A Successful Preventive Maintenance Program
-
Upload
jessica-sanders -
Category
Documents
-
view
219 -
download
0
Transcript of A Successful Preventive Maintenance Program
-
8/18/2019 A Successful Preventive Maintenance Program
1/8
-
8/18/2019 A Successful Preventive Maintenance Program
2/8
F a i l u r e
R a t e
X I
Preventive Maintenance Plan
. Inspection Plan
. Improvement Plan
. Recommendat on Items
Maintenance Records
Specification o f Equipment
Failure Data
L
e
/ / S
Bilm-in Useful
L i fe Year -oc r
ReZion Reqion
Repion
I
Ciaqnos i s Da ta o f Rena in i rg L i fe
Opera t ion
B
maintenance Data
1 Recomecdat ion o f i z soec t ion . i a o r a v e ~ e f i r
Rel ia b le 0pe ra : ion P la n t s
Fi g.
1
The Need f o r Devel opment of Compr ehensi ve
Preventi ve Mai ntenance Progr am
Thi s sys tem can i dent i f y the fo l l owi ng i tems
f r om t he manufact urer ( Hi tachi Ltd. ) to the us i ng
ut i 1 i ty.
1)
Recommendat ons f or i mpr ovement , or r epai r
i t ems, f or cr i t i cal equi pment and syst ems f rom t he
s tandpoi nt of h i gh ava i l abi l i t y and l i f e extens i on.
(2)
Recommendat ons of i nspect i on cr i t er i a and
i t ems f or cr i t i cal equi pment and components .
(3 ) Recommendat ons f or pendi ng ques t ons and
r emai ni ng mat t er s r el ated to r epl acement ,
i nspect i on of equi pment, and components .
Thi s system i s composed of dat a i nput s on
devi ces used i n the pl ant , the dat a base of each
equi pment , a common dat a base, and t he mai nt enance
pl anni ng and r ecommendat i on syst em
The st r ucture of t hi s syst em is shown in Fi g. 2.
Actual mai nt enance, operat i on, and f ai l ure dat a
of each power pl ant i s i nputt ed to t he dat a center.
I n par t i cul ar , t he mai ntenance dat a i ncl udes
di agnost i c data of deter i orat i on and t he r emai ni ng
1 i f e of equi pment i n t he power pl ant s. Nai nt enance
dat a ar e t hen anal yzed
f o r
t r ends o f deter i orat i on,
r emai ni ng l i f e
o r
i nspect i on resul ts .
Thi s dat a i s st or ed i n the common data base
o r
i n the equi pment dat a base.
The common dat a base i s used f or st or age of common
dat a such as pl ant operat i on r ecor ds, mai ntenance
schedul e and pl ant shut down f ai l ure data.
The equi pment dat a base st ores i nherent dat a of
equi pment such as di agnost i c r esul t s, mai ntenance
r ecor ds, and t r ends of remai ni ng l i fe.
The mai nt enance pl anni ng and r ecommendat on
syst em can prepar e the r ecommendat ons
f o r
r equi r ed mai nt emance i t ems at each pl ant,
systemati cal l y, wi t hout om ssi on, and i n a t i mel y
manner .
I
Pl ant]
1
I RK P an:Short Long Team I
Maintenance Work
.
Detection Crack and Diagnosis
. Nondestructive Test (NDT)
. Calib ration of Instrument etc.
Diagnosis o f Deterioration
and Remainina Lif e
I I
LJ
IMaintenance Records
I I
quipment Data Base
I
Trend of Remaining Life
CY
I I
1
Actual Data from Each Plant
Maintenance Records
Operation Records
1
I
Common Data Base
Plant Specitication
Operation Time
Start/Stop Times
Maintenance Schedule
Operation Plan
Fi g. 2 The S t ruc tu re of Hi tachi Compr ehensi ve Prevent i ve Mai nt enance
P l anni ng Sys t em
1993
PROCEEDINGS Annual RELIABILITY AND MAINTAINABILITY Symposium
147
-
8/18/2019 A Successful Preventive Maintenance Program
3/8
3.
REM
INING LIFE DIAGNOSTIC
SYSTEhIS
Rotor.
Blade.
Cas
ny.
Main Stop
Va ve
We have devel oped t he di agnosti c s yst ems f or
es t i mat i on of remai ni ng l i fe
f o r
bot h pr i mary
equi pment and auxi I i ar y equi pment f r om the
st andpoi nt of preventi ng pl ant f orced out age.
The di agnost i c syst ems of r emai ni ng
1
i f e f o r
equi pment ar e shown i n Tabl e 1.
By
way of exampl e, t he di agnost i c syst em
f o r
the
t urbi ne and generat or w 11 be expl ai ned.
Tabl e 1
Di agnosti c Systems of Remai ni ng Li f e
f or Equi pment
Equipnent Conponent
I
Diagnostic Het od
-
CREEP
lectric resistance test
Hardness test, Replica method
Replica method FAT GL
Hard etching
Turbi ne
h p
Fan
Shaft X-ray diffraction method
sys em On-condition monitorin g - lh.vibration
Optical micro-crack detection method
Generator I
Stator Coil
Motor D Map nethod
I
H-Y Yap method
Tube
Tube
Eddy current test for tube thickness
Data analyzer
Ultrasonic test (bl) for tube thickness
Trend of thickness with analyzer
(CAT)
El ectri c
Va
Ive
Con t r o l
Rod Dri ve
Mechanism
Packing actual operation data
carbon seal . Trend
of
cooling water teqperature
0 ring Comparison between life test data and
actual operation data
Beat
Exchanger
Pipe
3.1
Diagnostic system
of
remaining life
f o r
turbine
The st eam t urbi nes are operated under sever e
operat i ng condi t i ons, wi t h hi gh t emper ature, hi gh
pressure, da i l y s tar t - s top and l oad f o l l owi ng
operat i on. The mater i al of t he t urbi ne, under
thes e condi t i ons w l l de te r i ora te as a r es ul t of
damage f rom creep, f at i gue and t emper
embri t t l ement . I n the worst case, damage w l l l ead
t o dest r uct i on. Theref ore, t he i nspect i on and
di agnosi s f o r esti mati ng t he r emai ni ng l i f e of t he
t urbi ne i s i mpor t ant ( Ref .
2,
3 ) .
The di agnost i c pr ocess of det erm ni ng t he
r emai ni ng l i f e f or a turbi ne rotor is shown i n
Fi g.
3.
3.1.1
Electric resistance method
I t
i s known that met al l urgi cal changes and cr eep
damage accumul at e duri ng l ong- t erm use under hi gh
t emperat ur e, wher e car bi de compound pr ec i p i tates
and gr ows near gr ai n boundar i es. Thi s
prec i pi t at i on gradual l y decreases e l ect r i c
r esi st ance when cr eep damage occur s. The
el ect r i cal r es i st ance method ut i I i zes the
cor r e l at i on bet ween e l ect r i c r es i s t i v i t y and creep
damage.
3.1.2
Hardness measurement
I n order t o i ncr ease t he accur acy of eval uat i ng
t he r emai ni ng l i f e,
i t
i s necessary to obt ai n a
corr el ati on bet ween har dness deter i orat i on and
cr eep damage. Thi s r el ati onshi p can be der i ved
f rom the tes t of t he same,
o r
s i m l ar mate r i al as
t he or i gi nal one.
3.1.3
Observation
of
microstructure
The r epl i ca method dupl i cat es a condi t i on of
grai n boundary on r ep1 i ca f i l m The mcr ostr uct ur e
i s observed wi t h t he repl i ca met hod and cr eep
damage i s eval uated qual i tat i vel y.
For heat r es i s t i ng al l oy s teel , dete r i o ra t i on
has shown by car bi de compound pr eci pi t at es and
gr owt h near gr ai n boundari es. When t he
det er i orat i on cont i nues, voi ds appear on a t rack
where t he car bi de compound exi st s.
3.1.4
Temper embrittlement
Two met hods f o r measur i ng t he amount of t emper
embri t t l ement have been devel oped f or t he actual
di agnoses. One di agnosi s i s the grai n boundar y
et chi ng met hod. The degr ee of t emper embr i t t l ement
i s measured wi t h f r actur e appear ance t r ansi t
t emperat ure ( FATT) and proport i onal to t he l ength
and dept h of pi t s i n the grai n boundar y af t er hard
etchi ng of t he mater i al surf ace. We have
devel oped t hi s t emper embri t t l ement eval uat i on
t echni que and
i t
i s ready for f i el d use.
Another way t o measur e t emper embr i t t l ement i s
t he smal l punch (SP) t est met hod by usi ng met al
sampl i ng and a mechani cal t est. Thi s test i s not a
nondes t r uct i ve tes t , but onl y a- smal l sampl e (l ox
1 OX O. 5
mm
i n si ze) is used. I t i s poss i bl e to
r emove a sampl e f r om an act ual component.
Ac tual l y , i t i s not easy to det erm ne t he degree
of materi al damage, because t he fact ors of f ati gue,
cr eep, and t emper embr i t t l ement are compl ex.
1993 PROCEEDINGS Annual RELIABILITY AND MAINTAINABILITY Symposium
-~
-
8/18/2019 A Successful Preventive Maintenance Program
4/8
Ue s t r u ct i v e
1
I
I
E l e c t r i c Ha r d ne ss S amp l i n g
Res i s t ance Test Test
T e s t
T e s t T e s t
Po
ent a
1
Di f f e r e nc e
Decr ease
i n El ec t r i c
es i s tance
ecrease
i n
Har dness
E lemp.
FATT
P
Ener gy
Mai nt enance Recor ds
Fi g.
3
T he Di a gn os t i c P r o c ess of Re ma i n i n g L i f e Ev al u at i o n f o r Tu r b i n e Ro t o r
I n order to r ai se t he accur acy of di agnosi s, mechani cal , and t hermo-mechani cal cycl i ng
var i ous t echni ques ar e i mpl ement ed in combi nat i on. r esul t i ng f rom the st art and st op operat i on.
Mor eover, a1 1 t he procedur es are syst emat i zed on a The I nsul at i on det eri or at i on mechani sms and
comput er- ai ded basi s. Some of procedur es use t hei r detect i on method are shown i n Fi g. 4.
Expert Syst ems usi ng a knowl edge dat a base.
3.1 .5
Application f o r Turbines
r emai ni ng l i f e for t urbi nes have been appl i ed to
50 t her mal power pl ant s , and
250
equi pment du r i ng
seven years . As the resul t o f di agnoses , s i x t o a c r i t i c a l
mai n t urbi ne v l v e s
wer e
r epl aced. Equi pment
w h i c h
est i mati on methods and t hei r exper t syst em of
i nsul a t i on di agnos i s (Ref . 4, 5). These are
wer e eval uat ed and f ound to have adequat e
r emai ni ng l i f e w l l be di agnosed agai n in a f our - d i scussed be l ow
year peri od.
3. 2. 1
N Y Map
method
\\ %en new, i nsul ati on composi t es of m ca- r esi n
are a voi dl ess st ructure. Voi ds i n the i nsul at i on
The di agnost i c syst ems f o r eval uat i ng
t h e
ar e gr adual l y pr oduced by deteri orati on. Some
voi ds become l arge and combi ne wi t h ot her s. The
di el ec t r i c s t r engt h o f m ca r esi n then decreases
turbi ne ro to rs , f i v e t urb i ne
cas ings ,
and
e l e v e n
we have devel oped two new r emai ni ng l i f e
One est i mat i on i s t he N- Y map met hod t hat
caused due to vari ous f actor s af f ect ed
by
t he
n u be r of
star t -s tops
y ) and t he operat i ng t i me
l arge gener at ors i s caused by e l e c t r i c a l , Y ) . The dat a is t aken f rom the gener ator
operat i ng hi story.
3 * 2
emaining f e estimation ystem o f
det er m nes how much i n s u l a t i o n det e r i or a t i on
i s
stator coil in generator
I nsul a t i on det er i o rat i on Of the St at or Co i l s in
1993 PROCEEDINGS
A nnua l
RELIABILITY AND MAINTAINABILITY Symposium
149
-
8/18/2019 A Successful Preventive Maintenance Program
5/8
Part a
d scharge
test
;
Maxi mum
Fi g. 4 I nsul at i on Det eri orat i on Mechani sms and
its
Det ect i on Net hod
f or St at or Coi l i n Gener at or
The mai n f act ors i nf l uenc i ng i nsul at i on
det er i orat i on are the r epet i t i on of t her mo-
mechani cal f orc es due to st ar t s and st ops of the
power pl ant , t hermal agi ng, el ect r i cal s t ress
agi ng, and mechani cal f at i gue due to operat i on of
the pl ant . Each deter i orat i on rat i o o f the
i nsul at i on st r ength caused
by
t he above- ment i oned
f act or s can be obt ai ned f r om many model agi ng
t est s. Then, r esi dual br eakdown vol t age can be
obt ai ned by the pr oduct of the det er i orat i on
r at i os, as shown in eq.
( 1 ) .
Vr : r esi dual BDV ( Br eakdown Vol t age)
Vo: i ni t i al BDV
Ve: res i dual BDV due to el ectr i cal
Vt : resi dual BDV due to t her mal deter i orat i on
Vh: r esi dual BDV due to heat cycl e det eri orat i on
dete r i ora t i on
The i nsul at i on st r ength est i mati on curves of the
N- Y map ar e shown i n Fi g. 5 ( Exampl e of di spl ay
f am exper t system .
150
16
14
12
S M
tw
U
tb %
10
a
6
4
2
0
lm
0 1 2 3 4
-
q*lrpticil
hours Hr) x10o0O0
162217
hr
=5611 T i a a
VR = 73 54 I
Fi g. 5 I nsul at i on Str ength Est i mati on
Cur ve by N Y Map Met hod
The fi el d dat a was obt ai ned by exam ni ng t he
gener ator coi Is act ual l y used, aft er l ong
operat i on. The measured r esi dual br eakdo m vol t age
( BDV) and est i mat ed BDV of t he N- Y map aer e
compared. Both of them were near l y equal .
Compari son between t he est hat ed BDV usi ng the N Y
map and the measur ed BDV i s shown i n Fi g.
6.
1993 PROCEEDINGS Annual RELIABILITY AND MAINTAINABILITY
Symposium
-
8/18/2019 A Successful Preventive Maintenance Program
6/8
-
8/18/2019 A Successful Preventive Maintenance Program
7/8
3 . 2 . 3 E upert system for insulation
diagnosis
We devel oped t he expert systei n wi t h Hi t achi
workstat i on 2050/32 by usi ng a sof t ware
pr ogr amm ng t ool ESI KERXELI W .
I t
has a knowl edge
express i on of a hybr i d t ype consi st i ng of f r ames,
r ul es ( i f - , then- t ype) and C- l anguage program
The sys tem str ucture and j ob sel ect i on o f t hi s
system
o n
CRT di spl ay is s h o w
i n
Fi g.
10.
Each knorvl edge base i tem i s di vi ded i nt o modul es,
and t he system has a t ree- l i ke str ucture.
Thi s st r uctur e makes
i t
easy to add and r evi se t he
knwl edge bases.
After al I i nfer ence i t ems ar e execut ed, t he
i nspect i on r eport s , rvhi ch i ncl ude al l i t ems, are
pri nted out and present t he r esul t s of t he
i nsul at i on di agnosi s qui ckl y .
I .
F ig . 10 Syst emStr uct ure and J ob Sel ect i on of
Expert System
3 .2 .4 clpplicat ion
f o r
Generators
The r emai ni ng
1
i f e es t i mat i on s ys tem of the
gener ator st ator coi l s has been appl i ed to
60
generators over a 1 year per i od. As a resul t o f
di agnoses f or coi l s whi ch have al r eady reached to
l i f e expect ancy or c l ose to l i f e, 21 c oi l s of
pol yest er r esi n and 4 coi l s of epoxy res i n i n
gener at or s have been r ep l aced. Gener at or s whi ch
have adequat e r emai ni ng l i f e
r v i l l
be di agnosed i n
a per i od of f our years .
4.
RESULUTS
The comprehensi ve prevent i ve mai nt enance
pl anni ng system has been actual l y appl i ed to
mai nt enance pl ans of t her mal and nucl ear power
pl ant s . Ve obta i ned good res ul t s by i t s
appl i cat i on. For exampl e, a r educt i on of the
f orced out age rat e of nucl ear power pl ant s
(Boi
1
i ng Water React or t ype pl ant s by Hi t achi Ltd.
as mai n cont r actor ) i mproved f r om
0. 3
t i mes
/ pl ant - year to 0. 16 t i mes / pl ant - year . The f a i l ure
r at e in a t her mal power pl ant s al so i mproved f r om
1.4 t i mes
/
6 mont hs t o
1.0
t i me/
6
mont hs.
Fai l ure 'of a bo i l er t ube af t er i nspect i on ,
di agnosi s and repai r , has not occur red.
1. The comprehensi ve pr event i ve mai nt enance
pl anni ng syst emhas been devel oped Tor t he
systemati c, t i mel y pr eparat i on of
r ecommendat ons
on
i tems used for per i odi cal
mai ntenance and pl ant l i f e ext ensi on of power
pl ants .
have been devel oped f or t urbi ne r otor ,
gener ator coi , rotor bar , pi pe th i ckness ,
and
so
on.
2.
l he di agnost i c syst ems f or remai ni ng l i f e
3.
The det ecti on t echni ques for mater i al
f ai l ure and remai ni ng
1
i f e assessment al so
have been devel oped. That is:
(a) Combi ned ass ess ment met hod usi ng the
resu 1t s of el ec t r i c r esi s tance
measur ement , r ep1 i ca and har dness.
deter i orat i on wi t h t he N- Y map and t he
map met hod.
(c) Detect i on method of corr os i on pi t t i ng.
(b) Esti mati on method of i nsul ati on
The pl anni ng system can make a l arge
cont r i but i on t o a pr eventi ve mai ntenance progr am
of power pl ant s , especi al l y i n decreas i ng pl ant
f orced out age r ates wi t h th i s sys tem and
its
cont i nuous use.
152
1993 PROCEEDINGS
Annual
RELIABILITY AND MAINTAINABILITY
Symposium
-
8/18/2019 A Successful Preventive Maintenance Program
8/8
REFEREMES
1.
K. Onoder a, Mai ntai nabi 1 i ty Desi gn Techni ques,
1989; Uni on of J apanese Sci ent i st s and Engi neers
( J USE
.
2.
S
Mori et al . ”Di agnost i c Techni que f or
Remai ni ng L i f e of Operat i ng Power Pl ant s”, Hi t achi
Revi ew,
Vol .
69- 10, 1987
Oct.
pp
83- 88.
3. T. Tan et ai , ”Pr eventi ve Mai ntenance Technol ogy
f or Turbi ne Components” , Hi t achi Revi ew, Vol . 40,
1991 Apr i l ,
b . 2 .
pp103- 110.
4. S h i , L T an i g uc hi , ” Di agnost i c Techni que
f or Remai ni ng L i f e of Lar ge Rotat i on Coi l ” , The
Proc. of 17t h J USE Rel i abi l t y and Mai nt ai nabi l i t y
Symp. ,
1987
J une,
pp
295- 300.
5 K. Onoder a, ”Expert System and Appl i cati on f or
Re1 i ab I i
ty
and Mai ntai nab
1 i ty”,
Re 1 i ab i ty,
V01. 9-3, 1987 NOV. pp 44- 51.
BIOGRAPHIES
Kat sushi ge Onoder a, PhD
H tachi Works, H tachi . Ltd.
Operat i ng Pl ant Servi ces Cent er
1-1 ,
Sai wai - cho 3-chome. H t achi - shi , I barali i -Ken, 317 J APAN
Kat sushi ge Onodera i s
a
mannger
in the operati ng pl ant
servi ces center. Si nce j oi ning H tachi h’orks in 1961, he has
been i nvol ved i n re1 i abi I i ty anal ysi s and
qual i
ty assur ance
acti vit y in
a
nucl ear power pl ant.
H s PhD in engi neeri ng is fr omTokyo Uni versi ty in 1990:
hi s
B S.
in El ectr i cal Engi neeri ng is f romYamagata
Universi t y in 1961.
He
is
a member
of
I nternati onal El ectr otechni cal
Commssi on(1EC)- Dependabi l i t y ( I C- 56), and
a
di rector of
Rel i abi l i t y Engi neeri ng Associ ati on of Japan(REA ).
Yoomei Katoo, PhD
H tachi Works. H tachi , Ltd.
1-1 Sai wai - cho 3-chome.
H t achi - shi , I baraki - ken, 317 J APAN
Poomei Kat oo is a chi ef engi neer, Corporat e technol ogy i n
H tachi works. Since j oini ng H tachi , Ltd. in
1956,
he has
been engaged in i nstr umentati on and contr ol t echnol ogy
sect i on of nucl ear povi er and chemcal pl ant
i n
H tachi
Research Lab.
hi s
B. S.
in el ect ri cal engi neeri ng is f romKanncawa
Universi t y in 1956.
H s PhD in engi neeri ng is f romOsaka Uni vers i t y i n 1984:
1993 PROCEEDINGS Annual RELIABILITY AND MAINTAINABILITY Symposium
153
1
1