,,;,c.t.r.g incense. imposa ce:ause acceptable operaHon racis with a design which allow: closer
- } ls For a nowa ar puer plant. an was not iet demonstrated. This assumes spacing betwen stored spent fuel nu.rrose m authonzed trwmam cora that the operstmg restriction and the l essemblies or placing additional raOp .
I' po wer in cl. critens to be applied to a request for of the original design on the pool floor if NI A change to technical relief have been established in a prior ' space permits: spu ilu aiions or other NRC approval, a a e 7{ s (2l The storage expansiori methed mvolung a egmficant unrevmwed does not involve rod conschdation or , g
- I**
1 i e in plant operation designed to unprove safety but which.
,[" Pj, ,,
*** P *"
, ,g h an 3 of the poolis maintained operstmg facility, a relief gren*ed feoen less than or equal to 0.95; and "
i due lo other factors, m fact allows plant (4) No new technology or unproven operatmn with safety margma an operating restriction that was imposed because the construction wee technology is utillaed in either the usmfu;antly reduced from those construction procese or the analytical hcIwved to have been present when the nM yet complewd saudactrHy.This la intended to levolve only restricdone technigsse necessary to justify the lu.enge was issued. Id. where it is justifled that construction e*Penelon. c.[, . y ,'f Nemtu una flazards Considerotoons Are {f 8'
- has been completed estisfactorily, U U ^ * **** * * * ****'
- g. T==p===== le Comments on Interim Y^*IR'I**
in some incrosse to the probabu'ity or tsurd BcIvw. The statement of ,,,,q,,,,, g , ,,,,4 _ _ i y,,g The comments are described m, i on oderations for the interim final rules accident or sney reduce in some way a somewhat greater deta.4 in an heled the following examples of attachneet to ENN. safety mercia but where the results of amendments the Commission th c are clearly wisia ed considered nnt hkely to involve 4cggg6 acceptates cetteria with to the snimlicant hazards considerations. 44 gy,wm ,, -p g g, g, JJ Ceaunenes- A group of I'R 14uw it explained that unless the comuentere state that the three speaho circumstances of a liceaeo' Sunderd Revkw Ptaw e abange resulung the the appusudes of a emag standarde la 5 50,92(c) are unclear and amendnient request leed to a contrary ,,gge,,,,g ,g , -- , g arp that the examples in the statement rum.lusion when measured against the calculetkaal M m h~ ,,,mesmi. of consideret one-which they believe standards in 150 92 then, pureuset to (vii) A change to osaform a license to are clearer than the standarde-should stie procedures m i 5491, a propaced changes is the reguladoes, where the be made part of the rule: otherwise. they amendment to an operatietl icenas ter a liconee change results la very minor . argise the euamplee have no legal facihty licensed under g821(b) se changes to faculty operedens clearly la' significance. 4 W = or for a tee willlikely keeping with the regulations. " -T De Commission disagrees he found to invi.ao (viii) A change to a lit. ease to reflect a with the request. As explained in hazards considere operettes of minor adjustraent la ownership shares roepense to the comments on the the fecihty in acc w6th the among co ownere already shows la the propeasd rule (see 44 FR 14464), the proposed amend-@~ __ __ only one license. Id. commentere ese correct that the or more of the folicwig (As discussed below, the Cmanission examples have no binding legal (i) A purely administrouve change to has added examples (la) and (m)la siginflomnos. However, they do provide techmcul specifications for example, a twoonse to comments on the laterim - goldense to the staff, licensees and to chenge to achieve consistency final rules.] the general public about the way the throughout the technical spoof 9.ations. (ix) A repair or . f- - t of e standards may be interpreted by the correction of an error, or a change in maior component or systes important to t'a===4amiam The Commission did nomenclature. safetv. if the following condittene are consider combining the standards and (ii) A change that constitutes an met: examples as e espise set of criteria in the additional limitation. nstnet:ca, or (1)De repair or replamt process lateria flaal rules, out decided against control not presently la.luded in the involves prec: ices which have been this because (1) the standardt and technical specifications. e.g. a more successfully implemented at least once examples had proved useful over time. I strmaent nuweillance requirement. on similar components or systems (11) the staf hed usad all three standards 1 o
*8 -
. ,,,w-- , e*-w-*ew-* v.-e-*vm* TNT"F N C"f'# T "T ~ V W #
i b 3 i
- ENCLOSURL 6 TO TXX-92503 i.
I NRC Safety Evaluation of Point Beach Nuclear Plant. Units 1 1 & 2 application for License Amendment ralated to Turbine j Vaive Test Frequency Reduction 1 i , t l l ' d i f i. i i i l f l l l 4 i 4 i i r
/,p* %'o, UNITED STATES i y
e(i i NUCLEAR REGULATORY COMMISSION WASHINGTON, o C. 20$55 gv , SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION RELATED TO AMENDMENT N05. 129 AND IJJ I? FACILITYOPERATINGLICENSEN05.DPR-24AND6$-2] WISCONSIN FLECfRIC POWER CSMPANY POINT EXCH NUCLEAR PLANT, UNIT N05. I AND 2
~
DOCKET N05. 50-266 AND 50-301
1.0 INTRODUCTION
By letter dated September 7,1989 (NRC-89-107), Wisconsin Electric Power Company (the licensee) requested amendments to Technical bpecifications appended to Facility Operating Licenses DPR-24 and DPR-27 for the Point Beach Nuclear Plant Units 1 and 2. The proposed amendments would change the specifications related to two subjects. First, the amendnnts would revise the frequency of the surveillance testing of the turbine sto; and governor valves associated with turbine overspeed protaction. Specifically, Technical Specification 15.4.1, " Operational Safety Review," would be revised by changing the test requirements in Table M.4.1-2, "Minir.om Frequencivs for Equipment and Sampling Tests." Item 18 in that table currently sp W.fies that turbine stop and governor valves shall be tested monthly. Footnote 10 in the table waives this reouirement during end-of-cycle operation when boron concentration may present a practical limitation. The licensee would instead of monthly. amend item Existing 18 to10show footnote would that bethis test is done eliminated. annually (11) and (12) would be Footnotes-renumbered as (10) and (11) and changes would be made in the table to correct the references to the new footnote numbers. The second proposed change is to Technical Specification Section 15.7.B.4.A.5,
" Administrative Controls, RETS Reporting Raquirement, Semiannual Monitoring Report, Meteorological Data." The current specification requires maintenance of date on wind speed, wind direction, atirospheric stability, and precipitation (if measured). The specificaticn gives three options for maintaining the data: strip charts, magnetic tape, or joint frequency dittributions. The proposed amendment would dele'.e the requirement for maintaining rainfall records and would delete tL< option of-maintaining the data in the form of joint frequency distributions. The amendment would authorize the licensee to keep the data either in the form of strip charts or as hour-by-hour dverages stored in electronic form.
2.0 EVALUATION 2.1 Reduction in turbine stop valve test frequency. In a letter to Mr. James A. Martin, Westinghouse Electric Corporation Generation Technology Systems Division, dated February 2,1987, the NRC 0111180201 911016 c PDR ADOCK 05000,266 . _ _ _ . . .
. .- . . = . , = . . - . _ -. _. _
2 staff presented its views on precluding turbine missiles and consecuential damage to safety-related structures, systems, and components. Uttiizing testing and inspection to maintain an initial small value of the probability of a turbine failure resulting in the ejection of fragments through the , turbine casing simplifies and improves procedures for evaluation of turbine
; missile risks and ensures th&t the public health and safety is maintained.
j In its letter, the staff provided generic turbine failure guidelines for i total turt>ine missile generation probabilities to be used for determining 1 (1) frequencies for turbine disc ultrasonic i'ispections, and (2) maintenance j and testing schedules for turbine control and overspeed protection systems. Subsequent to the NRC guidance letter, Westinghouse Electric Corporation ' prepared : report for the Turme Valvr Test Frequency Evaluation Subgroup i of the Westinghouse Owners Group. That report, published as Westinghouse Electric Corporation Topical Report WCAP-11525, "Probabilisi.ic Evaluation of Reduction in Turbine Valve Test Frequenr.y," June 1987, provides a detailed probabilistic basis for extending the i.esting intervals of turbine valves. In a supplement to e safety evaluation dated February 7,1989 for Northern i States Power Company, Prairie Island Nuclear Generating Plant, Unit Nos. l 1 and 2, the staff found the methodology described in Westinghouse Topical Report WCAP-11525 acceptable. ! A letter dated November 2,1980, to the Chairman of the Turbine Valve Test l Frequency Evaluation Subgroup, Mr. D.M. Musolf, Manager, Nuclear Support Services, Northern States Power Company, provided the NRC staff's generic conclusions regarding license amendinent requests for changes in the
- surveillance intervals for turbine valve tests and the applicability of i WCAp-11525 to support these requests. The letter to Mr. Musolf included a model technical specification for turbine overspeed protection to guide i i licensees in preparing an amendment application. The letter Jlso set forth l the following plant-specific infonnation which the staff would need to
- review license amendment requests proposing changes in turbine valve test r frequencies:
(1) If applicable, provide a justification for us total turbine missile generation probability other than less than 10"g yofr" i (2) Provide a cons.'tment to work with the turbine vendor to maintain a turbine valve data base in the purpose of tracking changes in valve component l-failure rates. l (3) Provide a commitment to accumulate valve f ailure rate information in a maaner accessible for staff audit, and that this information will be reviewed by the licensee at least once every 3 years and updated when
- more than minor changes occur in the data.
i (4) Provide a connitment to review and reevaluate the turbine velve l test frequ?ncy probabilistic analysis (by the Methodology of HCAP-11525-A)
- when any mor changes to the turbine system are made in accordance with
! 10 CFR 50.M or when a signif t: ant upward trend in the valve failure rate ! data is id 2 ified. l i I .
., . . _ . . . . . . . . - - ~ . - - - - - -- --
l 4 1 The WEPCo application, dated September 7, 1989, was submitted prior to the issuance of the guidance to the owners subgroup. Nevertheless, the stindard represented by the guidance to the owners subgroup is applied in this 2 evaluation. It is also noted that the Technical Specifications for the i Point Beach Nuclear Plant were published before the development of the
- Standard Westinghouse Technical Specifications. The difference in format precludes direct adoption of the rodel supplied to the owners subgroup.
However, the technical speci : cations at Point Betch should include the substa 'ive requirements of the model provided to the owners subgroup. The ir. formation provided by UEPCo is evaluated relative to the information requirements set forth in the November 2, 1989, letter to the owners grcup as follows: (1) The testing frequency proposed by WLPCo is consistent with staff criteria. The licensae states that "WCAP-11525 demonstrates that the mean ! annual probabilities of tuM 're missile ejection for PBND Units 1 and 2 for turbine valve test intervais of up to one year remain well below tha NRC criterion for turbine valve missile generation probability." Therefore, there is no need for justification of an alternate probability. The proposed testing frequency of once per year is tnerefore acceptable. (2) In its application, WEPCo stated that "Westir.ghouse currently maintair.s a turbine valve failure data base and is developing a method to assist
- licensees in the review of failure-rate data and turbine valve test frequency changes, when required." The staff accepts this as a commitment on the part of WEPCo to work with the turbine vendor (Westinghuuse).
(3) WEPCo states in their application: "Since changes in failere rates ) would affect these probabilities, and thus the periodicity at which turbine , valve testing is required, Wisconsin Electric will review, at least once every three years, turbire valve failure-rate data to determine if the testing frequency requires modification." This commitment meets the information requirement with the understanding that the data vill be subject to staff audit. 1 (4) In their application WEPCo st/.ted: " Additionally, anytime that major
. changes to the turbine system are made or a significant upward trend in I turbine valve failure rate is identified, we will review the turbine valve i test frequency." The staff believes that this commitment meets the t information requirement.
Overspeed is a potential cause of turbine failure. Governor va' and stop valves are central to the turbine overspeed protection system, ans lus to the probability of a turbine missile. Point Beach .as experienced a sub-stantial improversnt in valve failure rate sirca tue original technical , specifications were issued. Prior to the use of All-Volatile-Treatr.ent (AYT) for secondary chemistry, the periodic cycling of tne turbine valves associated with overspeed protection provided assurance that the buildup of chemical deposits cid not i i
- as i
p , I t
- i cause mechanical binding of the valves. The AVT type-of secondary plant
! chemistry control reduced the specific failure mode of chemical deposit binding to a negligible value. In establishing a valve failure rate, the ] WCAP-11525 evaluation uses the fact that valve failure due to chemical i deposit binding is now virtually non-existent. i WEPCo proviced information to show that the Point Beach turbine stop aN } governor valves have been very reliable, especially since early 1975 after l both units switched to all volatile secondary chemistry control. Since January 1975, Unit I has undergone 174 surveillance tests and the Unit 2 valves have been tested 182 times. There were three instances where the i Unit 1 turbine valves did not fully close. In each of these instances the j appropriate governor valve fully closed and no threat of tuttine oversp,eed j existed. It is significanc to. note that the tests have not resulted in a
- failure to close since March 1978.
1 i In the same time- period, Unit 2 experienced no turbine stop valve failures j but two instances where a governor valve did not fully close. With the turbine stop valves closed, there was no threat of turbine overspeed.
- The licensee claims that the testing itself resultr. in an increment of risk.-
Existing Technical Specification Table 15.4.1-2, " Minimum Frequencies for
- Equipment and Sam. 'ing Tests," requires that turbine stcp and governor valves be functic ally tested monthly except during periods of refueling i shutcan and during end-of-cycle operations when the primary coolant boron I concentration is less than 100 ppm. The licensee conducts these functional
! tests by cycling each valve to the full closed position and-returning it j to the original position. Although not in the technical specifications, the ! licensee advised that the Plant Procedure for the tests requires direct observation of the valve mover 2nt. l To conduct this test, reactor power must be reduced to about 50 to 60 l- percent of full power. This power level corresponds to the maximum steam j flow which one' set of one stop and two governor valves can provide. The reactor power level reduction is achieved by adding boron to the reactor j coolant system and by incerting control rods. When-both sets of stop and ! governor valves have been tested, reactor power is returned tc pretest 1 l-conditions -by withdrawing control rods and by ruoving the ac'ded boron by J processing the reactor coolant in the chemical and volume control system. j- Each valve test with the prcrequisite power twing places an additional. r thermal and pressure cycle on plant equipment, increases the amount'of
- radioactive waste from reactor coolant processing to remove the added boron,
! atd places the plant in a condition more vulnerable to an inadvertent L r%ctor trip. Additionally, later in core life, these power swings cause ! alal power fluctuations and divergent xenon oscillations during which core
- power stabilization-becomes more difficult, i
5 L 1
l l l l l l l 4
' Us ;ng the dCAP methodology, the licensee found that the mean annual probabi- j lities of turbine missile ejection for turbine valve test intervals of up to )
1 year remain well below the NRC criterion for turbine valve missile genera- i , tien probability. Because new experience with valve failure rates could l
- alter the probabilities of turbine missiles calculated using the WCAP l methodology, the licensee has comitted to review turbine valve failure rate 1 data once every 3 years to determine whether the valve testing frequency
- should be modified. GPCo has also comitted to review the turbine valve testing frequency any time that major changes are made to the tuttine system, or anytime a significant upward trend in turbine valve failure rate 1
is identifi!d. The staff concludes that a reduction in frequency (using WCAP-11525 methodology) for turbine valve testing will reduce (1) the amount of time the reactor plant is placed in a vulrtorable position; (2) unnecessary stress due to the number of thermal cycles for piping, valvas, and other equipment; and (3) the amount of radioactive waste generated along with the resultant men-rem exposure. Based on the Point Beach Nuclear Plant operating experience, the benefits men *.ioned above, and the assurance by the licensee that they will meet the NRC guidelines for turbine missile generation probabilities and reduction of power transients, ;he staff finds the proposed ch'anges in j testing f requency acceptaole. The proposed change in the frequency of valve testing from monthly to
- annually is found acceptable as stated above, i
2.2 Meteorological data record keeping requirements. WEPCo proposes to delete the requirerant that " precipitation data be main- + tained if measured." The basis for their request is that indeed they do , not measure precipitation. The existing wording was not included as a result of a concern over precipitation at the Point Beach site. The exirting wording h consistent with the Westinghouse Standard Technical specification wh'ch would be applicable to those licensees which do measure l precipitation as weil as to those, like WEPCo, that do not. Removing the superfluous wording does not alter the intent of the technical specifications nor does it introduce the possibility for any different it.terpretation. Therefore, this change is considered to be adninistrative and has no safety significance, f WEPCo proposes to delete the option of storin; certain meteorological data in the form of joint frequency distributions rather than as hour-by-hour listings. The existing wording does not require that joint frequency distributions be computed and is not-included as the result of a safety analysis of meteorological phenomena. The wording is consistent with the standard technical specifications which would apply to licensees who calculate joint frtquency distributions as well as to those, like Point Beach, who do not. The licensze is not proposing changing any of its practices regarding
- the collection or maintenance of meteorological amendment nor would this
! amendment authorize any change in such praf.tices. Furthermore, the amendment 1
. . . s 4
l 4 ! does not introduce the possibility of any different interpretation of the 4 technical specifications. Therefore, this change is considered to be administrative and hcs no safety significance. Finally WEPCo has proposed changing the technical specification option of storing meteorological data in the form of " hour-by-hout averages listed on magnetic tape" to read " hour-by-hour averages stored in electronic ferm." This proposed revision would allow the licensee to take advantage of tech-nological advances in data hanoling. Although the change does authorize a change in procedures, it is confined to date maintenance and is not considered to be substantive. Therefore this change has no safety significance. The proposed changes to the meteorological data recordkeeping requirements j are found acceptable as stated above. t 3.0 STATE CCNSULTATION i l In accordance with the Commission's regulations, the Wisconsin State official I was notified of the propnsed issuance of the amendment. The State official had no comments. I
4.0 ENVIRONMENTAL CONSIDERATION
t j This amendment involves a change to a requirement with respect to the instal-lation or use of a facility component located within the restricted area as 4 defined in 10 CFR Part 20 or a-change to a surveillance requirement. The staff has determined that the amendment involves no significant increase in the , amounts, and no significant change in.the types, of any effluents that may be released offsite, and that there is no significant increase in individual or cumulative occupational radiation exposure. The Commission has previously issued a proposed findirg that the amendment involves no significant hazards consideration and there hss been no public comment on such finding (54 FR 49140). i Accordingly, the amendment meets the eligibility criteria for categorical i exclusion set forth in 10 CFR 51.22(c)(9). This amendnent also icvolves changes in t acordkeeping, reporting or administrative procedures or requirements.
- Accordingly, with respect to these items, the amendment meets the eligibility criteria for categorical exclusion set forth in 10 CFR 51.22(c)(10). Pursuant to 10 .FR 51.22(b;, no environmental impact statement er environmental assessment need be prepared in connection with the issuance of the amendment.
5.0 CONCLUSION
The staff has concluded, based on the considerations discussed above, that (1) there is reasonable assurance that the health'and safety of the public will not be endangered by operation in the proposed manner, (2) such activities wi'l be conducted in compliance with the Commission's regulations, and (3) the
- issuance of the amendments will not be inimical to the common defense and
} security or to the health and safety of the public. principal Contributor: Robert B. Sanworth Date: October 16, 1991 i __ . _ - __ _ -- m.,._ - , , , _ . , -
I ENCLOSURE 7 TO TXX-92503 NRC Safety Evaluation of Prairie Island Nuclear Plant, Units 1 & 2 application for License Amendment related to Turbine Valve Test Frequency Reduction l l l l l l t r- c- r sv.*
i
.f kj UNITED STATES NUCI EAR REGULATORY COMMISSION usecToN o c 20555 w\..v p
SAFETY EVALUATION BY THE OFFICE OF NUCLEAR P.EACTOP REGULATION 1 RELATED TO AMENDMENTS N05. 86 AND 79 TO FACILITY OPERATING LICENSES N05. DPR-a2 AND DPR-60 NORTHERN STATES POWER COP'ANY PRA1 PIE ISLAND NUCLEAR GENERATING PLANT, UNITS NOS. 1 AND 2 DOCKETS N05. 50-282 AND 50-306 1.0 INTRODUCTICN By letter dated September 28, 1987, 1987, and June 24, 1988, Northern States Power Companyas the licensee) supplemented requested by (letters d:.ted amendments to the Technical Spscificatiens (TSs) appended to Facility Operating Licenses Nos. DPR 42 and DPR-60 for the Prairie Island Nuclear Generating Plant, Units Nos. 1 and 2. The proposed amendments would change the Technical Snecifications by revising the surveillance tett frequency of the turbine stop valves, governor valves and the intercept vd ves associated with the tu-bine 1 overspeed protection. Surveillance testing of these valves is necessary to assure th7 perfor:.iance of their safety function in protecting against the consecuences of a turbine missile ejection accident. Specifically, Technical Specification Table TS.4.1-2A, dealing with the test frequency of these turbine valves would be changed from vonthly to a frequency consistent with the trethodology preser.ted ir, WCAP-11525, "Probabilistic Evaluation of Reduction in Turbine Valve Test Frequency," ed in accordance with ti e established NRC acceptance criteri 5f or the pt: bm:ity of a turbine musile ejection incident of 1.0x1C per year. However, the test interval shall not exceed one year. 2.0 EVALUATION The licensee tests six ! top valves, four intercept valves and four governor valves for each turbine during a typical monthly test. The periodic testing of the turbine valves consists of moving the valve stem from the position prior position. to testing, to full closed and returning the valve stem to the original The reactor power level must be reduced to approximately 55% to conduct the test because of the reduced stesm flow to the turbine generator and the limited steam that can bypass the turbine. The power level reduction is achieved by the addition of boron to the reactor coolant system which in turn must be removed when valve testing is completed in ordt to return the reactor to pretest conditions. The cycling of the reactor power as described above (1) places an unnecessary thermal and pressure cycle on he plant equipment, (2) increases the amount of liquid and solid radioactive waste that results in an increase in personnel exposure and (3) places the plant operator D . - .d ? U. . N
- Y
{ 1
2 in a vulnerable position where an inadvertent reactor scram is more likely during e the transient power reduction and increase. In addition, during such power swings even with the aid of control rods, it has taken several days for the power distribution between the top and bottom of the core to 7tabilize. The NRC staff agrees with the licensee that certain reload designs can be wch that power differences be ueen the top an(i bottom of the core are more sensitive to control and can devalop divergent xenon escillations when the power reducticn occurs during the middle of core life. Near the end of core life, stabilizing even larger differences in axial power distribution becomes more of a problem because of the larger isothermal temperature coefficient, lower boron concentration and larger differential xenon transients. Based on the above, the staff has concluded that the margia of safety is reduced when the plant is undergoing turbine valve testing. By letter deted June 24, 1938, the licensee informed the staff that during tu-operating life of the Prairie Island Units (equivalent to 29 years of combined operation) there has not been a single incident of an unplanned turbine overspeed nor a ringle turbine valve malfunction that could have led to a turbine overspeed condition. During this period, the Unit 1 valves have unde @ ne 112 surveillance tests and the Unit 2 valves were tested 104 times. There htve been two instances of turbine valve failure that occurred durino surveillance testing. In one case, the small (approximately l's inch) bypass valve was founa in the open position because it had failed to reclose during the previous surveillance test due to mechanical binding. In a second case, the reheat intercept valve failed to close due to binding of the operating lever on sne actuator. Both of these failures did not represent a threat to turbine overspeed because backup valves ( cperated as designed. In support of this amenc; ment request, the licensee submitted by letter dc.ted Septemair 28, 1987, an evaluation performed by Westinghouse Electric Corporation, the results of which are contained in Westinghouse Electric Corporation Topical Report WCAP-11525, "Probabilistic Evaluation of Reduction in Turbine Valve Test Frequency." This report provides a detailed probabilistic basis for extending the testing intervals for turbine valves. The probability of a turbine missile ejection has been calculated for turbines at the Prairie Island Nuclear Generating Plant, Units 1 and 2. The effect of extending the time interval of turbine valve testing has been included in the analysis. In a supplemerit to this safety evaluation, the staff finds the methodology described in Westinghouse Topical Re, cort WCAP-11525 acceptable. In a letter dated February 2,1987, to the Westinghouse Electric Corporation, Generation Technology Systems Division (Mr. James A. Msrtin), the NRC staf f l stated its belief that maintaining, through testing and inspection, an initici small value of the probability of turbine failure resulting in the ejectiun of fragments through the turbine casing is a reliable meaur, cf ensuring that the l objectives precluding turaine missiles and unacceptable s mage to safety-related structures, systems, and components can be met. Maintaining an initial small valuc o,' the probability of a turbine failure as discussed above simplifies and improves procedures for evaluation of turbine missile risks and ensures that the public health and safety is maintained. To implement this emphasis, the staff proposed, in the letter dated February 2,1987, turbine failure guidelines for 1
I l i I total turbine missile generation probabilities to be used for determining (1) frequencies of turbine oisc ultrasonic inspections and (2) mainbaace and l testing schedules for turbine control and overspeed protectice sy;; ems. In the letter dated February 2,1987 to Westinghouse, the NRC issue; reihb4ity criteria for raintaining the turbine in generation probabiifty less than 1 x 10~gervice per year fo;dealing witt % turbine missile a favorably-oriented turbine and 1 x 10~ per year for an unfavorably-oriented turbine. This provides adequate assurance that the guideline values of Section 2.2.3 of the Standard Review Plan are satisfied. The WCAP-11525 calculated rean annual probabilities of turbine missile ejection for Prairie Island Units 1 and 2 based on the available data, show gradual but steady increases in the missile ejection probabilities as the mean test interval increases from one conth up to twelve months. Thus, small increases in the test interval would net be expected to result in large increases in the missile ejection probability, in addition, the calculated values over this ragge of test intervals are all well within the acceptance criterion of 1 x 10~ per year. 1he staff, tierefore, considers that the calculated values for Prairie Island contain adequate ir4rgins for orotection against potential adverse effects due to l discrepancies in implementation. However, it must be kept in mind that, while the WCAP-11525 methodology is deterr:ined to be acceptable, all calculated values us Ng the methocology are external to the methodology and are subject to change due to the availability of more recent failure data tnan the failure data used in cilculating the values proviwed in WCAP-11525 for Prairie Islar.d. Units 1 and 2. The staff believes ' that, in considering missile ejection probabilities calculated by using the WCAP-11525 methodology based on more recent failurt data, the licensee should assure . that the test frequencies contain adequate margins for protection against potential adverse effects due to discrepancies in implementation. The staff requested that the licensee work with the turbine vendor to maintain ( a turbine valve failure database for the purpose of tracking changes in valve 1 failure rate. Information on valve failure rate will be included in the plant Updated Safety Analysis Report (USAR). The failure rate infcreation included in the USAR will be updated at least cnce every three years. The licensee was also j L requested in accordance with 10 CFR 50.59 to review and re-ava'aste the Turbine Yalve Testing Frequency probabilistic analysis (by WCAP-11525 n.ethodology) any l time that major changes in the turbine system have been made, or a significant i upward trend in the valve failure rate it identified. This matter was discussed l eno agreed to by +he licensee. l In conclusion, during turbine valve testing, it has been demonstrated that the l- plant is forehhat more vulnerable to undergoing a pla7t accident, and therefore the safety margin is reduced. Operating experieace shows that during 29 years I of operation there has not been an incident of unplanned' turbine cierspeed nor. a turbine valve malfunction that could have led to a turbine overspeed condition. Based on this operating experience, the Westinghouse analysis'ior Prairie Island which demonstrated how Prairie Island will meet the NRC guidelines for turbine missile gereratior, probabilities and a reduction in the frequency of power transients, the staff finds the proposed change in testing frequency acceptable. l
=y , ,.,e ~, .,.---,..m-_...,-w.,,-, ,,,,_,,,,.,e,
- - _ .._,o,_,,.,m.s ,_,%.m ,,-,,,,,,,.,n-.r.,,,.,
(- _. ? 4; 4 )
1.0 ENVIRONMENTAL CONSIDERATION
l An Environmental Assessment and finding of No Significant Impact.has been
- issued > ar these ' amendments (54 FR 5706, February 6,1989).
d
4.0 CONCLUSION
4 We have concluded, based on the considerations discussed above, that-(1) there is reasonable assurance that the health and safety of the public will not be endangered by operation -in the proposed manner, and (2) such ! activities will be conducted in compliance with the Commission's: regulations, 3 and tha issuance of the amendment will oct be inimical to the common defense [ and security or to the health and safety of the public. ! 1 Principal Contributor: D. C. Dilanni Dated: Februarv 7, 1989 v i 4 e i l (- l l l i i i t i f r } o T 'l .
, - - . . - - - ,- , . - - - ----n,, ,,---,..m--- , , - . . . . ,-n...n,-,-, , . . .~,,~-~w n.}}