ML20247B655
| ML20247B655 | |
| Person / Time | |
|---|---|
| Site: | Brunswick  |
| Issue date: | 03/16/1989 |
| From: | Ballard B, Crellin G GENERAL PUBLIC UTILITIES CORP., LOS ALAMOS TECHNICAL ASSOCIATES, INC. |
| To: | |
| Shared Package | |
| ML20246L184 | List: |
| References | |
| NUDOCS 8903290440 | |
| Download: ML20247B655 (25) | |
Text
~'
T x
8!L"A*87H ANNUAL NATIONAL ENERGY DM840N CONFERENCE
/N " '
oPsaATIous ParvorrAnvE mumotancs l
quaLITT courInnsCE OP 7srasquzur PREVENTIVE MINTENANCE I
- 3. E. as11ard, sr.
Manager - TM1 QA ModificationsNyerations CPU Nuclear Corporattom i
l l
P.O. Boa 480. Trailer 24 Middletova, PA 17057 1
'Dr. G. L. Cre111a, P.E.
Assistaat Department Manager Engineerlag Evaluation Department Los Alamos Taehafrat Associates, Inc.
2444 Moorpark Avenue, Suite 209 San Jose, CA 95128 1
ABSTEACT (FM) and the potential for degradation of. equipment, it is essential to assure by asseurement and control that FM activities are accomplished at adequata levels of quality. This paper addresses the use of confidence 4
I estimatas for various quality measures, and the small-sample difficulty commonly encountered in obtaining such estimates, for FM activ'ities that are performed daily or less frequently during an operattag year. The paper presents layesian statistical methods for overcostas this difficulty and develops the appropriate relations fpr commonly encountered statistical situations. Illustrations of the applicscion and interpretation of the Bayesian anthods are compared to the commonly employed methods of forming confideoca statemaats. The paper also discusses the results from an Laitial pilot study of implementing the layestaa methods _to provide a means of obtaining rational measures of PM quality achievement for use in future quality assurance decisions imi=GDUCTIou The GPU Nuclear Quality Assurance Plans contain the following state-i.
amats:
"A Preventive Maintenasca Program including procedures as appropriate for structures, systems and composeats important to safety shall be established which prescribes the frequency and type of maintenance to be performed. Preventive Maintenasca shall be performed in a. timely manner to ensure that important to safsty items are adequately maintained."
8903290440 890316 y*
PDR ADOCK 05000324 P
PDR p
ELEVENTH ANNUAL NATIONAL ENERGY DIVISION CONFERENCE
~.
.7 1
The assurance of quality is of paramount importance to ancisar power plaats. This objective has transcended the ususi action of quality that historically has dealt with defects, or defectives, is produced items of hardware. The GPU Nuclear Quality Assurance staff has espanded their approach to quality to taclude the majority of activities routtaaly conducted is their anclear power plaats. Their assessmaat, and control, of quality thus esteeds into asay activities that have been classically lefa out of the verification processes moraally associated with nuclear 7-19y assurance.
While there is no doubt that this signals sa estressly high dedication to controlling all facets of nuclear plaat operation la a quality amaner with significant emphasis en safety, it also has created special diffical-stas for the quality Assurance (Q&) staff. It is virtually impossibis to review, monitor or taspect every activity that takas place. Many activities oesar at estreme1y amall frequencies such,that even ebat may be visued as a normal statistical lot sina is small at best. Necessarily, ga is forced to deal with small samples. At the same time, the concern for quality demands that the activities be toaducted satisfactorily with a high confidence.
-._ _._.Dafortunately; - -
- -1
-1===1sa ltachniques..of,estimattag. quality level,at high confidence requires mich larger sample size tham is practical.
This is partially due to the fast that'such estiastas are based os classical seafidense interval techniques and, in some cases, it is difficult to pick the appropriate quality parameter to be estimated. GPU Nucisar QA has been revisning this proklam is the past year and has asked los Alamos Technical Associates to assist them in developtag suitable techaiguas for forming confidence intervals from small samples.
The techniques discussed in this paper are' designed to allow the MMy Assurance staff to develop confidence statements not only la preventive maintenance but in other areas where getivity occurrence levels can be categorized as statistically low population events. FN activities routtaaly occur daily, weekly, monthly or,at less frequent intervals.
These astivities savolve a large maiority of ' equipment and systems ta the plant. Botatise of pumps, replacement of packing, adjustments of ses potats, tolerances; etc., are all important to maintain the quality of the unit and its operational readiness to function appropriately A daily FM oscars three-hundred and sixty-five tians a year. Using souventional statistical methods, poor confidence statements are obtained unless a significant amount of" sampling of the total population is conducted.
Quality Assurance staffs are not sufficiently sised within the suelaar Ladustry to produce the amount of sampling which would be needed. Sayestaa methods appear to be osa approach which can allow muclear Q& staffa to develop appropriate, confidence statements in regard to PM quality achieve-meat. Uban activity occurrences (let sise) are of small frequency and the processes too varied and numerous to 'esable a significant amenac of 9
8 2.2
ELEVENTH ANNUAL NATIONAL ENERGY DIVISION CONFEREN.CE
=
1 sampling to be performed which will provide desired confidence statements, the Bayesian approach can be the answer.
QUALITT MEASU188 Measurement plays a castral role in quality assurance. In its staplast and most direct form, measuremasta can demonstrate, and thus '
assure, that desired standards and levels are is fact belas achieved.
More importantly, measurements saa identify the ased for controlling actions when desired standards and levels are not being achieved. Providing quality assurance of PM activities thus necessitates the consideration of suitable measures of M quality.
h visu is takaa that a FM activity, like an itas of hardware, is giroduced by a process. This process, which ambodies the writtaa procedures and the skills and training of implementing mechamies, produces the product PM activity. As with a hardware product, the resultias PN product may be judged as good, defective, or deficient in one or more attributes. N oe judgments may be based on whether appropriate procedures were correctly followee, whether proper scumentation was developed, or whether the.
_ hardware.itself. was affecmed adversely._hs.,_ the quality _of;an individual
._c product PM activity can be measured either' by whether it is a defective activity or by the number of deficiencies in its tapiamentation.
h quality'of the process which produces the product PM activity any aise be measured. This is achieved by characterizing the process in terms of an underlying parameter which governs statistically the outcase of any application. N appropriate parameter depen $s on whether the FM quality is measured by defectives or deficiencies.
THE SERNOULLI PROCESS he the quality is characterized by defectives, the appropriate' parameter is the Bernoulli parameter. This parameter is interpreted as the probability,a FM activity is defective. This is analogous to the tossing of a ceia where the Bernoulli parameter is them the probability of obtaining a " heads" la a single toes. h a a group (i.e.. a lot) of FM activitias are produced over a period of time, some any be defective and same may be properly performed. N esast ausbar of defective FM activitias produced is determined probabilistically by the binomial distributions p(5lpL)=
p gg )bt (0$$
(E4 =%)
l where L = the size of the lot (i.e., the number of Pk activities)
B = the number of defectives in the lot (no us.re than the' lot size and no less than sero)
/
[
ELEVENTH ANNUAL NATIONAL ENERSY DMSlON CONFERENCE
' g.-
f: c
\\
\\
1
- r; ~ ; (
ye
- Ny.,,.
es and p = the probability of a defectivs activity.
Equaties (1) repressats the probability that exactly 5 defectives are produced by the process is a group of L.
& sampling of activities from the 1st allows two possibilities:
(1) it presents information concerning the underlytas parameter p that we can use with respect to the possible quality level of future lots and what to expect as a long-run average, and (2) it presents Laformation conceratag the number of defestives contained (or remaining) La the current lot. This latter point corsarns the actual quality level partsiains to ths current lot and thus is of immediate concera, while the former deals with what may happen la the long-rum. Both may be important items of Laformation for decistaa making.If the cascars is the long-run process parameter, them the statistisJal, process for generatias the reasics of.the sample are of interest. The exact number of defestive Pit activities produeed ta the sampling is aise determined probabilistically by the binomial distributions
~
p (1-p) ~
(Opg)
(EQ-2) 7(Dlps) = g g
vhere S = the sampla size (i.e., the==bar of Pit activities monitored) and D = the number of defective FM activities observed.
When the concars is the level of defects in the current 1st, it is ao longer necessary to be concerned about the underlying parameter. In this case, the primary concern..is the==har of defectJves in the lot. The exact number of defective Pit activities produced in tts sampling is controlled only by the number la the entire lot. The probability *of. observing exactly D defectives in the sample S is thus stves by the hypergeometric distribution:
St St(L-5)!
(L-5)t N
I
- Bt($-0)!
Li (3-D) I(L-8-MD) i Equations (2) and (3) bot,k represent statistical models of how the observed data (defectives) are generated. Seth can be used to maka estimates (measuransasp) of their underlying paramators. Equation (2) ta' appropriate when it is desired to control or assure the louerus' process average as characterized by the Bernoulli parameter. Equatima (3) is appropriate when coetrol of the defective level is desired.
TER POISSou PtocEs3-When a FM activity has a quality characterized by deficiencies, the 8 2.4
a ELEVENTH ANNUAL NATIONAL ENERGY DMSION CONFERENCE
~
1.
1 appropriate parameter,is the Poissoa parameter. This parameter is inter-preted as the deficiency rate or the long-run average ausber of deficiencies is each activity. It is analogous to 'the density of flaws found La carpettag. Even though the average density of flaws is==all, a single square yard any have amitiple flaws. N probability of having a specified number of flave la any given square yard is governed by the average fisw density. Wham a group of FM activities are produced over a period of time, some may be, properly performed and some may be deficient ta one or more ways.
N esact number of daticiencies la the group of FM activities is desarmined probabilistically by the Poissos distributions p(3lE)=
a (Os,se)
(EQ-4) where L s the size of the lot (i.e., number of FM activities) 5 = the ausbar of deficiencies in the lot (no less than sero) i and i e the deficiency rata (par activity).
J Equation (4) controls the. generation of.deficiascies and represents the prohalitlity'.: bat azactly 8 deficiencies are produced by the process la a stoop of L.
A sampling from a lot of this type also allows two possibilities:
(1)*
it provides information on the underlytas parameter 1 that can be utilized
~
with respect to future los quality levels, and (2) it presents information concerning the number of deficiencies 'contataed (or remaining) in the aprrant lot. Agata, this heter point deals with the actual quality laval of the current lot and thus is of immediate int'airest while the former concaras what may happea la the long-rus. Both are important items of information for decision-making.
When the concars is the long-run process parameter, the statistical,
process for gemarating the sampte results are of intarast. N exact number of deficiencies produced in the sampling is also determined probabilistically by the Poissoa distributtaa p(Dl18)=
e' (0,<0e=)
(EQ-5) where S = the sampla size (i.e., the number of monitored Phi activities) and D = the ausbar of deficiencies' observed.
N underlying parameter is no longer important if the concern is the level of deficiencies in the current lot. N total number of deficiencies 8 2.5
ELEVENTH ANNUAL NATIONAL ENERGY DMSiON CONFERENCE
~
q, u-
., aqf
- h.% -6?
I la the 1st already has been determined by the process and thus controls the probability of sheervias deficiencies ta the sample. Bowever, because the Poisses process ideally regstres the deficiencies are spread evenly over the lot, the assignment of the deficiencies to the sample can be viewed as a ts. r 111 process controlled by a racia of the sample to the lot. Thus, the probability of sheauving exactly D deficiencies ist D
S-5 I"
p(D[58L)=Dt
)i Equatissa (S) and (6) thus provide statistical models of how the p
Both can be used to prov14e estimates observed deficiencies are generated.
(nessuremsats) of their underlying parameters. Equation (6) ir, appropriate l
uhan it is desired to control deficiency level. nihan it is desired to esmerol the long-run process average as characterized by the Poissos paramatar Egnation (5) is appropriate.
~
MDENCE STATDerFS Ristorihally,confidencestatements,orconfidenceintervals,havebeen l
the esclusive-purview.of the classical _ stat _isticina.ussch intervals, or
~~~
the confidasca level for such intervals, provida decision makers with a~
measure of how certain they are, in the conclusions they might draw from observed data. Thus, the confideoce level is often used as though it is sa assessment of the probability that the true parameter value is within the interval. nihile this any met be strictly appropriate, based om schcle
====arte distinctions, it is none-the-less a consoa.and often aseful view-point for desistes'makars. Thus, while there are useful alternatives, the classical technique of confidence interval estination is widely used eyes though oftsa misunderstood.
Ttun Bayestaa theorist offers sa altern'acive to the confidenes interval.
This alternative more directly addresses the assessment of the probability that the true parameter is withis the interval. As such, the result is, strictly speaking, not a confidence interval, but a probability interval.
Bowever..because of the widespread misinterpretatias of the confidence laterval as a probability interval, the Bayesian probability facarval is oftas referred to as a leyestaa seafidemas interval.
In the mest sectime, consideration will be given to forming confidenpa intervals by both approaches.
CLASSICAL courIDruct STATEMartS The forming of confideoca intervals is a standard technigne of alassical statistics allowing a measure of the strength of inferences made only from observational data. Such intervals tall one what any.be 8 2.6
1 f
ELEVENTH ANNUAL NATIONAL ENER8Y DMSISN CONFERENCE u
tafarred about a parameter from only the data, or statistics, produced by The maderlying philosophy is based on the interp ner ios that N t parameter.
if one repeatedly samples sad forms estiastas of an interval using the eenfidence incarvel procedure at the same confidence level, them the persassage of an such intervals actuany containing the true parameter That is, with a confidence laval values win.egnal the confideoca.1svel.
of 905, one will be correct ta slaiming the true paramsta'e value is contained ta the estimated interval for 905 of the incarvala se estimated.
Mathematically, the upper interval estimate is defined as the value of the parameter for which the confidence level (Cu) equals the probability of Thus, for a observing a statistic greater than that actuany observed.
true parameter valus outside the upper incarval, the probability of seeing a statistia equal to or less than that observed is less than 1-Cu.
As a resuit, tbs probability of forstas an upper interval that does not contain the true value is less thsa 1-Cs.,'Thus, it aan be conversely stated that the
' probability of foretag sa upper interval that does contain the true value must be at least equal to the confidence laval (Cu).
' For' azample, if 'ose wishes to make as upper interv'ai estimats at the 991 confidence level on the underlying probability (p) of a " heads"
. occurring-la-a-single-toss-of.a_acia when_maa_haa_ observed two;" tails" in _
two tosses, them one would find from squation (2)' that p=0.9 (or' greater) provides a value of 0.99 (or greater) for the probability of oma or more
" heads". Staca one ar more heads did not occur, we say we are 991 confident that the true value of p is.ta the interval 0 to 0.9.
That is, there is a less than 0.01 chance for p > 0.9 that two tosses'would preduce as " heads" and lead us to form an interval that does not contata the true parameter.
Thus, there is at least a 991. chance that,the incarval contatas the true j
parameter.
j
.s.
This examp1's, besides illustrating the process of obt=*= M s' classical If upper confidence interval.,also demonstrates the "sman-sample" problem.
obtaining a " head" on a. toss is equivalent to a defective ectivity, them.
observtag no dalective activitiss la a sampling of two would allow an upper interval estimate of sono to 0.9 for the long-run percentage of defectives at the 991 confidence level. ilhile the interval is formed at high confidence, the interval itself is "as broad as a barndoor".
The 1starvs1 can only be nada'smallar at the expense of reducing the confidence level or increasin6 the sample size. Staca Classical confidence intervals are based solely on observed data, it is an incontrovertible fact that small upper intervals at high ceafidence can only be obtained from large sample sisks.
~
Cace the basic idea of forming confidence intervals is understood, it is a simpis, although sometimes tedious, matter to form intervals for any e
B 2.7
c....
- >.u.
- .i
.m... m=
ELEVENTH ANNUAL Nail 0NAL ENERGY DMSION CONFERENCE
.o f
i I
- %,4.*
parameter controlling a statistical procese.
In doing so, it is only ancessary that one understand and be capable of mathematically describing the probability of the statistics or observations resulting from the 1
Thus, it is a relatively staple matter to define confidence prosess.
1starvals for almost any statistical parameter of interest is a decision Appendix A presents the equaticas for forming conf Wence incarvals process.
appropriate to the statistical models discussed previously.
BATg5IAN CONFIDENCE STA,Tue rts Bayesian probability intervaIs are formed.! rom probability distributions To understand tM ir obtained from the application of Bayesism theory.
functions, it is necessary to first be fadme with the gayesian tuhaique.
The Bayestaa process is a useful technique for rationally amploying all that is knous about a specific paraseter. The taformation is both specific (e.g., direct test results) cad non-specific (e.g., subjective knowledge of the situation of interest). Non-specific data alone leads te some level of knowledge conceraias the parameter. For azample, one's asperience with tossing quarters, nickmis and dimes, while not specifis to
-the tossing 4f-a parti-w =" war _dellait.. is_palavant;and forms;a part of -
one's knowledge about tosses of that part1 W silver dollar. En a similar amener, tihere is non-specifis data conceraias PM activities derived from J
past experience sad engineering judsmaat.
The Bayesian process assumes,that this initial state-of-knavledge is just as important and valid after conducting specific' tests as it was bafore the tests are performed. Thus, the initial information should set be ignored arbitrarily, but rather it should be upgraded or modified to,
arrive at new conclusions based os both the initial knowledge sad the This use of non-specific'information in combination' specific test ra==1ea.
with specific test data contrasts markedly with the classical approach, j
which utilises only the test data.
l The Bayesian p'rocess aconsplishes this combinattaa of saformation by l
In assence, it l
employing mathematics to simslate a procaes of learning.
takes the taitial knowledge and modifies it proportionately to the chance l
et sestas the specific data. In practice, the initial knowledge is expressed j
anthematically as a probability distributes (as11ed the " prior")
representing one's uncertainty about thi parameter value. This uncertainty is than modified in light of the specific data. The teamit is a new probability distribution (cslied the " posterior"), which describes the sodified (sombined) hacutedge and uncertainty.
The resulting " posterior" distribution, when formed for a quality para-meter such as the percent defectives (the Bernoulli process) or the I
deficiency rate (the Poissoa process), can be used te determine the probability that the parameter value is within a swifia interval. Thus, 1
m
m
~~
~ ~
a ELEVENTH ANNUAL NATIONAL ENERGY DMSION CONFERENCE
~
\\
1 ene may detaraine (1) the probability that the paramatar is less than a specified value (an upper probability interval), (2) the probability that the parsneter is ' greater than a specified value (a lower probability interval), and (3) the probability that the parameter is between two spesified values (a two-sided probability interval).
These probability 1starvals, while obtained from a differsat philoso-phical basis, are the Bayestaa equivalent of the Classical confidence 1starvals. For purposes of making decisions, they may be employed La the desistas process in a similar manner. The major distinctica is that the Bayesian process. admits to some asasure of confidence (uncertainty) befora Aftar all, data are obtained and involves that confidence with the data.
someosa had enough confidence to decida for taplementation without, having i
sampling data.
The key to successfully empleytag the Bayesian' process is in realistie -
cally determining the " prior". Utmost integrity'is required in assuring that it truly represents a reasonable interpretation of the taitial knavledge This is and subjective data concerning the quality parameter of intarast.
the only way to assure credible postarter results.
MMifii macGr.aiisal formulatiibai of a""pridr"'i's'often~ chosen for-mathematical conveniemes to be a " conjugate prior". Such a prior has the characteristic that when combined with the data, the resulting posterior is of the same mathematical form as the original prior. It also has the Staaral characteristic that the prior may be incorpreted as though it represents p
'-ta obtained from pseudo-casting.
1 For asample, if the ;-11ey paramatars to be assessed is the percent defective (i.e., the Bernoulli parameter,, p) and the monitortag results are D defects is a sample of 5 units,' then the Bayesian technique would find the poetatier as follows:
I
~ ~ ' '
'(*I o(Dlos)
)
p(pjDS).=g
/ p(p)p(Dlp8)de where p(p) is the prior said p(Dlp8)*D1(5D)i II9)
M}
is the 1 N H W function (probability of obtaining the test results).
For the Bernoulli process, the
- conjugate prior is a Beta distributions l
1 Et&9 l
..x..s.
ELEVENTH ANNUAL NATIONAL EN$GY DIVISION CONFERENCE 9
_l e
-g of f
.s c
t s _p y
p*"I(1-p)61 (so-9)
' 93, r(e+s)
,e-1 gg,)6-1, (e(e+s-1)I r(s) r(s)
-L)t(s-1):
Thee, the posterior becomes:
(1 M MM p(plDs)=gg y p It is seem that the posterior is of the same asthanatical fo're as the It is alas seen that the prior with etD replacing a and 3+S-D replacing 8.
form of the prist, the' posterior, and the 1NHhaad function [p(DlpS)] are The prior parameter a appears to have the same role in the all similar.
prior and postartar as the number of defeats (D) has la the likalthood feastime. Stailarly, 8 appears to have the sans role as the sample sise Less the defeats (5-D). Thas, e-1 any be thought.of as pseudo-defects ebeerved is a pseudo-sample of ets-1.
This provides a simple laterpretattaa to the prior parameters and allmas For example, a prior assignment em the basis of " equivalent" test results.
-the-prior-kasvledge-asy-he-t990 of aa_the.squivalent _of_having _seen le _1_.
~
defectivas la a? sample of etS-1.
The upper probability interval is found by determining the area under the posterior probability distr'thutions N
e.&D-1(1-p> --
1de (so-u) pesalsD) = l p(,p(o+6&S) e
,s3,(,,p_s3 The above tatsgral can he expressed as a finite sum when e is $
.n.
tacaser. In this forma p (1-p)
(
12) p(p3*[DS)=1-dl(
y.
=1
)l g
and the similarity with the Classical upper interval, given by pk1@ M
,ps cuGlDs)=1-
.,, 4 3,
is
- d..bvi APPLICATION PILOT STUDT Demonstrattaa of the application of' the Bayestas rasmits to real plant 8 2.10
~
ELEVENTH ANNUAL NATIONAL ENERGY DMS40N CONFERENCE
\\
l i
data was accomplished by a pilot study.
In the pilot study, data generated by tem different procedures in a r.; yar time span were subjected to l
analysis. Table 1 lists the data for each of these procedures.
l The first major issue was to eht. priors which made sense. Ideally.
l the priors abould be selected by the % engineer and should be based on a 1
seabimattaa of historical experience, engineerias judgeest and assagensat l
espertise. In a practical sense, they abomid not be as weak as to I
duplicate the " classical" results; however, they should not be ao strong i
as to distate the conclusion for aceton without any consideration of the i
data.
In the case of PM sativities, there is as abaada~e of previous l
experiences. These have been utilised to shape the structure of procedures.
l It is fair to say that if there wasa's some adequate policies and training., he underlying process, them the engineer or the level of confidence in t manager weald not allav the activity to be implemented. It would be re-l vrittaa or re-defined until. it was viewed as having a raaaaankle chance of l
being implemented carrectly. If a anchasis was thought to be so poorly l
trained as to inspire ne confidence, them' he wedd not'he allowed to perform the activity, but would be replaced or re-trained until there was as
. adequate-level of confidences-The-levd--of-what's-adequate-may-vary...h*
it seems reasonable that there is some "a priori" level of confidenes or the process would not be taplemented.
On the other hand, the level of what's adequate anst not be that level eventually desired or else there would he no esed 'to give additional data. Tless, we are ma===Aa* like the gambler checking his' cards for fairness.
We are pretty sure, but not quite sure enough.
.1.
For illustrative purposes, the proeddures were viewed to be subject to.
multiple deficiencies. Thus, the Poissoa process was viewed as represen-
. tative. It was decided that the desirabia level was 1-0.05, and that there was an initial confidence of 903. This led to a prior assignment of a=2 and $=78. This prior was deemed as appropriate for each of the procedures. This, of course, was eiguivalear to havtag zampled 78 procedures and observtag 1 deficiency.
Figure 1 shove a plot, of the confidgmes levels,for >0.05.
The classical results are obtained by starting at the origia (5'=0, D'=0) and aeving along a until the sample size is reached, and them moving up the D axis until the observed deficiency number is reached. The achieved seafidence for %=0.05 ces them be interpolated from the surves. This nakas quite visible the small confidence generated only by the data and reveals the need for exceptionally large samples la order to achieve high confia s
dance levels.
The leyesian procedure has the affect of changing the starting origin.
B 2.11
- t.
ELEVENTH ANNUAL NATIONAL ENERGY DMS40N CONFERENCE
~
p%.
- i s
1
[ %,s-By introducing the pseudo-data, one is essentially starting at 5'=74 and D'=e-1=1 with a 901 confidence. Plots of the data from this origin are shoum.
h second major issue was to formulate a reasonable method for interprettag the resulting posterior values in order to define decision and actica guidelines. For illustrative purposes of the 711st study.
the confidence curves were interpreted as accept-reject lir.as. N philesy; we to reflect a desire for an 853 plant. Thus, say procedura pleuing above the $53 curve was deemed as rejectable and necessitattag some sort of corrective action or tavastigation. Any plot beyond the 951 curve was deemed as acceptable and perhaps worthy of reduced observation.
Betvean the,853 and 953 curve, the peocasses could be viewed as adequata, but still requiring nomitoring.
Figare 2 shoue an alternative plotting ta which curves of the Poisses parameter are provided at a confidence of SSI. Suah curves any be use-fully employed if one desires to set acceptable and unacceptable levels of 1.
h interpretattaa of Figure 2 is analogous to that of Figure 1.
Figare 3 shows a sta11ar plotting for the. case in which one desires to
- -control-tha number-of-. deficiencies. ;.h: curves;repressat_tha: total._ _.-..
2 m
deficiencies in a lot of 24 at a confidence of 953. h interpretation of these plottians for actions to be' takes is also dependent on what is viewed as acceptable.
CoucLyszcu$.
h development of the Bayesian confidence intervals suggested the
, possibility of solving the small sample probles encountered La assuring the quality of FM activities. h pilot study ha's' demonstrated that it is possible to determine realistic and defensible prior' distributions. In addition, the pilot study demonstrated that the Bayesian confidence inter-vals can be implemented in staple to use graphical techniques which can k employed by inspectors and monitors. More importantly, the results seen interpretable as actise statemaats for future control.
h results of the pilot study ara encourastag. It is felt that these results held the promise of significant utilization la assuring the quality of FM activities as well as other infrequent but routine plant activities. Novaver, it is also felt that a significant amount of learning is still secessary. Procedures for selecting priors must be formalised.
and perhaps standardized if the results are to be acceptable to regulatory agensias. Appropriate guidance for selecting appropriate processes, quality levels and acceptable confidence levels also must be developed. h as efforts are to be the subject of the future development sagdies for
- implementing Bayesian et.nfidence intervals.
Ia -ry, the Bayesian methods discussed appear to be a valid 8 2.12
ELIVENTH ANNUAs. NATIONAL ENERGY DIVISION CONFERENCE
~
i i.
a...
statistical assas for anciaar % staffs to develop appropriate confidence statements for management on the achievement of quality La FM and othat sativities. Emelear plant activities are typically of a nature that they tavolve freguest process changes and variations. They occur infrequently la comparison to what saa he viewed as a neraal statistical 1st sise. The amount of different processes occurring significantly limits the amount of sampling capability that a typical-sised nuclear % staif saa achieve. Non-removable variations from activity to activity due to,
personnel performance, materials sad plant equipment are commes The Emyestas approach allows % to properly assess these factors and to develop confidence stacaments that can be used by managensat to assess the quality of the activities boiss conducted. The techniques discussed, preparly taplemented by trained % Engineers and Quality Control personnel, can produce valuable statistical data. These data should enable % to achtsve a more " graded approach" la the application of % requirements and a better utilization of its resources.
y t'.
r.
e B 2.13
I-f,..,,,
ELEVENTH ANNUAL NATIONAL ENERGY DIVISION CONFERENCE
. p*c} "
\\
1 APPENDIX A f
This Appendix provides the equations for obtaining classical confidence levels for set. intervals for sach of the parameters discussed above.. Only upper interval forms are presented in the interest of brevity.
Bernoulli Parameter Estimates The classical confidence level for an upper interval estimate of the Bernoulli parameter is given by the*following equation:
E
~E}
~}
Cu(p;D,S) = 1 -
dl(S )!
d=0 Note that when D=S, then an upper interval cannot be established (1,s.,
the confidence is sero for all upper intervals).
Bernoulli Defectives Estimates _
s The classical confidence level for an upper interval estimate of the number of defects present in a lot (based on the Bernoulli process) is given by the following equation:
D St Bl(L-B)I (L-S)!
E Cu(B;p,S 1-) = 1 -
dl(S-d)!
Li (B-d) l (L-S-5+d) I g
Note that when D=S, then the confidence is sero for all upper intervals.
Also, when D=0, the upper interval estimate of B=0 is formed with a confidence of zero.
Poisson Parameter Estimates The classical confidence level for an upper interval estimate of the Poisson parameter is given by the following equation:
Cu(1;D.S) = 1 -
(AS)d,- M (EQ A-3) d=0 DI Poisson Deficiencies Estimates The classical confidence level for an upper interval estimate of the number of deficiencies present in a lot (based on the Poisson process) is given by the following equation:
B 2.14
(
ELEVENTH ANNUAL NATIONAL ENERGY DMS40N CONFERENCE 1
1
\\.
- i gggh), - (h (1-h -4 (M p) e 5
a ca(Esa,s.L) = 1 - E d=0 Este that the upper interval estimate of B=D is always at a confidesca of sere.
i AR EDlI_5 This Appendia presents the equations for obtaining the Bayesian probability (confidesca) isvols for set intervals for each of the parameters discussed above. only upper interval forms are presented in the interest of brevity.
Bernom111 Parameter Estimates N Reyesian probabt?,ity for an upper interval estimate of the Bernoulli parameter can b given by the following equation:
pk1M (M 5-1)
~
pu(sia,8;D,5) = 1' -
ggg g
),
vben a is sa integer. For non-integer values of s, an integral form is necessary.
It is worthwhila nottag that ta this form,.the Bayesian probability is equivalaat to the classical confidence.svaluated with the" pseudo-data added to the ke,1 data, i.e..,
(EQ B-2)
Pu(p;s,8;D,5) = Cu(pge+D-1, a+9+S-1)
Thus, for a=1 and p=0, the two are equal. Otharvise, a and $ have the role of "psende" sampling resalts of prior superience which are added to the rea.1 samplings. For assiganaat purposes, a-1 any M thought of as the number of pseudo-defeats and e&S-1 as the number of pseudo-samples.
Bernoulli Defectives Esriantes' The Bayestaa probability for sa upper incarval estimate of the number of defects present ta a lot (based on the Bar em process) is given by the following equation:
5 2.15
~
b-ELEVENTH ANNUAL NATIONAL ENERGY DIVISION CONFERENCE
. 2 A..
l oj c.
og N. I
.s a
%gr Pm(5;s,8sD,s.L)
(L s)t (e+s-1+s)t(e*D-1+b)I (s-1+L-D-b)t (34 5-3) be(L-s-b)I (e*6-1+L)8(e-L+D)I (5 1+S-0)I g
This relation holds for"ase-integer values of.e and 8 when the factorials are replaced by Games functions.
)
Poi-m Parameter Estimates The Bayesian probability for an upper interval estiaste of the para-
)
assar saa be given by the following equations e2-1 4
III*IIII e"II* IIA
@ M)
Pa(1gs,83D,5) = 1 o E
di g
whom a is'as,iueger. For non-tateger values of.s, am integral fora
- I*
It is noted that la this form, the Bayesian upper probability is equivalent to the classical upper confidence evaluatian with the pseudo-data added,to the real data, i.e.,
Pu(1s,8;D,5)=cadsa+D-1,9&S)
(BQ 5-5)
Thus, for o=1 and 8=0, the two are equal. OEttervise, e and 8 have the role of " pseudo" sampling results of prior experience which are added to the real samplings. For assignment purposes, a-1 may be thoughs of as the pueber of pseudo-dgfacts and 8 as the number of pseudo-samples.
Poisson Deficiencies Estimates The Bayestaa probability for sa apper interval estimate of the number of deficiencies present in the cummulative lets (based en the Poisses process) is given by the following equatima a+D 5-D h
Pu(5:n,8;D,8,L) = (
)
E
(
)
(EQ 5-6) pg g
b=0 when a is an integer. Per aco-integer values of e, the factorials are t
replaced by Gasmia functions.
Eere there is as obvious direct relationship with the classtaal cam-J 8 2014
t ELEVENTH ANNUAL NATIONAL ENERGY DMSION CONFERENCE 4
l fidassa laval. Eowever, the prior paramatars a and 8 m y still be thought
'-ta and the probability for the interval behaves analogously of as p to the classisal confidenes evaluated for the pseudo-data added to the real data.
e e
9 e
e a
e o
S 9
9 e
0 e
e 4
9 alt e
e q
O e
4 4
9 0
e e e
e
- O O
4 s
e 8 2.17
ELEVENTH ANNUAL NATIONAL ENERGY DMSION CONFERENCE
,..jAq
,ft l
l f
c I.v.
J.
I
%l
^j N;),,_p 313L10CIAPRT l
l (1) Tanaan. T., seme<=e4en. Am tatroduerary Analveis. 2nd Editism, Earper and Esw. New Terk,1964.
(2) Mama, M. R., et al., Methods for Statistie=1 Analysis of Baliability and Life Deta. John Wiley and Seas New York, 1974.
l (3) Seklaifer, R., probability and Statistics for Business ' Decisions.
Macraw E111. New York, 1959.
(4) Ces111a, G. L. and A. M. Smith, "Ra11 ability Measurement and confirmatias;" p.,93 la 84=k **=k Safety Technoloav, edited by 1
A. E. Green, Jehm Miley and Seas. New York, 1982.
(5) Crellin, c. L. (editor), "Special Isame o's Bayesian Ballability Tashaiques," IEEE Tr-----ti== on Ea11 ability. August,1972.
~~
- (6) GPU Nuclear Corporattaa, " Operations 1 Qaality Assurance-Plaa "
r
- p. 87. September, 1982.
I 1
)
e b
's Y*
o9
- ,g e
e O
D
'6
,e e
e e
O e
i -
s 2.is
V ELEVENTH ANNUAL NATIONAL ENERGY DMS40N CONFERENCE
/
4 r
s h
l.
e I&EluE 1 m _=ossunnomnu Frequency of Emmber of Number of Ossurressa observattens Deficiencisa Procedure (2 Taar Period)
(3)
(D)
L 730 10 0
2 730 36 2
3 730 6
2 4
104 3
0
'S 104 7
1 4
0 6
24 7
to 4
'1 S
S 6
0' 10 8
7 2
e 9
e e
e e
9 6
OO g
e e
e 9
S O
OS
- .*e 9
e e
e 4
ti 2.19
~-.V e
q l
l 4
I L
n d.
- 3
=
+
.f S.
s t
M M
m l
e m
e l
8 e
e a
+
M
. g l l#,
=
m 8
M e
e
.=
. p o
- g M M M MM MMM M M n
M M
M
. M W
e W
m
+
" g.W g
m 8
99 o
i g
M M
.L
=
W W
e g
e W
8
. g.
4 g
K
=
g
=
O W
e.n.
t 4
~
l 1
l l
l 4
4 q
w
'O+ T-w m,4
~
I
$=
I I
t.
s t :-
a 4
4:
- 6
+.
1 j
,.a 1
b '
t
."g 1
em
~
~
m%
e s.
.n s
mg ir,
g M
~
+
O
~
=== ---- ~ ~.
s g
to yu2 3
I!!
,u
.t-g t
- to e
p
~
et
~
g o
e n_
g g
I l
_1 l
I 4
as G + T - 3s s,0 B 2021
c r
e O
a i
i i
I i
M
'4
- s
~
W W
W e
e 4
.:: 3 h.
W
" g W
M e
M s * *S n*
ne
~
e h
4 O I
~
- g
_M_
m 9
W m
'O
~
n
=
~
. og g --
N be o
2 ana 9
- 3
=
e a
~
UR
.s.
8 tt 39 s
@ +
h 9 g~,.
80 g
O i
e e w
M t
g 3
HB l
l l-l C+T-me,4
_^n e..
D; t
,, +..
3.
i-
\\'
s 5.
-ENCLOSURE 8 s
BRUNSWICK STLTi ELECTRIC PLANT, UNITS.1 AND 2 NRC DOCKETS 50-325 6.50-324
.c OPERATING LICENSES DPR-71 & DPR-62 REQUEST FOR LICENSE AMENDMENT ROSEMOUNT ANALOG TRIP SYSTEMS ANAIDG TRANSMITTER / TRIP UNIT SYSTEM
-FOR ENGINEERED SAFEGURAD SENSOR TRIP INPUT i
i.
1 l
.i
~
i
______________._____w
ENCLOSURE 8 BRUNSWICK STEAM ELECTRIC PLANT, UNITS 1 AND 2 NRC DOCKETS 50-325 & 50-324 OPERATING LICENSES DPR-71 & DPR-62 SUPPLEMENT TO REQUEST FOR LICENSE AMENDMENT ROSEMOUNT ANALOG TRIP SYSTEMS ANALOG TRANSMITTER / TRIP UNIT SYSTEM FOR ENGINEERED SAFEGUARD SENSOR TRIP INPUT
a j
i i
I i
1 LICENSING TOPIC L bPORT-(NEDO-21617-A).-
"ANAIDG TRANSMITTER / TRIP UNIT SYSTEM FOR ENGINEERED SAFEGUARD SENSOR TRIP INPUIS"
- 1 i
9
]
i k
0 9
e 4
9 i
e
==
3
- _ _ _ _ __;_ _ - - ~
i, NEDO-21617-A j'
_c ir s s_,,L,,
, Ecembs D 978 )
O 02 Q,/~
g 0
/
LICENSING TOPICAL REFORT i
g J
'N j
l
\\
/
J 81 12 01
/,y s-ANALOG TRANSMITTER / TRIP UNIT SYSTEM FOR ENGINEER'ED SAFEGUARD SENSOR TRIP INPUTS
\\
s.s s
N p.
'N
/
4 R. W. Strong L. H. Youngborg-i l
P0H" ihrusuinu;.) 'ON.1 n
Approved:
D. G. Scapini, Manager Control and Electrical Engineering l
l 1
l NUCLEAR ENERGY PROJECTS DivissON e GENERAL ELECTRIC COMPANYl SAN JOSE CALIFORN4A96125 j
(G EN ER AL kp*LECTRIC e
m f /) N M d %'
,1
\\
s
_ j
. N~
J' I
]
DISCI. AIMER OF RESPONS188UTY TNs document was propewd by or for the General Coctric Company. Neither General Sectnc Company nor any d the conMbutore to this document:
Makes any warranty or representadon, expraes or knpNed, wilh respect to the accuracy, completeness, orusetukene d the Normadan contained kr Na docu-A.
ment, or that the use d any Idormadon discheed M this docuunt may not inMnge privately owned rights; or
~
- s. Assumen'eny moponetwary tse nabiny or damage d any und which may roe korn the use d any Normat6on diecioned M thie document.
)
4 s a G
4 o
e
-s
f NEDo-21617-A-
~
q TABLE OF CX)NTENTS E!1*.
U.S. NRC LETIER OF ACCEPTANCE AND REPORT EVALUATION xi ABSTRACT 1-1 15 INTRODUCTION 1-1
1.1 Background
1-1 1.1.1 Justification for Design Change 1-2 1.2 General Design Requirements' 1-2 1.2.1 Applicable Documents, Codes and Standards 1-3 1.2.2 General Requirements 1-3
1.3 System Description
1-3 1.3.1 General 1-4 1.3.2 Master Trip Unit 1-4 1.3.3 Slave Trip Unit 1-6 1.3.4 Trip Relays 1-6 1.3.5 Power Supply 1-6 1.3.6 Calibration Hardware 1-7 1.3.7, Card File 1-8 1.3.8 Bench Test Facility 1-4 1.4 Application 2-1 2.
SYSTEM COMPONENTS 2-1 2.1 Transmitters 2-2 2.2 Trip Units and Calibration Hardware 2-2 2.3, Power Supply 2-2 2.3.1 115 Vac to 24 Vdc Ferroresonant 2-7 2.3.2 125 Vdc to 115 Vac Inverter 2-12 2.4 Trip Relay 2-14 2.5 Component Acceptance 2-15 2.5.1 Transmitters 2-15 2.5.2 Master Trip Unit 2-17 2.5.3 Slave Trip Unit 2-47 0
1' 5.4 Card F11e 2-17 2.5.5 Calibration Unit 111
~..
NEDD-21617-A s
j TABLE OF CONIENTS (. Continued)
!. age, t
2-18 2.5.6 Randout Assembly 2-18 2.5.7 115 Vac to 24 Vdc Power Supply.
2 2.5.8 125 Vdc to 115 vac Inverter 2-20 2.5.9 Trip Relay
~
3-1 3.
SYSTEM DESIGN 3-1 3.1 Component Interconnection 3-1 3.1.1 Emergency Core Cooling System Applications 3-6 Reactor Frotection System Applications 3.1.2 3-7 3.1.3 Card File Arrangement 3-7 3.2 Trouble Monitors 3-7 3.2.1 Card out of File 3,-14 3.2.2 Gross Failure Low 3-14
'3-15
)..
3.2.3 Gross Failure High 3.2.4 Card in Calibration 3-15 3.3 Farformance 3-15 3.3.1 Accuracy and Repeatability 3-17 3.3.2 Response Time 3-20 3.4 Component Availability 3-20 3,4.1 Availability Criteria 3-21 Mean-Time-Between-Failure (MTBF) Analysis 3.4.2 3-23 Failure Mode and Effects Analysis (FMEA) 3.4.3 3-24 3.4.4 Conclusion of System Availability Analysis 4'- 1 4.
QUALIFICATION TESTS 4-2 4.1 Environment 4-2 4.2 Seismic and Fragility 4-2
- 4. 3 Power Supply Regulation 4-5 4.4 Electromagnetic Interference (EMI) 4-5 4.4.1 EMI Transients 4-5 4.4.2 Radio Frequency EMI 4-6 4.5 Margin i
iv
- - ~
v l
NED0-21617-A q
i TABLE OF CONTENTS (Continued) fa.g,e, 5-1 5.
APPLICATION 5-1 5.1 Operating Power Plants (Phase C)-
5-1.
5.1.1 Trip Unit Cabinet Assembly 5.1.2 Trip Unit. Cabinet Design for Reactor 54 Protection System (RPS) 5.1.3 Trip Unit -Cabinet Design for Emergency
$-4 Core Cooling System (ECCS) 5.1.4 Location and. Interconnection of Trip 5-4 Unit Cabinet and Existing Cabinets 5-9 5.2 Early Dasign Plants (Phase B) 5-9 5.3 Late Design Plants (Phase A)
J 5.4 Interfaces 5.-9 5.4.1 Specific Instrument Loops 5-9 5.4.2 Trip Cabinet j
.,5-10 I
5.4.3 Environmental Interface 5-10 I
5.4.4 Specific Plant Interconnection.
1 5-10
,j 5.4.5 Pield Calibration Rack I
6-1 SITE CALIBRATION AND SURVEILLANCE TESTING TECHNIQUES 6.
6-1 6.1 Transmitter Calibration Prequency of Calibration Test Interval 6-1 6.1.1 j
~
Calibration Procedure 6-3 6-1 6.1.2 6.2 Trip Unit Calibration 6-3 6.2.1 Punctional Surveillance Testing 6-5 6.2.2 Setpoint Calibration 7-1 7.
CRITERION CONPORMANCE 7-1*
7.1 IEEE Scandards 7-1 7.1.1 IEEE-279-1971
.7-1 7.1.2 IEEE-308-1971 7-2 7.1.3 IEEE-323-1974 7-2 7.1.4 IEEE-336-19'i V-Y l
i mm
NEDO-21617-A TABLE OF QNTENTS (Continued)
'i f.agg 7-2 7.1.5 IEEE-338-1971 7-3 7.1.6 IEEE-344-1971 7-3 7.1.7 IEEE-379-1972 7-3 7.1.8 IEEE-384-1974 7-4 7.2 NRC Regulatory Guides 7-4 7.2.1 Regulatory Guide 1.22 7-4 7.2.2 Regulatory Guide 1.29
' 7-4 7.2.3 Regulatory Guide 1.30
~
7-4 7.2.4 Regulatory Guide 1.32 7-5 7.2.5 Regulatory Guide 1.47 7-5 7.2.6 Regulatory Guide 1.53 7-5 7.2.7 Regulatory Guide 1.62 7-5 7.2.8 Regulatory Guide 1.75 7-5 m
7.2.9 Regulatory Guide 1.89
- ),1 7-6 7.3 other Documents l
7.3.1 Title 10, Chapter 1, Code os Federal Regulations, 7-6 l
Part 50 Appendix A l
8-1
\\
8.
CONCLUSION 9-1 9.
REFERENCE A-1 APPENDIX A - RESPONSES TO REQUESTS FOR ADDITTONAL INFORMATION t
l vi i
i i
m i
LIST OF ILLUSTRATIONS Title pg Figure 1-5 1-1 Typica'l Trip Unit / Calibration System Elementary 1-8 1-2 Typical Trip Unic/ Calibration System File ~
2-1 2-1 Typical Pressure Transmitter 2-3 2-2 Typical Schematic of 24 Vdc Power Supply 4 2-3 24 Vdc Power Supply 2-7 2-4 DC Power Supply Regulation 2-8 125 Vdc to 115 Vac Taverter (CE Special) 2-5 2-9 2-6 Typical Schematic of de to ac Inverter 2-13 '
2-7 24 Vdc Trip Relay 3-2 3-1 Typical ECCS Connection Diagram 3-2
{
3-2 Master Trip Unit Functional Block Diagram.
3-3 Sample Combination Arrangement of Master and 3-4
)
Slave Trip Units 4
J 3-5 3-4 Use of Trip Units as, Parallel Drivers 3-5 3-5 Trip Unit. Driving Parallel Relay Coils 3-6 3-6 Typical RPS Connection Diagram 3-7 Total Repeatability Computation for 3-17 Master Trip Unit Output 3-8 Total Repeatability Computation fo'r 3-17 Slave Trip Unit Output 3-18 3-9' Time Delay Response Contribution of Each Component f
3-10 Fast Scram Sensor and Logic Time Delay 3-19 Compared to Specification 3-11 Safety Valve Sizing Pree'sure Transient Envelope 3-19 for BWR/6 3-26.
3-12 Plot Locus of Equation 3-2 4-4 4-1*
Voltage Regulation Conservatism of Power Chain 5-1 5-1 Standard Retrofit Panel 5-2 Perpendicular Separation Method for Divisional 5-3 and Nondivisional Wire 5-6
]
Trip Unit Cabinet and Cable Arrangement (Local Installation) 5-3 Interconnection Between New Transmitter / Trip Unit 5-7 5-4 and Existing Wiring (Local Installation)
O vkk
- i
- - _ _ _ - _. _ _-.;. a,, u.c
~
NED0-21617-A' LIST OF ILLUSTRATIONS (Continued)
Title _
g Figure Interconnection Between New Transmitter / Trip Unit 5-8 5-5 and Existing Equipment (Control Room Installation) 6-2 6-1 Transmitter Calibration Setup 6-3 6-2 Dif ferential Pressure Transmitter 6-3 6-3 Pressure Transmitter
~
9 P
e a
e
..)
s l
~ viii U
Q@ p#p
,_ gy A
NEDO-21617-A n' -
LIST OF TABLES Title g
Table 2-11 2-1 Inverter Specifications 3-8 BWR/6 ECCS Trip Unit Card File Assignments 3-1 3-13 BWR/6 RPS Trip Unit Card File Assignments 3-2 3-16 Trip System Repeatability Error Specifications
- 3-3 3-18 Total Analog-Trip Unit Relay Tandem Response Time 3-4 3-22 3-5 Failure Rate and MTBF Summary Failure Rates (1)'as Computed per MIL-HDBK-2178 for 3-24 3-6 35'c Temperature High Stress Requirements for the Power Supply 4-1 4-1 l
and Trip Relay Qualification Susmary Showing Maximum Environmental 4-3 r
4-2 Abnormal Exposure Conditions (Spec vs Test) f Qualification Summary Showing Maximum Seismic and
.4-4 4-3 fragility Exposure Conditions (Spec vs Test) 4-7 Highest Successful EMI Test Parameters 4-4 4-9 Margin Test Scoreboard 4-5 5-2 J
q'.
Recommended Retrofit Application 5-1 7-1 IEEE Standard 279-1971 (. Criteria for Protection 7-7 l
Systems for Nuclear Power Generating Stations) 5
(
i
()
ix/x m.
^,
)
ABSTRACT This licensing topical report describes the Analog Transmitter /
Trip Unit System for Engineered Safeguard Sensor Trip inputs.
The system essentially replaces pressure, tevet and temperature j
switches with analog tm nsmitter/ trip unit ocmbinations, which provide continual monitoring of critiaat pammeters in addition to perfoming basic toda trip opemtions.
The principat ob.ieo-l tive of the system is to improve sensor intstligence and reti-e.
t
- ability white greatly enhancing testing pwoodures.
posign criteria and justification of application are included, as vett as per'formance and essential qualification data for att System imptamentation is designed to be suitable new harduare.
for att BWR power plants and is individually treated for openting plants, plants in late design phase,' and plants in early design phase.
I i
s t
i
('
xi/zii t
v-tet,DU-c:101 ( -n
/* C' tty **o UNITED STATLS O
- 7.,
NUCLE AR REcULATORY COMMisslON i
wasw N<pos.o e sous
, {
."A, "!
JUN 2 71978 MFN 2"9-78 General Electric Company
~
ATTN:
Dr. G. G. Sherwood.. Manager Safety and Licensing 175 Curtner Avenue San Jose. California 95114 Gentlemen:
REVIEW OF GENERAL ELECTRIC TOPICAL REPORT NEDO TRANSMITTER / TRIP UNIT SYSTEM FOR ENGINEERED SA
SUBJECT:
TRIP INPUT" We have completed our review of the subject topical report and conclude that it is acceptable for reference in license applications as specified in the enclosed evaluation.
The staff does not intend to r'epeat its review of this topical report Q
when it appears as a reference in specific license applications, except to assure that the report is applicable to the specific plants involved.
Should regulatory criteria or regulations change such that our conclusions concerning this topical report are invalidated, you will be notified and will be given the opportunity to revise and resubmit your topical report for review, should you so desire.
In accordance with established procedure, we request that General Electric issue a revised version of NE00-21617 to include any supplementary infor-mation provided for our. review of the topical report, this acceptance letter, and the enclosed evaluation.
Sincerely, 4
Olan D. Parr, Chief Light Water Reactors Branch No. 3 Division of Project Management
Enclosure:
Topical Report Evaluation cc: See page 2 Q
I.
u
r.
NEDO-21617-A-
!'l d-T General Electric Company
/i cc w/ enclosure:
Mr. L. Gifford General. Electric Cogany l
4720 Montgomery Lane Bethesda, Maryland 20014 O
N e
e J
O e
4 9
h I
t B
~
i 4
' ~ ~ ~ ^ ~ ~ - -
w
m NEDO-21617-A O
T0pICAL rep 0RT EVALUATION NEDO-21617 Report No:
Analog Transmitter / Trip Unit System for Report
Title:
Engineered Safeguard Sensor Trip Input April 1977, Amendment No. 1 - January 1978 Report Date:
l General Electric Company-Originating Organization:
l Instrumentation and Control Systems Branch.
Reviewed by:
Division of Systems Safety 4
1.
Sunnary of Topical Report The topical report provides a description of General Electric's proposed equipment to be utilized (i.e., possible replacement in I
older plants and planned implementation in new plants) for certain f
,...s l
V Engineering Safeguard sensor trip inputs including inputs for the' i
1 reactor pmtection system, emergency core cooling system, and nuclear steam supply shut off system.
i The proposed equipment can replace pressure, level, or temperature This switches with analog transmitter / trip unit combinations.
equipment then pmvides the 6quiva. lent logic trip inputs of th'a 1
1 switches while providing additional features such as:
i I
j a) continuous monitoring of parameters 1
b) improved testing procedures and capabilities l
c) improved operational characteristics such as fewer instrument testing related scrams and fewer instrtsnent valving errors.
Q t
I l
Jy
.. /j l
The report outlines the design bases and cMteria for the equipment with the emphasis on the trip unit.
The analog input portion can be implemented with any qualified _ process transmitter that can
- satisfy the interface requirements imposed by the GE trip units.
Thus the analog transmitter portion was not included as part of The trip unit is composed of electronics for the topical report.
conversion of a 4-20 milliasp transmitter signal input to the requimd '
logic signal necessary for the various logic systems.. The trip unit includes means for trip point adjustment, calibration.- failure indication, and periodic testing.
l The tHp unit system is designed to be nuitable for all' Bl6L power
.)
~
The top'ical report provides the implementation for three plants.
. ' distinct category of plants:
- 1) operating plants'.
- 2) plants in late design phase, and
- 3) plants in early design phases.
Included in the report are the specific plants of each category as presently detemined.
2.
Sumary of Regulatory Evaluation The General Electric Company has investigated design improvements for the safety system instrumentation of Bidts based on reports from L
i e
1
. O This led to proposed modifications for operating operating reactors.
plants to some instrumentation for the reactor protection system (RPS), nuclear steam supply shutoff system (NSSSS), and emergency l
core cooling (ECCS). The instrumentation can provide for the moni-toring of various plant pressure, level, and temperature parameters.
We have reviewed the topical report with emphasis on the design bases and criteria of the proposed trip unit system as compared f
to the existing sensor (switch) applications.
l GE determined that the proposed modifications are a design improve-8 ment to be supplied on all future BWR/6 product line plants and f
'i the output of the trip unit can be directly compatible with the
(,..
o Therefore, this t
I solid state logic of the GESSAR design plants.
new equiment is presently an integrel part of the GESSAR preliminary i
design as approved by the staff for a pDA.
As noted by GE, there are' major advantages of this system over the t,
The advantage with respect to regulatory require-previous method.
1 l
ments is in the area of input testing.
Based on these improvements, l
l-we concluded that the proposed system could be an. acceptable design-Our evaluation improvement and, therefore, proceeded with our review.
concentrated on the ability of the new equipment to continue to h
satisfy the criteria for the existing equipment.
In addition, we reviewed the system with respect to current criteria because of the proposed utilization in current and future designs.
\\
Q f
l
--.....e...v.....
.m
'NEDO-21617-A 4-
]
)
The trip unit (with analog transmitter) and trip relays (for non-solid state plants) provide the input intelligence to system logics
{
for the RPS. ECCS, and NSSSS. These systems are composed of various
)
/
independent divisions of equipment.
The trip unit and associated trip relay equipment are implemented on a divisional basis (i.e..
a trip unit through the trip relay connunicates to only one division)
For new solid state plants, the trip units 1
in the older plants.
are required to provide outputs to all four divisions of the solid
{
state two-out-of-four logic for the reactor protection system.
The divisional interconnection is implemented through' isolation '
devices which are outside the scope of this report and which have been reviewed for the PDA of GESSAR.
.b, During the review, the staff expressed concern that the trip unit system may compmmise the independence requirements of certain plant It was detemined that for operating and current plants functions.
f each individual application shall be evaluated in accordance'with the specific divisional separation requirements of that plant as outlined in the interface requirements.
(See Section 3 of this report.)
One area 3f concern for the GESSAR plants is the requirements for The design the new safety relief valve actuation system logic.
basis for the system includes the requirement that no single failure 31 m
ns.w-c m. -o n This would require that certain shall cause actuation of the system.
coincident signals within the divisions be independent. ~ This infor-mation can be ascertained only when the final design details are i
available, and thus will be reviewed during the review of the GESSAR FDA. or the FSAR of plants referencing GESSAR.
f Another area of concern is equipment qualification; i.e., _how well k
the equipment can perfors.its required function under all plant
~
conditions.
In all cases the only conponent associated with the proposed'new j
l equipment located inside the reactor building (but outside' primary For the BWR 4 plants the trip l
containment) are the transmitters.
All other
,,)
units are also located inside the reactor building.
'~
e equipment will be located in control room like environments. There-fore, GE qualified the trip units 1or the more stringent BnR 4 reactor E
In a similar manner, GE qualified all of building requirements.
q l}~
the equipment (seismically and environmentally) to the most severe conditions expected for any application. We require an applicant 8
to demonstrate that, for his specific application, the equipment is located in an area with se,ismic and envirorsnental conditions i
within tha specified test envelope.
GE has stated that qualification data for the transmitters and trip relays are already in existence, since these. devices are being used i
e **9, 2
' */ P'M M@gS A O"WI e
~
) in operating plants, and that the applications.of these devices as proposed in the topical report does not deviate from current applications.
For GESSAR plants, the qualification of the equipment to the require-ments of IEEE 323-1974.is part of the generic review of the qualift-Therefore, cation program for all Class IE equipment in GESSAR scope.
the acceptability of this equipment for newer plants will be based on the outcome of this generic review.
For the previous sensor inputs implemented through switches at the I
process interface, periodic testing involved valving out the sensor.
This
)
applying a test signal, and then checking the switch response.
method of testing had disadvantages as identified by GE in the With the new design, the periodic testing require-topical report.
ments of the instrunent channel, can be changed to be similar to
.present technical specification requirements for pressurized water plants.
Specifically, a " channel check" could be perfomed at the meter on each divisional output, a " channel functional test" could be perfomed with a test signal input at the trip unit, and a " channel calibration" could be perfomed at the transmitter at major plant outages.
1he The testing advantages for this equipment are significant.
only damerit in this method is that the transmitter itself is not Y
L;
NE00-21617-A f.
i tested as frequently.
However, the likelihood of an undetected failure of the transmitter is minimized by the ability to compare the output of one transmitter to the other identical divisional i
transmitter output at any time.
I 3.
Regulatory Position _
We required General Electric to provide interface criteria which address the required supporting infonnation for implementation of the equipment described in the topical report on a particular plant GE docunented tn interface requirements section which t
application.
addresses. the following areas:
\\
{}
- 1) Transmitter vendor and model number to be utilized 2)
Divisional separation assignment
- 3) Cabinet layouts
)
- 4) Environmental and seismic qualification 5)
Interconnection diagrams We shall require that each applicarrt, referencing this topical report.
address the interface requirements of the report.
For GESSAR plants, the specific area of circuit independence discussed
\\
in Section 2 above will be reviewed as part of the FDA review of
~
O I
- 0 m
. /
Also, it GESSAR or the FSAR review of plants referencing GESSAR.
is our intention that the review of equipment qualification to will be part of the generic rtview of GESSAR equip-IEEE 323-1974 4
~
ment qualification.
Based on our review of NED0-21617 as sunnarized in Section 2 a subject to the conditions concerning interface requirements and GESSAR plants noted above, we conclude that the subject topical mport is an acceptable 'mference.
4 e
1 e
9 e
d
)'
v NEDO-21617-A 1.
INTRODUCTION
~'
1.1 BACKGROUND
Reports from operating reactors have led to an investigation of potential design improvements for the safety system instrumentation of General Electric It has been observed that direct pressure and boiling water reactors.
iiifferential pressure actuated switches, which provide process parameter out-of-limits trip into. safety systems, are subject to setpoint drif t.
Additionally, technical specification requirements on this instrumentation have led to a high rate of surveillance testing which requires a large amount of technician time to ' accomplish.
Such testing of some instrumentation requires the plant to be placmd in a half-scram condition when the instrument
^
is out of service for calibration or test. The General Electric Company, therefore, has proposed changes in the instrumentation for tha rgactor' -
protection system (RPS), nuclear steam supply shutoff system (Ng ), and emergency core cooling systems (ECCS), for power plants which were in the design phase and which could have the circuits changed without adversely affecting hardware delivery or fuel loading-schedules.
Similar changes r.re being proposed to owners of operating plants, suggesting that they incorporate similar modifica-tions to their safety system instrumentation.
It has been shown by some owners of operating plants that the modification is cost justified because of anticipated improvements in the plants' availability and the simplified calibra-tion procedures.
1.1.1 Justification for Design Change The proposed protection system changes are designed to:
reduce the time the RPS logic must be in a half-scram condition to a.
functionally test or calibrate a safety trip; b.
reduce the functional test of calibration frequency for the primary sensor from once per month to once per operating cycle for multi-channel variables, to allow calibration of the primary sensor to be performed when the reactor is shut down for refueling; 1-1
's NED0-21617-A eliminate likelihood of instrument valving errors; c.
virtually eliminate instrument testing related scrams; d.-
i reduce undetected primary sensor element drift due to the utilization e.
of a meter in each primary sensor signal' loop; decrease the amount of time required. to functionally, test or cali-f.
brate the safety trip point from approximately one hour to a matter-of seconds; eliminate the need for Iydraulic anubbers at the input to the primary 3
sensor; significantly reduce the number of abnorami occurrence reports' h.
filed with the NRC for setpoint drift; allow the location of the trip units in an area where the background -
i.-
This radiation will remain at a low level for the life of the plant.
will reduce the exposure to the instrument mechanics because -sons sensors that need monthly ' calibration are in moderate radiation areas.
I'. 2 GENERAL DESIGN REQUIREMENTS Applicable Documents, codes and Standards 1.2.1 The following codes were applied as part of the qualification of this equipment:
General Guide for Qualifying Class 1 Electric Equip-IEEE-323-1974 -
a.
ment for Nuclear Power Generating Stations.
i Recommended Practices for the Seismic Qualification-b.'
{
of Class 1 Electric Equipment for Nuclear Power Generating Stations.
y' a
e 1-2
NEDO-21617-A-
- 7 1.2.2 General Requirements The trip unit must provide a process trip that does noit. vary with time due to instrument drife.
The electronic trip. unit must be testable in place. This greatly reduces time required to perform a surveillanen ' test which minimizes the time the sensor trip is in the inoperative state.
The. trip unit shall incorporate a readout for each sensing loop that continuously ~
In addi.
displays the prameter b'eing monitored during operation of the plant.
. tion, the test' current 'shall 'be displayed numerically with sufficient resolution-to achieve setpoint accuracy ' requirements impesed by system specifiestions.
Inputs to operator-interface displays shall be provided to indicate s'tatus of trip systems for such actions as ic,Cac " trip", transmitter current loop failure,.
card out of file, power loss, and trip. unit in "Inop" status.
()'
The trip unit calibration system shall be capable of introducing a positive or negative step of variable amplitude which is used to perforn response time testing of the trip unit and downstream logic elements.
1.3 SYSTEM DESCRIPTION 1.3.1 General i
i The Trip Unit / Calibration System is implemented as a technically superior I
8 approach to meet the safety system process trip input requirements of the BWR when compared to other methods of generating process trips (e.g., " blind" The Trip Unit / Calibration System is an all solid-state electronic sensors).
crip system designed to provide stable and accurate monitoring of process parameters.
The system consists of master trip assemblies, slave trip assemblies, calibra-1 The master trip unit tion units, card file assemblies, and other accessories.
C&.-
interfaces with a 4-20. milliampere (mA) transmitter or'a three-wire resistance t**perature detector (RTD), located at some remote location within the power 1-3
~~~~~~~q~~-**
- --_- - r ~ ; Y :L - ~ ~~_~ 3 _- - - ---- - - -
1
.l The slave trip unit is driven from the analog output of
')
I J
plant installation.
The calibration unit has the capability of providing
.j a master trip unit.
either a stable or transient calibration current that can be routed by a s
switch to any mastar trip unit. An elementary drawing depicting a typical application is shown in Figure 1-1.
Test jacks are provided on the master trip unit face for precision measurement A two-position logic invert switch internal to of actual parameter values.
each trip unit allows for the selection of either a high trip or low trip, i
thereby allowing the trip relays'to be either energized or de-energized The system requirements dictate the position ~ of the during normal operation.
logic invert switch.
1.3.2 Master Trio Unit The master trip unit is a plug-in printed wire assembly designed to accept a 4-20 mA signal from the remote transmitter or accept the input of a three-wire The Selection of input type is identified on the purchase part drawing.
RTD.
trip unit contains the circuitry necessary to condition these inputs and pro-
.The master vide the desired switching functions and analog output signals.
trip unit provides an output to energize a trip relay at any point.within the It also conta' ins an isolated panel 4-20 mA or resistance input signal range.
meter that displays the value of the measured parameters which'can be scaled in the units of the process variable.
1.3.3 Slave Trio Unic l
The slave trip units are used in conjunction with master trip units when it is The sisves desirable to have different setpoints from a, common transmitter.
Up to obtain their input from an analog output signal of the master trip unit.
thus allowing a seven slaves can be driven by a single master trip unit, I
Unlike possible eight differant setpoints from a singis measured carameter.
the master, there is no direct connection of the slave to a transmitter, nor i
However, each sieve has its
)
are any analog signals generated 'oy the slave.
own output logic switching function for either high or low trip which is These outputs may be T
independent of its master or other parallel slaves.
. used to ' supply " independent" sensor logic to any system or combina* ion of systems within the same engineered safeguard division.
1-4
\\
i
J
,4 m"p.#
n 5
w s'
/
I P
Tj R
T 2
a L
E V
t n
$I\\
E i i ts L
m 6
Uoo K
t r
m e
L R 9" y
E T
r V
E E
a V5 9 l
L t
A R
n L
E e
S 2 9I 1
T m
AW e
4 9 l
E a
x W
m j-R e
4 9 T
R os t
A o_
i E
t T2 9 s
ET 1
e t
t 1 RT y
g E
Ul U
V S
Se l
o e
4(
SS E
P i EN L
n IR s
R A.
PR n o4 T 0 0 K
o 1
1 i
T R
\\
t E
99l a
~
TS r
AA 8 9 l'
b I
i 6i l
n a
C
/
t in U
O p
i r
E o ois T
R i
E i
T -
l l
R T e
UI t l 0
SM Ui Y
a P
P I
A c
SS EN I
0 0 R
YL i
R1 T
AE p
RA.
R n o4 T LR P
3 E
y E
T R9 9 l
R E
T R
E U
R T
S EU S 8 9l S
ML A
E UI A
M R
L 1
n P
l F 6i A
L R
L F
1 E
E SW W
S Y
O O e
R P r
l GR u
i 8j D
i R
os R O.
g o_
E i
~
i E
T I O. E F
T R
I I T
1 I
F EN TO I
A R
R A i o3 T 99l nT 4
P R r eU T
R ll O E 8 9 U
I C AD TS CR A
A M
N 2 9 6
34 KK O
.0 Iil{
- ll; -
- 3
- 11. t I i
p
- 3"
. ~,
1.3.4 Trip Relays _
)
Each master or slave trip unit is capable of supplying trip relay loads up to
~
1 amp at nominal 24 Vdc.
Contacts from these relays provide the necessary variable input. The trip units are designed logic function for the procest with output diode "isolatic.," which allows parallel output connections of sev.eral trip units into one relay.
1.3.5 Power Supp1r The trip units are designed with individual power regulation circuits so that This allows the main power supply voltages need not be precisely regulated.
use of a highly reliable ferroresonant type power supply which is not likely to fail in such a way as to introduce a high voltage to the system.
This feature precludes catastrophic failure of all trip units on a single The power supplies are designed with built-in bus due to power supply failure.
diode " isolation" at the' output so they may be connected in parallel for load Power sharing and/or bumpless transfer, given a single power supply failure.
leads bypassing the diodes are also brought out for single unit applications or for individual unit voltage sensing when several power supplies are operated in parallel.
l l
1.3.6 Calibration Hardware 1.3.6.1 Calibration Unit The function of the calibration unit is to furnish the means by which an in-place The cali-calibration check of the master and slave trip units can be performed.
The normal use of brator contains both a stable and transient current source.
the stable current is for verification of the calibration point of any given The transient current source is used to provide a step current input channel.
into a selected channel such that the response time of that channel can be determined from the trip unit input to any point downstream in the logic to and including the final element.
1
\\
lt 1-6 f
1.3.6.2 Readout Assembly The readout assembly is a portable measurement and display device which is in-serted in the front of any calibration unit. It has two four-digit displays which track applied calibration currents for any trip unit within the card file, The lower as selected by a rotating switch on the front of the calibrator.
display always shows the test current as supplied by the calibrator.
The upper display does likewise except, when the selected trip unit changes state, the upper display is " frozen" to read the trip setpoint current.
The portability of this device yields two important advantages:
Only one readout assembly is required to calibrate all the trip units (1) of this design used in the power plant; however, it is General Electric's practice to furnish three readout assemblies for each plant. site to provide maximum availability of calibration hardware.
e The unit. is easily removed and calibrated against the laboratory bench (2)
Then the standardized readout as-()
standard to confirm its accuracy.
v sembly is used to calibrate all the trip units, thus assuring maximum precision and consistent process trip setpoints.
1.3.7 Card File _
The card file contains 13 slots,12 of which may be used*'for any combination of The thirteenth is a double-width master or slave trip units or blank fronts.
slot designed for the calibrator only. Figure 1-2 illustrates the complete card file containing typical sets of masters, slaves, blank fronts, calibrator and readout assembly.
i Each card file is furnished with its own calibration unit regardless of the The files are installed in standard' quantity of trip units within the file.
19-in. relay racks in quantities as required within each divisional cabinet.
In essential safety systems incorporating multichannel logic design, cards can be configured within the files such that it is only possible to calibrate or ma fe--e preuedes inadvete= sy-em -uvity O
test -e channet - a ume.
because of erroneous test procedures.
1-7 i
~. -
v -
NKUV-4Loat=3 CARoplLe
..a...t ?/////fh)555f J
- *~ ~ **~ ~ w d i.~,,. s...,E$i [fs r * - a - m
%f%%
f 1
amme
- ,A,to.g c
d e, ' *!.
l
...i.
le e
l 1
4 4 9 @
s;i 1
1-[
i i
u-i
',5h 1
/
w.
L
. d.
3 j
.}
..O e
%L- --C,C.,ti, g ~g, t
l--
MAsTsR t
l-
- L-TRIP UNIT 3 M ASTE R BL NK CALIBRATION o
u AN LoQ TRI U. N TS PANELS UNIT
\\
QUTPUT Figure 1-2.
Typical Trip UnitMalibration System File
)
1.3.8 Bench Test Facility A power-up device is provided to each site to facilitate standardization and trouble-shooting the components that comprise the Trip Unit / Calibration System.
The functions, of the Bench Test Unit are:
to provide a means to standardize the readout assembly to an on-site (1)
~
standard;
- to provide a facility to trouble-shoot a failed trip unit; and (2) to provide a way to perform a functional test procedure on a trip (3) unit that is equivalent to the acceptance test procedure originally This procedure performed on the trip unit when it lef t the factory.
will be performed any time the trip unit is repaired.
l t
I 1-8 s
v NEDO-21617-A 1.4 APPLICATION three time stages of plant construction are defined:
For purposes of this report, A plant in Phase A is in its initial design stages with elementaries (1) and functional drawings being actively worked on by system engineers.
Most projects in this category are of the BWR/6 design.
A Phase B plant has been released for construction with system docu-(2) l ments " frozen" but the plant not as yet started up.
A Phase C plant is fully turned over to customer, started up, and (3) operating.
4 Most Phase A plants have the analog sensor feature built in for all safety sys-tems as a standard package with transmitters mounted locally and trip units The analog transmitter / trip mounted in divisional panels in the control room.
unit sensor scheme is not intended to be an option here, but is an integral part
^
of the standard safeguard system design.
i r
Most of the Phase B and all of the Phase C plants have been designed with direct General Electric Company is offering a retro-acting mechanical switch sensors.
If they elect
~
fit program on an individual basis to several Phase C plant owners.
to purchase the new hardware, it may be installed at the old switch location or in the control room, relay room, or in the reactor building, as desired.
O i
i 1-9/1-10
'{
i
-l 0
-._.a.---........
~ '
2.
SYSTEM COMPONENTS
,s 2.1 TRANSMITTERS Any type of analog sensor having a 4-20 mA signal (or a 3-wire RTD) may be used with each master trip unit. Transmitters qualified for nuclear safety applica-tions have been available as " shelf" hardware for many years, so a detailed description of their internal mechanisms will not be discussed here.
A typical transmitter is shown in Figure 2-1.
Transmitter performance characteristics as required for the tandem elements used to generate a trip are discussed in Subsection 3.3.
T i
l M-?g..
, Y, f-],$~% '
i
,,.. 1.} 'W,'....
}
.3 13g h
- 4. ;
's w' p
m a
.[)
I s
i
...' g.
N y
N
)
l N
\\
%k Figure 2-1.
Typical Pressure Transmitter 2-1 i
.m m-
n e,w.c m,.
for Phase A plants utilizes the new hardware The standard BWR/6 design concept Future applica-
)
for sensing pressure, differential pressure, level and flow.
retrofit packages can include temperature sensing by tion for B and C plant using a trip unit that connects directly to a three-wire RTD.
2.2 TRIP UNITS AND CALIBRATION HARDWARE How-All of the trip units used in this application are electronically linear.
ever, nonlinear scales are available on the master trip unit meter f aces for Flow measurements may use in direct measurement of flow using a AP sensor.
also be taken by coupling a [AP transmitter with the trip unit having a linear See Reference 1 for complete detailed descriptions and meter scale if desired.
specifications of the trip units and associated supplied hardware.
2.3 POWER SUPPLY 2.3.1 115 Vac to 24 Vdc Ferroresonant The primary concern in selection of an appropriate power supply (Figures 2-2 and 2-3) for 'the new hardware was reliability. Most power supplies available on the
~
market employ series regulating techniques which provide almost unlimited preci-However, greater regulation means more series elements creating sion if desired.
Furthermore, a short circuit of these. series voltage regulat-less reliability.
ing elements could cause high voltages' to be transferred to the loads, which, in For this application, would be all the trip units of an entire power division.
this reason, precision voltage regulation is done internally within each trip individually so that any single failure of this circuitry could affect unit In addition, the trip units are naturally isolated from only one logic train.
each other by virtue of their individual regulators.
The ferroresonant type of power supply was selected because it is extremely reliable due to its simplicity, since relatively few components are used in its the ferroresonant transformer, design. The unit consists of only six parts:
two capacitors, a diode bridge, a resistor and a diode.
The units have no series regulation components and have reasonably good line Adequate voltage regulation for the trip unit is provided l
voltage regulation.
by the regulation circuits in the trip units themselves.
l l
2-2
=
O.
m 2
men V w
VOLTAGE REGULATION 2
b 26
(
> ALLOWABLE RANGE l
( S PINS 4 4 5)
O 25 h
min V 24
=
F 23 3 l
l g
23 l
l n
i l
1 1
I I
0 20 40 m
30 100 l
CURRENT (%)
VOLTAGE REGULATION CURVE
( +l' 3
s I
CR1 CR2 y
f).
w
?
I iO I
CI Ic
> OUTPUT' 7
+
INPUT C2 % [
gy h
NEUTRAL GROUND C1 3&
(-)3 (COMMONI 5,
f ON g
f t
t
' UNITS OF SAME PART NUMSER MAY BE TIED IN PARALLEL l
AT PINS 4 AND 5 WITH PIN 6 USED FOP. VOLTAGE MONITOR SEN60R.
IF SINGLE UNIT IS USED, PIN 6 MAY SE USED FOR POWER OUTPUf.
I 9
Figure 2-2.
Typical Schematic of 24 Vdc Power Supply O
2-3 l
t
')
/
?
s
?
t
\\,
.se
\\
Figure 2-3.
24 Vdc Power Supply
']I l
2.3.1.1 Theory of Operation Thedepowersuppliesincorporateaferboresonant transformer with a magnetic shunt, which provides a high reactance to all secondary currents above a certain One of the transformer secondary windings is resonated with a capacitor, level.
which causes the voltage to' rise until the core under it saturates and no longer limited and sustains additional voltage. Thus, the secondary peak voltage C1 maintained constant and, after rectification and filtering, provides a nearly j
The resonance begins l
constant de voltage for any current within its range.
Vell below the input voltage range, assuring good output voltage regu at on f
l i i
for changes in input voltage.
Under absolutely no-load condition, the output voltage is higher than desired.
l To prevent this condition, a resistor is permanently connected internally to preload the power supply to about 7 to 10% of the rated current.
l l
i
)f l
2-4 I
mI
A I
- Thus, The secondary copper conductor's resistance increases with temperature.
b' the' output voltage reduces with temperature, but this reduction is small and The change in the diode's forward is within the limits of the specifications.
voltage drop compensates partially with temperature or change in the copper The diode voltage drop is in the order of 0.6V, which reduces
)
resistance.
with temperature at about 0.002 V/*C.
)
Due to the magnetic shunts, the transformer secondary windings cannot exceed a 1
Thus, the output can be shorted indefinitely without any certain VA limit.
o damage to the unit.
Voltage regulation is provided by the ferroresonant transformer, which is a Therefore, frequency variations of 11% will frequency sensitive component.
This condition result in a corresponding 11.3% change in output voltage.
remains linear from 55 to 65 Hz.
DC power supplies of identical rating can be connected in parallel to obtain r
Output voltage connections must be made by connect-
.)
a higher output current.
l If the polarity is not observed, there will be no voltage ing like terminals.
output from the paralleled bank. A diode is provided on the output of each power supply to protect and isolate individual power supplies.
l 2.3.1.2 DC Power Supply Specification i
Input Voltage e
h Normal:
115 Vac At a Power Factor Maximum:
127 Vac >
from Minimum:
102 Vac 0.75 to Unity e
Input Frequency:
jj 60 or 50 Hz as specified e
Output Voltage:
)
Normal:
24 Vdc j
Maximum:
28 Vdc At No Load Minimum:
24 Vdc At Rated Load 1
2-5 j
1
v--__-__.
output Current:
2.5 to 30.0 Amperes, as specified Output Voltage Ripple:
e 1% RMS Naximum Full Load Efficiency: 64% to 79%, as specified.
.e Power Supply is capable of continuous operation at no load.
Environmental Conditions:
o Temperature:
Minimum.: 40*F Eormal:.
60 to 90*F' j
Maximum:
120*F Humidity:
Minimum:
10%
Normal:
40 to.50%
Maximum:
90%
l Qualification Tests e
(See Section 4) 2.3.1.3 Actual Test Data (20 Amp Unit)
Efficiency 115 Vac 75.16%
e Power Facter 120 Vac 0.89 e
115 Vac 0.91 127 Vac 0.90 Ripple PK/PK at 100% load 0.6 Vac e
0.88%
and 115 Vac 9
dl i
1
)
i 2-6 j
.__. __. i
1 28 I
(
REGULATIONS MEASURED FROM OUTPUT l
27 PtNS 4 ANO 5 AS SHOWN ON FIGURE 2 2 2.
1 s
I i
- 5 25 8
J I
E 2s a
so 23 l
D
=
i t
i i
I I
I I
I I
)
0 20 40 GO 80 100 100 200 SHORT CUMMENT (%)
Figure 2-4.
DC Fower Supply Regulation I
i
~
- 2. 3.'2 125 Vdc to 115 Vac Inverter
, _j The inverter (Figures 2-5 and 2-6) is the necessary power interf ace between the station batteries and the ac to de power supply in the nonfailsafe ECCS The GE units are essentially standard with three modifications:
systems.
The on/off power circuit breaker has been removed because of the (1) risk of inadvertent power rem' val in essential systems.
o (2) An input filter choke was added to absorb electromagnetic inter-ference (EMI).
Mechanical strengthening of the transformer base place and circuit (3) board mounting hardware in order to qualify the unit for seismic specifications of Class 1E essential equipment in compliance with IEEE-344-1971.
l 1
O 2-7
- x e.. -. sm.r.- =a :-:-
_. - _ w n.
1
-,z-----.
'h
..i
.A
- xg W*
Q M
- t.
n P44 Q.
- t. 4 tra E
f
's
/
.k. '.. a
=
i'
%u#
-' 'o. O*.*
-lj y
djf. 4.., '.Mf*M;*z _ ;M^,...1 W
M e
.yl?,'....a V*'QX**ff...n,..f V.'**
h*'.g./.9JsMW ;..*b*:. - o *% %f._9Q % 'd g
.g N
h, '
k~
.A* %gf,.b.'y,.
,e
.y 1 h*
- b
~
.f j
o
- 8 e
., a(
4 t%
g
,, e /,
'p O
i da N
e.2.'.j'e N,
- 7.Q
<<c..
'E.&
m sn
n
,26}yjg;gW 0-A
- W h. mp.~~
- t. Car Se g
pm=*).-
. y y,. -..
Sie e..p.* =.=
~
- jk. t..ru.
e te, ' S.%
al i
,e*
a f*n d,
. g ; 43..
h..., J.... ;'
3 A
ssw u.... -
.n
'M
)
I a
.s 2-8 I
=o
V
~
3
,m,.
t f...)
E V
HH E e se HH y
HH g
HH 4
i h,
- auA*
3 u_m a
a g
_______vv1 g.,
___________1,
- r111 g1 1
$2 E R
El E 8th_._@._@@.@_@]
n, g
\\
a
.x h U
a nQ o
oQ q y
e ss.oo, g
ao m
a N_
4
,a g
j Y
na
,gg
=g 1
s a
A.
=
I d>+
n
.s
.s
=.,
=
4 5
.. A.a.
J an i
5 i
.V&
-y a
a
,4
.u "a
8 E
w m
bu=h..,M i
.E se g L
<M.
~2 [
6 a
N b
I s.
I 5
"3 i
I z-l
(]
lE[$
_k i
2-9
? -
Theory of Operation and Circuit Description 2.3.2.1 The. inverters convert a nominal.125 Vdc input voltage to sinusoidal single-phas The units' are all solid state ac output at a specified voltage and frequency.
More than one inverter and provide output overload /short circuit protection.
unit may be connected in parallel to increase output power capability.
i The 125V input inverters have a reverse polarity protection diode (CR1), which f
~
{
h will trip an external circuit breaker if the input source is connected with t e l
A surge limiting resistoi R1 limits the -initial charging cur-wrong polarity.
An input rent to input filter capacitor C1 to prevent breaker trip at turn-on.
f power relay closes to connect the input filter capacitor to the center tap o h input level i
the inverter transformer T2 only af ter C1 is nearly charged to t e On closing, and only af ter the input level is above the low input voltage limit.
R1 is shorted out by one set of the K1 contacts. to prevent voltage drop and f
power loss due to R1.
Af ter K1 closes, the A1 inverter control PCB assembly will, af ter a slight ' dela w
initiate gate drive pulses to the inverter SCRS Q1 and Q2 to start the inverter.
The inverter always starts with a quarter cycle with all subsequent pulses of This special quarter-start festure is mechanized normal half-cycle duration.
on the A1 control assembly to avoid high inrush currents,that could otherwise result due to initial " transformer set" if the power transformer were driven for a full half-cycle in the same, fluz polarity of its retained residual flux.
Output voltage regulation and overload /short circuit protection are provided When two units are paralleled in master-by the ferroresonant transformer.
slave. configuration, the inverter control board is removed. from the slave unit The output of the mester unit inverter and replaced by a slave control board.
control board is input into the slave control board, thus synchronizing the The inverter control assembly contains slave inverter to the master inverter.
/
circuitry for fixed frequency drive, quarter-cycle starting, input undervoltage overvoltage detection, and automatic restart.
2-10 t
_._,a o-
i1 2.3.2.2 Inverter Specifications-
'O DC input voltage range is 100 to 140V for all versions (Table 2-1):
Table 2-1 INVERTER SPECIFICATIONS I
Maximum Input Output DC Output Output Current (VA)
Input Frequency Voltage.
(A) 200 125 50 115/230 2.5
'400 125 50 115/230 5
800 125 50 115/230 9
l 250 125 60 115 3
j
{
500 125 60 115 6
1000 125 60 11'5 13 l
l 1500 125 60 115 20 1
1600 125 50 115/230 18 j
i
]
l
. ~.
e output Power l
The VA ratings in Table 2-1 apply as output power in watts at unity Full VA ratings (in volt-smps) a'pply with load power factor only.
power factors of 0.75 minimum leading or lagging.
4 Regulation e
115 Vac 25% with line, load and camparature changes, 4
e Frequency 50 Hz 10.5% with line, load and temperature changes.
60 Hz 10.5% with line, load and temperature changes.
l output Waveform e
The output is sinusoidal w1th 5% maximum total harmonic distortion at full load (typically 3%).
l v
s 2-11
_-__ _ _ _ i NEDO-21617-A J
Y e
Protection low and high input voltage detectors turn the unit off when-Input:
the input volt. age is below' and above the'specified input voltage limits.
When the input returns to within the specified input voltage limits,.
the unit will again turn on.
the output is protected -from overloads and short circuits
-Output:
with automatic recovery upon removal of the. overload /short circuit.
f e
Environment j
operating Temperature':
14 to 131*F
-40 to 167'F i
Storage Temperature:
Parallel Operation '
e The 1500 VA, 60 Hz output system consists of a 1000 VA, 60 Hs unit and.
a 500 VA, 60 Hz unit operated in parallel, with the 1000 VA unit as a master control unit and the 500 VA as a slave unit incorporating a Similarly, the 1600 VA, 50 Hz system consists of two slave PCB card.
800 VA, 50 Hz units in parallel, one. master unit and one slave unit.
}
e Cooling All units are fan cooled.
d 2.4 TRIP RELAY i
l The nominal output of the trip units is 24 Vdc for logic 1 and 0 Vdc for logic 0.
Therefore, to replace the contact logic of the former mechanical switch, a relay j
It The relay has been qualified for use in Phase A and B plants.
I is required.
is a standard unmodified relay, as shown in Figure 2-7.
Each relay provides four single-pole-double-throw (Form C) contact sets and consumes less than 6 watts of power within the coil.
Its design. life is
~
mechanical operations which is significantly more than the expected 5
5 x 10 usage in 40 years of operation. Other specifications of interest in this l
application are noted below:
)
1 4
2-12 t-
i l
_ _ _m
9 I
/~
L.
y f
i
(
.. '.n.e -
g,'w..
c l_...
g.8'
- 5. 0 g *I 8 8 g,8 i
g.' s 8 ' s I
. i, f Figure 2-7.
24 Vdc Trip Relay
)
1 O~
Contact nominal voltage and current ratings:
10 A at 120 Vac, e
1.5 A at 125 Vde; 0.2 A at 250 Vde, 10 A at 24 Vdc f
e Contact resistance: Less than 10 milliohms.
I transfer time requirements for the 24 Vdc coil:
e Contact (1)
Between energizing and opening of normally closed contacts:
12 msec, maximum (2)
Between energizing and closing of normally opened contacts:
20 msec, maximum (3)
Between.de-energizing and opening of normally opened contacts:
8 maec, maximum (4)
Between de-energizing and closing of normally closed contacts:
18 msec, maximum Voltage Requirements for 24 Vdc coil:
e (1) Minimum sustaining voltage at 68'F:
19 Vdc
.Q (2) Minimum pickup voltage at 120*F:
20 Vdc (3)- Maximum continuous use voltage:
27 Vdc 2-13 I
}
NEDO-21617-A Coil inductive load:
0.28 Henries e-The relay may operate for extended time periods Operational Use:
e (months) in the energized or de-energized state before safety operation is required.
Maintain state during seismic:
contact must not falsely. transfer e
during a seismic event with the following maximum chatter time -at
. the specified maximum seismic _ g load.
De-energized: Normally closed contact with 10 msec chatter time at 6.7 g's.
Normally open contact will not have transfer contact at 6.7 s's.
Normilly open contact with 10 usec chatter time at Energized:
17 s's.
Normally closed contact will not have l
transfer contact at 17 g's.
1 Environmental Conditions:
e Parameter Minimum Normal Maximum (1) Temperature ('F) 40 60/90 120 1
to 40/50 90 (2) Humidity 1.0 in. WC (3) Pressure (State Gauge) 0.1 in. WC 2.5 COMPONENT ACCEPTANCE Each component specifically identified above and associated equipment for Class 1E application is required to undergo a nondestructive acceptance testing procedure to assure its functional ability prior to shipment to General Electric.
These tests are identified separately from the qualification test procedures, which may or may not be destructive and are usually done only once on a rep-resentative sample or set of samples. Qualification testing is treated in l
Section 4.
.The acceptance tests are performed in an environment considered " normal" by General Electric Company in accordance with the installed location of the The following is a listing of the principal hardware featured in the device.
l' l
2-14
- w.
- i
_______._.__.m.
._.m.
s..~
m
_.m
.v
... __= ;
1l cuales transmittst trip unit ctusing concept, with ecczptanco t20t procedurso.
l
[ ~ briefly summarized for each component.
.I j
f' 2.5.1 Transmitters, 4
Procedure calls for hydrostatic og overpressure test, calibration, visual and mechanical inspection.
Standard environmental test conditions include:
77'F 19'F Ambient temperature Relative humidity:
80% or less 28 to 32 in. of mercury Ambient pressure:
a Tolerance, failure and ratest criteria, along with recordkeeping and tooling requirements, are defined in this procedure.
0 4
3 2.5.2 Master Trio Unit 9
Each master and slave trip unit will be " burned in" by the vendor prior to their I
The burn-in consists of 48 hours5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br /> of ON/0FF power cycling 9
acceptance testin~g.
at room temperature with maxistan loads on both the trip output and the. gross This burn-in procedure is intended to detect any infant failure output.
mortalities caused by defective components or poor worlenavahip.
By cycling at 0
the maximum operating voltage (28 Vdc) and with maximum loads on the output, the components are subjected to maximum power conditions which should preci-pitate a congionant failure if a component is on the verge of failing or I
indicate' a workmanship problem if such a problem exists.
The acceptance test procedure defines test conditions as follows:
I I
l 70 to 80'F Temperature:
t Relative humidity:
20 to 90%
e 0.5 mR/h, maximum Radiation:
Power supply:
23.90 to 24.10 Vdc Units are visually and mechanically inspected, then energized and functionally Q tested with a semiautomatic test set specially designed by the vendor for I
i.
2-15 l
1 a n. _...... -
m NEDO-21617-A The acceptance test proce-
]
acceptance testing of master and slave trip units.
/
dure for the master trip unit verifies that the adjustable parameters and logic functions meet the acceptance criteria shown on the data sheet (See l
Reference 1,Section III for full explanation of adjustments and displays):
I (1) Low and of trip adjustment range (2) Maximum low current (gross failure) trip point (3) Minimum low current (gross failure) trip point I
(4) Meter zero reading (5) Analog output at 4 mA (analog output units only) c (6) Midscale trip point adjustment '
(7) Maximum trip point re' set dif forential I
(8) Trip point repeatability (9) Trip point repeatability for 23.5 Vdc power (10) Trip point repeatability for 26.5 Vdc power (11) Minimum trip point reset differential (12) Analog output et 10 mA (analog output units only)
Tl (13) Analog output at 15 mA (analog output units only)
~
(14) Analog output at 20 mA (analog output units only)
,15) Meter full-scale reading
(
(16) High and of trip point adjustment range (17) Trip output logic select operation
(
(18) Trip status light logic select operation (19) Minimum high current (gross failure) trip point (20) Maximum high current (gross failure) crip point (21) Calibrate command function Acceptance test data sheets are provided with specified tolerances or conditions for each step as shown above. The test operator compares entries Quality Assurance against the specified limits and initials each data sheet.
verifies all data and makes a stamp entry on page 1.
n I
2-16 t
I wa - --. -
I
~
~
~
NED0-21617-A l
2.5.3 Slave Trip ~ Unit _
The ' acceptance test procedure for the slave-trip unit is the same as ' for the master trip' unit except for the following steps:
(2) Maximum low voltage (gross fsilure) trip point (3) Minimum low voltase (gross f.ailure) trip point
. (4) (Deleted - slave has no meter)
(5) (Deleted
' slaves have no analog output)
(12) (Deleted
-slaves have no analog output)
(13) (Deleted - slaves have no analog output) 1 (14) (Deleted - slaves have no analog output)
~
(15) (Deleted - slaves have' no meter)
(21) (Deleted - calibrate command function is taken at the master unit only).
2.5.4 Card File
~
The acceptance test is performed for the card ' file in the same environment that
(
is used for the master trip unit acceptance test. Dummy trip unit cards are plugged into the file, and continuity checks are confirmed with a lamp indicator for all previred or unwired card files for all printed wire card pins to the The presence or absence of continuity is recorded in the terminal screws.
Then a thorough visual inspection included data sheets for each terminal pair.
-l 1s made to determine if shorts exist between wired points on each card connector.
Repairs are made as required, then reinspected and approved.
2.5.5 Calibration Unit The acceptance test is performed for the calibration unit in the same environment The test step headings that is used for the master trip unit acceptance test.
are outlined bel'ow (See Raference 1 Section III for explanation of adjustments and displays.):
(1) Chassis. ground viring J.s (2) Maximum stable current 2-17 i
t l
n (3) Minimum setblo currsnt (4) Stable current powsr cupply cansitivity
]
(5) Minimum transient current (6) Maximum transient current (7) Transient current power supply sensitivity (8) Stable 'and transient current summation (9) Trip status continuity (10) Calibration location select (11) Readout assembly engaged relay test (12) Trip status location select f
(13) Current source rate test Data sheets and record documentation procedures are similar-to those previously discussed.
2.5.6 Raadout Assembly-The acceptance test is performed for the readout assembly in the sans environ-
{
asnt that is used for the master trip unit acceptance test (See Reference,1,
~N l Section III for explanation of adjustments and displays.):
.J (1) Minima display reading
)
(2) Display accuracy and tracking at 4 mA (3) Display accuracy and tracking at 12 mA (4) Display accuracy and tracking at 20 mA (5) Maximum display reading (6) Trip current display blanking (7) Trip current display gating (8) Trip dispisy reset 2.5.7 115 Vac to 24 Vdc Power Supply Before testing, each unit is " burned-in" for 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> under full load and nominal The environment in which the acceptance tests are input voltage conditions.
conducted is as follows:
60 to 90*F Temperature:
Humidity:
40 to 50%
Pressure:
Atmospheric I
2-18 I
a:
I b'
Af ter mechanical / visual inspection, the unit is connected with an ammeter and g) a voltmeter to the input and output of the power supply.
All test steps must satisfy the following criteria:
Requirements
}
output Voltage (Input Voltage = 115 Vde) l Normal 24 Vdc Maximum 28 Vdc At no load i
)
Minimum 23.5 Vdc At rated load The following items are measured and recorded:
}
ferroresonant jump (Voltage in and Current in);
(1) open circuit (2) open circuit Voltage out with Voltage input = 115 Vac; I
l short circuit Current out and Current in with voltage input at 90, (3) 102, 115, 127 and 140 Vac (data recorted for each voltage); and o O rated load Current out, Voltage out and Current input with voltage (4) input steps as in (3) above.
p Data sheets are signed by test and Quality Assurance people prior to shipment, a
2.5.8 125 Vdc to 115 Vac Inverter t
The inverter test procedure specifies the same environment as the de power The units are mechanically and visually inspected, supply in Subsection 2.5.7.
)
then subjected to the following test steps:
{
i (1) energize and observe open circuit waveform on oscilloscope; J
(2) trip voltage sensor and observe internal voltage balance across filter capacitors; 2-19 i
i w
c--y---__----_
)
(3) measure control circuit board input voltage; (4) adjust for correct frequency; (5) confirm saturation of output driving SCR's; (6) check total harmonic distortion with distortion analyzer; (7) short output and measure input current response; (8) adjust and recheck idw voltage sensor; (9) adjust and recheck high voltage sensor; (10) record input current and output voltage at low, nominal and high input and at no load an'd full load (This test confirms proper line and load regulation and static voltage regulation.); and a).
(11) run the snit at nominal input and full load for a minimum of 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />; record actual time.
The data in each case are checked against the specified limits, then recordad, stamped, and signed when all tests are satisf actorily completed.
2.5.9 Trip Relay 1
The test environment is specified as room temperature (2515'C) and normal atmo-spheric pressure. The releys are mounted vertically in a test fixture, which provides measured rated voltage to the coil and cycles (energize and de-energize) the units 5000 times. ' All contacts are monitored throughout the test and any that fail to perform their function constitute relay failure. All units are mechanically and visually inspected, then dated and stamped by individu21 j
performing the test.
..I o
2-20 j
i m
NEDO-21617-A-i A
3.
SYSTEM DESIGN
/
)
1 3.1 COMPONENT INTERCONNECTION l
i i
The transmitter / trip unit hardware can be used for process trip inputs into the emergency core cooling system (ECCS) and the reactor protection system All of the trip unit card flies and power supplies for ECCS are con-
)
(RPS).
tained within the core spray (CS) system panels that are in the control room These devices operate with logic in the energize-to-for Phase'A plants.
actuate mode, using the 125 Vdc station emergency battery for their power Similarly, the RPS contains its own power supply and trip unit card source.
4 files within the RPS panels. These devices operate-with logic in de-energize-to-activate mode, using the 1.15 Vac power from RFS power source.
1 3.1.1 Emersency Core Coolina System Applications A typical ECCS interconnection arrangement for one diyision is shown in l
All components shown in this figure are fully redundaht in th4 Figure 3-1.
The 125 Vdc power is brought into the. inverter from the same other divisions.
The inverter con-station bettery as that, supplying the system logic itself.
verts the 125 Vdc to 115 Vac. This power la then fed to the 24 Vdc power Pin 3' of the trip unit is the power common, and pin 4 is the signal supply.
These conunons are kept separate throughout' the trip unit circuitry; coimmon.
This arrange-l however, they are tied together at the power supply negative bus.
ment of the common return keeps the sensitive analog signals separate from the devices within the trip unit that are either on or off. Therefore, transient '
loads (e.g., the internal calibration relay, SCR) will not aff ect the sensitive analog circuits.
Figure 3-2 is a block diagram depicting the internal functions of the ' rip t
unit and their relationships with the connecting pins for the trip unit designed for accepting a 4 to 20 mA signal.
The trip unit designed for an RTD input is similar, so it is not shown.
It can be seen that current from It is then pin 2 is passed through fuse F1 and an 80-ohn resistor to pin 14.
routed out to the transmitter and returned at pin 15, where it is carried t
3-1
?
\\
l I TO 125 VOC SUS I
FROM PROPER
( INVERTER OlVISION R
SUPPLY
(- 8US)
(+ sums)ll ; - ;;;\\\\\\
^ ^:
\\\\
F
/////
TO OTHER TRIP UNITS i
' TO OTHER
'^
UNI 3
,r-
$1I T"E
~
9 0==
PRESSURE C l'
^
0
,3 TO SYSTEM LOGIC TRANSMITTER _
()
8 l
15 g.20 mal O'==== =C O
TRIP CARO i
UNIT ouT y
1 4 13 12 10 ALARM s
o o i
y p; n
(_
7' ~
gy
-TO LOCALPANEL i
~
AUXILIARY TRIP UNIT J
\\,,A,,,,*
ANALOG TROUBLE" JUNCTICN SOX OUTPUT TO J
RECORDER
\\ GROSS' NM EN RACEWAY Y
FAILURE h
TROUSLE" TRIP UNIT TO SLAVE OUTPUT CAstNET IP USEO Figure 3-1.
Typical ECCS Connection Diagram
]
u cwAesas oROJNO,%
m.ie.mm.AL
-==o-4.7 V o.
- o...v tar status Lao
'U"L Y
====O +e2 v mer
*"' **" 'I cow anAron 3 /
- e2W asp o outputitso 24 v powen nefumn a h LOGIC swr 7CM
==
=
s, TRIP ST LTuS y
3, OUT7ut AOJUST e
teGNAL RETURN 4 COMPARATC'M it2 A
~
s<-
h y
- "n g, Q 0
i i========
CAAD OUT ALARM i II GR l
nasat
}
%,;}
- 90 0 enas
- s. -
. n - nA> arm y
Aa'n 7*Myn 84,
I '""""
g u
EXCffATON vi i 't? A to aAituma N'
'My CALienAftCOMIA48 0 4 '
r I' *=*n'd f n
g
- "x
.m.
.A
.A......
, g
.A.,.
?~]Yq
- . +6
~ ~r,=. :
i..
,,A,h,,N m,,==
- l A,v.s.
f 1A.a A
g
'c"v"*
y,g, m
..m..
At,,,
p ga
'f!'1u
.=,;= t V
/
Q ;c,==A sc46 oA-LuAsm wm. A=Atoo oureur o=tv
_ s A Figure 3-2.
Master Trip Unit Functional Block Diagram l
i 3-2
NEDO-21617-A O.s Here the 4-20 mA current through K1 contacts to a precision 250-cha resistor.
The series signal is linearly converted to.1-5 V as required by the trip units.
combination of the two resistors acts to limit current to 73 mA maximum in the event of a short circuit on the transmitter loop.
Both short and open loop failures' are annunciated by the " Gross Failure" detection systems, as explained in Subsections 3.2.2. and 3.2.3.
Master Md' slave trip units may be arranged in combinations as required by the application of tha l system logic.
A sample connection is shown in Figure 3-3.
Redundant drywell pressure sensors and master trip units -are shown with trip The contacts from these relays may be connected as "A" and relays K1-and K2.
"E" in a system requiring (4+E) + (C+G) logic such as the automatic depres-
~
surisation system (ADS). Since. ADS sensors perform a seal-in function, the K1 and K2 relays in this example would have to pick up additional relays powered from the 125 Vdc bus designed to accomplish the seal-in functions themselves.
It is important to note that the trip units cannot " seal in" their trip relays.
When a trip signal is once generated, it is sustained only until the parameter
]
raturns to a value as determined by the setting of the reset differential' adjustment pot (Table 9, sheet 15 of Reference 1), which is a maximum of 7.5%
The pressure span of the drywell penitor' for ADS is 0 to 10 psi with of span.
its setpoint at 2.0 psi (increasing). Thus, if the reset differential pot were set at its maximum, the trip unit wuld automatically reset the trip relay at 2.0 - (0.075) (10) = 1.25 pai.
For this reason, if seal-in logic is required, it is designed external to the trip unit system as if the trip relay contact were a pressure switch.'
The right side of Figure 3-3 shows how a single differential transmitter may be used with a master and slave combination to sense two different water levels.
is connected The 1 to 5 Vdc analog output of the anster (pin 12 of master card) to the inpuc of the slave (pin 12 of slave card). The'crip output logic switch (number 5 of Figure 9 Reference 1) allows selection of-either increasing or Other connections decreasing trip points on master and slave independently.
shown on pins 2, 8, 9 and 10 are discussed in Subsection 3.2.
3-3 I
=._ m.. _ __
~
NEDO-21617-A PRESSURE aP PR ESSURE PR ESSUQE TRANSMITTER TRANSMITTER TRANSMITTER
.n n-
.n n-
.n c-
)
Ly-gJ Lp-_d----y Ly -Vc--fJ SLAVE TRIF UNIT MASTER TPflP UNIT MASTER TRIP UNIT MASTER TRIP UNIT rH-HHh rHHb rH-M 7,rifW L.J s
s-DRYWELL PRESSURE TFilP LEVEL 1 THE LEVEL 2 TRIP/
w WATER LEVEL K3
of Master and Slave Trip Units Figure 3-3.
Sample Combination Attangement:
Figure 3-4 illustrates how trip unit outputs may be tied in parallel to drive
' one relay set (up to 4 relay ' coils in parallel).
This is possible because the output signal of the trip unit (at pin 11) has built-in dioda isolation Thus, any one or more of the parallel units may pick up the (Figure 3-2).
The logic, as shown relay (s) with no adv'erse af fects from the other units.
in the solid lines of Figure 3-4, is actually used in the ADS for RHR and CS It is seen that operation of one of pump operacing permissive confirmation.
two RHR pumps or one CS pump satisfies the permissive condition to energize-Of course, K1E (not shown) must also be energized by a duplicate set of K1A.
sensors to satisfy the (A+E) logic to initiate depressurization from division 1.
Figure 3-5 The relay sets need not be confined to one cabinet or system.
shows how ADS relays for vessel water level B21-K1 and K2 are being controlled in parallel with CS (E21) relays with a trip unit housed in the CS panel.
However, relays wired in this manner nust be within systems sharing a conunon division.
1 3-4 i
v NEDO-21617-A CORE SPR AY Ram PUMP PRES $URE RHR PUMP PMESSURE
.g-PUMP PMESSURE
____q f
PRESSURE PRESSURE PRESSURE I
TRANSMITTER TRANSMITTER l TRANSMITTER I PRESSURE TRANSMITTER I
' L-7F l
i I
l Ett TRIP UNIT f~~
E21 TRIP UNIT E11 TRIP UNIT ku 4"
(_ __ r! _ j l
I l.
- 1. - (ETCs l
I I
I toiv )
H12Mf7
() (SENSOR INPUT) l olv i RACEWAY
=
=
toevs) g,,
(ADS CASINET) g y
H12MuS '
Figure 3-4.
Use of Trip Units as Parallel Drivers s
REACTOR WATER LEVEL A PRESSURE TRANSMITTER 321 TRIP UNIT 1
l l
l 1
0 11 l
l 1 osv s I
rRACEWAY' O
--= -
====5 p-l 821 M2 E21 #1 821 K1 l
E21.M2 l
(SENSOR INPUT)
+
l I
+
v l
tolv i s (408 C_AS_INET) H12M26 tolv in H12M17 Figure 3-5.
Trip Unit Driving Parallel Relay Coils i
3-5 i
~, m
J NEDO-21617-A N
j 3.1.2 Reactor Protection Syscam Applications l
The connection design for RPS applications is shown in-Figure 3-6.
Here, the control room inverter is not required because power is derived directly from The remaining hardware is similar 115 Vac supplied by the RPS power supply.
to ECCS (Figure 3-1) except that the trip output relays would be energized continuously by the trip unit during normal operation of the power plant.
For an increasing parameter, 'such as high reactor pressure, this requires a rever-sal of the trip output logic select switch (Reference 1) so that the trip units This contrasts go to "0" (OV) to drop out the trip relays and initiate scram.
f the ECCS logic,1Aich requires the trip relay to be energized via logic "1"
)
(24V) from the trip unit to initiate a safety action.
)
q I TO 115 VAC Sus
, PROu PROreR l
OlvislON
,g,,,
SUPPLY s
(+ sus
(- sus) 555554 Fir!!
ruu TO OTHER TRIP UNITS TO OTHER TRF TO CALIBRATION UNIT UNITS L
6 VMn
=Pur M
9C>m==
PRESSURE C fl
l
.s.
C l
TRANSMITTER 1
8 w
TRIP CARO M.20 M 3 UNIT OUT
" F
.1 4
16 12 to ALARM d
o o i
P/ o e
(_
e1 g
l1 LOCAL PANE L
- TO AUXILIARY ANNUNCIATOR TRIP UNIT ANALOG
%M H
JUNCTION SOX OUTPUT TO k
TROU6LE" R ECORDER
\\-GROSS
'?OWER SUPPLY RACEWAY Y Y PAILURE h
j TROUBLE" TRIP UNIT TO SLAVE OUTPUT CAslNET IP USED Figure 3-6.
Typical RPS Connection Diagram
~
3-6 l
_ _.. _. =.
- -. - -L
I 3.1.3 Card File Arrangement O.
Card file slots 1 through 12 may house master or slave trip units in any The thirteenth slot of every card file has its own calibration, j
combination.-
When the rotating switch on the calibrator. is linked with unit installed.
one of the trip unit card file slots, the sensor and logic action may be fully For' this ~ reason, logic pairs of one divi-simulated with the calibration unit.
sion, such as (A.E) or (A+C), can be ' arranged within the same card file for When functional temting the trip units, only one trip unit ECCS (Table 3-1).
Therefore, it is not possible to inadvertently at a time can be selected.
activate a saf ety action while using the calibrator to routinely check'setpoint trips or individual logic chain permissives.
In the RPS, cards are arranged in files by individual logic trains (Table 3-2).
j This was necessary because the de-energize-to-activate logic (A+C) ' * (B+D)
For example,'A and.C cannot
{'f requires each train to be in a separate cabinet.
E be in the same cabinet because a single panel failure could destroy both A and o
e Also, though A.and B could be in the e
23 C, causing a potential failure to scram.
e the A and 5 trip units must be in E". M r) a,mee panel from a single failure. standpoint, a:}
E Unlike the ECCS dif fersnt enclosures to allow separation of A and B power.
arrangement, it is possible to initiate system action (.a scram in this case) by S a.
- c.,
us'ing the calibration units; however, in order to do so, the technician would have to place logic from both power divisions in calibration simultaneously and ignore the annunciators.
l 3.2 TROUBLE !ONITORS 3.2.1 Card Out of File l
Each trip unit card is equipped with an internal shorting link between pins 8 Cards arranged within a common division are then connected and 9 (Figure 3-2).
so that these links form a series loop between' the positive supply voltage and f
a normally energized relay coil.
In Figure 3-3, the positive supply is con-nected to the loop by the jumper between pins 2 and 8 on the laft trip unit.
Then current is sustained by jumpers from pin 9 of each unit to pin 8 of each successive trip unit in the chain until the final card, which ties to the 3-7 1
i p....;...-.
l
,,,_y-l
v NEDO-21617-A l
s
/
\\
\\
Table 3-1 BWE/6 ECCS TRIP UNIT CARD FILE ASSIGNMENTS DIV 1 (H13-P629)
Card Function File Location Device No.
Reactor Low, Water Level Z1 1
LIS-B21-N691 A,
Reactor Low, Water Level 2
LIS-B21-N691E J'
3 PIS-821-N694A.
High Drywell Pressure 4
PIS-B21-N694E High Drywell Pressure Low Differential Pressure Across 5
JPIS-E21-N650 Injection Valve E21-F005 l
LPCS Min. Flow Valve E21-F011 Control f
6 FIS-E21-N851 Containment Pressure High 7
PIS-E12-N662C 8
PIS-E12-N662A.
Containment Pressure High
.]
RIEL A Inject Valve Dif ferential Pressure 9
dPIS-E12-N658A Low 10 PIS-E12-N651 A Steam Supply Pressure High Containment Pressure High 11 PIS-E12-N650A 12 dPIS-E12-N652A RER Flow Below Setpoint Calibration Unit Z1-1 1
Z1 14 t
S l
3-8
w
~m NEDO-21617-A i
(O l
Table 3-1 i
BWR/6 ECCS TRIP UNIT CARD FILE ASSIGNMENTS (Continued) i DIV 1 (H13-P629)
" Card Function File Location Device No.
z2 1
PIS-E21-N652 LPCS Pump-Discharge Pressure Above
'Setpoint for. ADS System L
2 PIS-E12-N655A RER Ptsmp Discharge Pressur.e Above Satpoint for ADS Systen-3 PIS-E21-N653 LPCS Ptsp Discharge Pressure Above Setpoint for ADS System 4
PIS-E12-N656A RER Pump Discharge Pressure Above Setpoint ' for ADS System 5
PIS-C11-N654A 6
.PIS-C11-N655A L First-Stage Turbine Pressure I
\\ ')
7 PIS-C11-N654C j t
8 LS-B21-N692E Reactor Vessel Low Water Level 9
LS-B21-N692A 10 LS-B21-N69 3A Reactor Vessel High Water Level 11 dPIS-E31-N683A Steamline High Dif ferential Pressure 12 dPIS-E31-N690A (Steam or Instrument Line Breek) 22-1 Calibration Unit I3 o
Z2 14 11 9 i
O 3-9 l
L--- _
-l
)
~
Table 3-1 BWR/6 ECCS TRIP UNIT CARD FILE ASSIGNMENTS (Continued)
DIV 1 (H13-P629)
Card Function File Location Device No.
Steamline High Differential Pressure-Z3 1
dPIS-E31-N684A h
dPS-E31-N684A')
-(Steam or Instrument Line Break) 2 3
PIS-E31-N685A
.Emactor Pressure Low Turbine Exhaust Pressure High f
4 PIS-151-N656A Turbine Exhaust Pressure'High 5
PIS-E51-N656E Turbine Exhaust Diaphragm Pressure High 6
PIS-E51-N655A Turbina Exhaust Diaphragm Pressure High 7
PIS-E51-N655E 8
FIS-E51-N651 Ptssp Discharge Flow Low 9
FS-E51-N'659 Pump Discharge F1ow Righ
.,}
l 10 PIS-E51-N653 ump Suction Pressure Low Pump Discharge' Pressure High 11 PIS-E51-N650 Reactor Water Levtl.Below 3 12 LIS-821-N695A Calibration ~ Unit 13 Z3-1 y
Z3 14 J
e y
of a
3-10 3
I J-w
.v NEDO-21617-A l
l
( N.
l i.
l I
I Teble 3-1 BWR/6 ECCS TRIP UNIT CARD FILE ASSIGNMENTS (Continued)
+
DIV.1 (E13-P629)
Card Function File Location Device No.
Z4 1
PIS-E21-N654 LPCS Pump E21-C001 Discharge Line
(
O I
Pressure Low L
()
2 PS-E21-N655 LPCS Pump E21-C001 Discharge Line Pressure High
)
dPIS-E31-hi680A LPCS/RER A Line Break (Leak Detection) 3 p
4 PIS-E31-N692 Vessel Plange Seal Leak Off l
b 5
PIS-E12-N653A RER Pump E12-C002A Discharge Pressure Low l
6 PS-E12-N654 A RUA Ptsap E12-C002A Discharge Pressure High
)
Shutdown Cooling High Suction Pressure
{
P 7
PIS-E12-N657
/
8 Spare I
- n
'..s i
9 PIS-E51-N652 RCIC Pump Suction ' Pressure High
'l 10 PS-E51-N654 RCIC Water Leg Ptany Discharge Pressure Low 11 Spare 12 Spare Z4-1 Calibration Unit l
I3 y
Z4 14 I1 5
e h
1 l
l 3-11 i
~
j
.s}
Table 3-1 BWR/6 ECCS TRIP ITdIT CARD FILE ASSIGNMENTS (. Continl DIV 1 (H13-P629)
Card Function File Location Device No.
Reactor Pressure for ADS System Z5 1
FIS-B21-N668A PS-321-N669 A '
Reactor Pressure for ADS System b
2 Reactor Pressure for ADS System 3
PS-B21-N670A.
Reactor Pressure for ADS System 4
PS-321-N671 A Rasetor Pressure for ADS System f
5 PS-B21-N668E Reactor Pressure for ADS System' 6
PS-821-N669E Reactor Pressure for ADS System 7
PS-B21-N670E Reacter Pressure for ADS System 3l 8
PS-B21-N671E 9
Spare 10 Spare f
11 Spare 12 Spare l
g Calibration Unit Z5 13 Z5-1 I
14 9
Y' b
m I
NED0-21617-A t:
1 l'
O Table 3-2 BWE/6 RFS TRIF UNIT CARD FILE ASSIGNMENTS
. Division 3' (H13-P695)
Division 1 (H13-F691)
Card
. Device Card Device File Location No.
Function.__
File Location No.
Function 21A 1
521-N678A Vessel Pres, Scraa Z1C 1
521-N678C Vessel Pres Scras 2
821-N679A Vessel Pres, ISO 2
- 821-N679C. Vessel Pres, ISO 3
C71-N650A Drywell Pres, 150 3
C71-N650C Dryve11 Fres ISO 6
C71-N651 Cond Vacuum, ISO 4
521-N675C Cond vacuum, 150 i
5.
C71-N653 Steamline Pres, ISO 5
521-N676C Steaaline Pres, ISO 6
C71-N652A 1st Stage Turb Fres 6
C71-N652C' 1st StaBe Turb Pres j
7 321-N680A ' Sz Level, Scras 7
57,1-N680C Rx I4 vel. Scram f
f 8
821-N681A Rx IAvel. ISO 8
521-N681C Rx Level, ISO 9
E31-N686C-MSL now, 150-10 E31-N687A HSL Flow, ISO 10 E31-N687E MBL Flow, ISO 11 E31-N688A H5L Flow, ISO '
11 E31-N688C-MSL Flow, ISO
~
12 E31-N689A HBL Flow. ISO 12 E31-N688C MSL Flow, ISO Calibration l3(
ZIC-1 9*
11A-1 Calibration Unit 11A 14j
' Unit Elc 141 4
(ISO = Isolation; Rx = liesctor; IG5L = Main Steam Line]
'T
,,_ /
1 Division 4 (,R13-F694)'
Division 2 (H13-F692)
Card Device Card Device File Location No.
Function File Location No.
Function Z15 1
B21-N6788 Vessel Pres Scram Z1D 1
521-N678D Vessel Pres, Scram 2
521-N6798 Vessel Fres, ISO 2
521-N679D Vessel Free, ISO-b 3
C71-N650D Drywell Pres, ISO 3
C71-N6505 Drywell Fres, ISO I
4 B21-N6755 Cond vacuum, 150 4,
821-N675D Cond Vacuum ISO 5
521-N6768 - Steaaline Free, ISO 5
B21-N676D Stesaline Pres, ISO 6
C71-N6528 1st StaBe Turb Pres 6
C71-N652D 1st StaBe Turb Pres 7
821-N6805 Rx Level, Scras 7
821-N680D Rx Level, Scram 8
821-N6818 Rx J.evel, ISO 8
521-N681D Rx Level, ISO 9
E31-N686D MSL Flow, ISO 10 E31-N687B MSL Flow,150 10 E31-N687D M5L' Flow. 150 11 E31-N688D MSL Flow 150 11 E31-N688B MSL Flow, ISO 12 E31-N6890 MSL Flow, ISO 12 E31-N6898 MIL Flow, ISO Z1D-1 Calibration k
115-1 Calibration 14f P
Unit Z1D Z18 14)
Unit
<O 3-13 t
I
~~ v NEDO-21617-A
[
Removal of any trip unit will break the current loop and coil of relay K4.
cause relay K4 to drop out and annunciate via normally closed contacts wired from the relay to the annunciator.
3.2.2 Gross Failure Low The.1 te 5 Vdc analog control signal discussed in Subsection 3.1 is fed into Each comparator has a reference volt-two comparators handh (Figure 3-2).
The age which can be adjusted by its gross failure adjustment potentiometers.
When low gross failure trip can be adjusted to trip between 0.5 and 4.0 mA.
the current is below the trip, comparator@ output generates a latching gross failure output signal (approximately 24 Vde) and turns on a gross failure indi-cating LED Otaference 1).
This digital output via pin 10 is diode isolated so that any number of trip -
units may be tied together (Figure 3-3). Relay K3 is shown as the gross fail-ure sensing relay and may have its contacts tied to the annunciator, computer, The relay is normally de-energized and is energized to cause
]
etc., as desired.
annunciation upon signal nom any trip unit of the chain.
The most important function of the low gross failure alarm is to sense an open (broken) transmitter loop; however, some failures within the transmitter or
'It can be trip unit are detected by the low gross failure detection circuit.
seen from Figure 3-2 that an overload of the trip output could.. force fuse F1 This action would also interr'upt transmitter loop, current and, there-to open.
fore, annunciate the power loss to the trip output as well.
3.2.3 Gross Failure High_
The high comparator@ function is identical to the low gross failure (Subsee-that it senses high loop currents between 30 and 40 mA.
tion 3.2.2), except Its output signal is tied to the same logic as the low comparatorh. The high sensing range is designed above the saturation current of the transmitter in order to distinguish between an actual gross failure and a transient condition that overpressurizes the transmitter. The primary function of the high sensor j
3-14
_ o epp
v NEDO-21617-A-
'3 is to annunciate a short circuit of the transmitter or its loop.
Some compo-
.,4 nent f ailures within the transmitter or trip unit are also detected by th's high gross failure detection circuit.
3.2.4 Card in Calibration is selected by the calibration unit and placed in the cali-
.When any trip unit brate mode,- a 24 Vde' signal is transmitted from the calibrator to pin 6 of the This is the calibrate command signal, which energises the input trip unit.
signal relay K1 (Figure 3-2) to switch input current from the transmitter to This voltage. is also imposed at the cathode side of ' the the calibration unit.
This' turns on the gross failure driving SCR for the gross failure detector.
LED and provides a gross f&ilure output signal at pin 10 as long as the card remains in the calibrate mode. However, unlike the othier gross failure cir-cuitry, the calibrate command signal does not latch the gross failure output.
Therefore, when the card is taken out of the calibration mode, the gross fail-ure output automatically resets and the annunciator may be immediately cleared.
)
3.3 PERFORMANCE i-3.3.1 Accuracy and Repeatability Accuracy and trip point repeatability for the trip units are discussed 10 Ref er,
Accuracy for the transmitters is specified in the v'endor's literature ence 1.
as 0.25% for temperatures up to 80*F with a temperature effect of 1% zero and l
span shif t for each.100*T above or below the calibration temperature..
The analog transmitter / trip unit feature utilizes tandem combinations of units Table 3-3 gives with repeatability calculated as shown in Figures 3-7 and 3-8.
the results of the calculations, where R, is the total worst-case repeatability the master trip unit output.with both master and transmitter considered.
R, at is the total worst-case repeatability of the slave trip unit output with slave.
master, and transmitter considered.
l
)
.O e
Li 3-15 1
1 x.
(
,$${
3 7
6 v.
5 7
0 9
1
)
a%
R
(
0 0
0 1
1 1
1 1
8 8
3 4
2 6
5 0
)
aZ R
(
0 0
0 1
1 1
1 1
0 2
8 0
O 2
3 2
5
)
S%(
0 0
0 0
1 1
1 1
S 1
2 2
4 I
5 5
0 0
A N
'M 1 0
0 0
0
)
O
(
1 1
1 1
I T
~
AC IF 3
0 0
0 IC 1
2 2
3
)
E M1(
0 0
0 0
P 1
1 1
1 S
R O
Q RR 3 E 5
5 9
2 6
4 0
)
3 Y T%(
0 0
0 1
T 1
1 1
1 e I l L b I a B T A
)
5 TA ees 0
1 5
0 E
std 0
0 1
5 6
P oaa 0
0 0
E DRR (
0 0
0 R
M ET my S
ut Y
mi S
id) 0 0
0 9
xl%
5 6
9 9
P aa(
I Mm i
R H
T m.
um p) 0 0
4 6
imF 9
2 0
5 1
1 1
xe*
aT(
M tne l
l m
a a
n m
m h
h o
r r
g g
r o
o i
i i
N N
H N
vnE g n s
n o e
e ti i l
s l
s nt t a
r a
r i
aa i m
e m
e l r d r
v r
v Pe n
o d
o d
p o N
A N
A O C
=
h
v NEDO-21617-A 3
\\
\\
MASTER TRIP UNIT TRANSMITTER TRIP RELAY (REPEATABILITY = M)
(ACCURACY = T)
J t
Y RM = TOTAL MASTER REPEATABILITY =[T 2, M2
)
yigure 3-7.
Total Repeatability Computation for 4
g Master Trip Unit output i
1 i
I MASTER T/U SLAVE TRIP UNIT TRIP Wy TRANSMITTER ANALOG OUTPUT (pgpgAyaggyyy. gg
==
C W ACY =TI (ACCURACY = M7 J
I L
7 R$ = TOT AL SLAVE REPtATAslLITY =/T2 + (M 1 + 52 2
b 8
Figure 3-8.
Total Repeatability Computation for Slave Trip Unit output 3.3.2 Response Time The transmitter response time is measured in terms of the " time constant," which t
f is defined as the time taken for the output to reach 63% of the final value Laboratory tests of the transmitters have verified this
{
following a step input.
to be between 0.19 and 0.20 see for both ascending and descending time constant step inputs of any magnitudes between 0 and 100% of span.
a l
I Trip unit response tests consisted of subjecting a master / slave tandem combina-e tion to both ascending and descending intput current steps with various setpoints The maximum trip unit delay time measured was throughout the O to 100% range.
With 0.5 msee con-1.5 meec af ter application of the current input step change.
C servatism, a figure of 2.0 is considered contributed by, this combination, i
1 3-17 I
NEDO-21617-A The time response of the trip relay (Table 3-4) varies in accordance with the The tandem combination time contributors specification given in Subsection 2.4.
are shown in Figure 3-9.
Table 3-4 TOTAL ANALOG-TRIP UNIT RELAY TANDEM RESPONSE TIME (msec)
~
CONTACTS Normally Open Normal 1Y Closed Initial Coil Conditions l
h h220 210 Energized h
h214 222 Not Energized
^
'E TMF RELAY TRAN9MITTER y
gy QeM NORMALLY 2
12.20.8.1s TIME 200 (mel l
l Time Delay Response Contribution of Each Componerit (msec)
Figure 3-9.
It can be observed from Table 3-4 that the response time of the tandem combina-tion varies according to which relay contacts are used (normally open or nor-mally closed) and on the initial conditions of the relay coil (energized or The response time of 222 usec, as shown in quadrant III, would de-energized).
apply to ECCS because of its energize-to-activate logic.
Quadrant II's time of 210 mese is applicable to RPS due to its de-energize-to-activate logic.
The f astest sensor response requirement for this new hardware is contained within the Safety Valve Sizing Pressure Transient envelope of Figure 3-11 for fast scram BWR/6 applications. Series time delay contributors for the instru-v mentation chain are as shown in Figure 3-10.
3-18
m NEDO-21617-A'
- D a
360 ms c
106 me C
2id ms Ag TIME 8
14 (mees),
20' 202 h
~
h a
a SPECIFICATION TRANSMffTER TRANSMITTER TRIP RELAY CR 105 L8 SENSINO LINKS AND TRIP / UNIT SCRAM CONTACTOR
- SASED ON SPEED OF SOUNO IN WATER FOR 100 ft OF TUSINO
)
Figure 3-10.
Fasti Scram Sensor, and Logic Time Delsy Compared to Specification 1410 CODE LIMM H3M W g-A BC 1370
)
1330 i
s
- 1. '~
W ALLOWA5LE SPECIFICATION LIMIT 1250 w
E SYSTEM ACTIVATION DELAY TIME (WORST CASE) 1210 g
c 0
1170 I
d SLOPE = 200 psilsec y
1130 A8C I
I I
I I
I 1
2 3
4 5
6 7
,,O TIME (seel Figure 3-11.
Safety Valve Sizing Pressure Transient Envelope for BWR/6 0
9 W
l 3-t9 i
s 1
NEDO-21617-A The allowable time specified from point A to point B is 10.35 sec (350 msec). }
Actual mechanical and electrical time delays as shown in Figure 3-10 are 244 Thus, 30% or 106 maec remain msec, which is 70%. of the specified time limit.
as a conservative margin.
j The information contained in Figures 3-10 and 3-11 is the worst-case d===nd on i
h as the. response time requirement of analog transmitter / trip un t sensor sc eme j
The only parameter requiring a faster it is applied in RPS and ECCS systems.
For this pur-response time is the turbine control valve fast closure, sensor.
drift pose, a pressure switch is still used, since the reqdired accuracy.and Should the need concern are not as critical here'.as is the high-speed response.
arise, the time response of the ' analog sensor chain can be greatly improve One vendor's use of a faster response transmitter to replace the existing one.
literature claims response time of 11msee for direct pressure transmitters This would and 1100 meec for differential pressure transmitters are available.%
shorten the chain response time in the Figure 3-10 example by 82% and 41, With the substituted direct pressure transmitter, the total time respectively.
from A to B would be only 45 meec.
COMPONENT AVAILABIL'ITY 3.4 3.4.1 Availability Criteria _
as the characteristic of an item Availability is ' defined in IEEE-352-1975 l
d expressed by the probability that it will be operational at a randomly se ecte
~
future instant in time:
Up Time Availability Up Time + Down Time
=
Availability allows for failure (i.e., complete loss of function) and subs'e-Unlike reliability, availability is not expressed as a function quent repair.
If the mean time of time, since the long-term availability is time invariant.
d unavail-to repair is very short compared to the cese interval, availability an ability may be expressed for a single component as:
~%.,#
l i
um
-. A.
T..
NEDO-21617-A Availability, and 1-A
=
A =
Unavailability "A
A
=
=
where
)
failure rate of ites, and A
=
average test interval.
0 =
further provides a method for adapting this expression to the 1EEE-352-1975 l
system level for the (1/2) x 2 logic common to redundant channels 'in nuclear
)
' This four-channel logic, which satisfies the Boolean RFS and ECCS systems.
success expression S = (A1 + A2)' (31 + 32), yields the following expressions
' for systee availability models:
Perfectly Stanzered Testing Simultaneous Testing E = (2/3) (10) 2 E = (5/12).(Ae)2 g3,3)
An _ = 1 - E = 1 - (2/3) (10)2 An = 1 - E = 1 - (5/12) (A6)2 I(3-2) 3.4.2 Mean-Time-Between-Failure (MTBF) Analysis The mathematical models for availability or reliability require the failure
' rate (A) of the individual components in the logic chain to be determined.
This is comunonly axpressed as " failures per aillion hours" and may be deter-mined with reasonable accuracy af ter several years of operating experience with Unfortunately, since the new hardware used for the many samples of hardware.
testability application does not have years of operating experience, an alter-
' native method was necessary to obtain A.
This method consisted of ' determining failure rates for each individual component (resistors, capacitors, etc.)
within each piece of hardware by use of MIL-HDBK-217B developed in 1974 by the Use of this documsat involves determining the stress Department of Defense.
ratio of each component (i.e., the ratio of the device's operating parameters to the device's rated parameter) and correlating them with the device's quality The individual component part failure rates are then added to and environment.
determine the overall A of each piece of hardware in the trip system chain.
i i
3-21
. )!
NED0-21617-A f
The Mean-Time-Between-Failure. (MIBF) is defined as the inverse of A and repre-sents the number of hours a system or piece of hardware within a system may be Each vendor of the new hardware was there-expected to operate without failure.
the results of which were used fore required to submit an MTBF analysis report, to determine availability and appropriate testing intervals as defined by Equa-The MTBF numbers submitted are summarized in Table 3-5.
tions F1 and 3-2.
The failure rate (A) and MTBF of each device are shown in the table.
Table 3-5 FAILURE RATE AND MrdF SUlt(ARY
- "P*#**"#*
A
('C) 6 (Failures /10 Er)
MrBF _
Device 25 20.00 50,000 Transmitter 56.12 17,820 35 Master 35 42.48 23,540 Slave Trip Unit 63.09 15,850 35 Slave / Master Combination 50 23.75 42,100 Inverter 80 835,000 1.20 Power Supply 6
50
<1.0
>10 Trip Relay The general consensus among the vendors is that data compiled via MIL-HDBK-217 For example, in the trip unit the failure tend to be somewhat conservative.
rate for R17, a nonwirewound (Carnet) trimmer, at 65'C, with a power stress ratio of 0.1, was calculated to be 8.64 failures per million operating hours.
Howeyer, when two of the approved vendors were consulted, failure rates at and maximum rated power (a power stress ratio of 1.0) were 0.8 failures per f
million operating hours (based on a 3.2 million-hour operatinr, sample) and no failures per aillion operating hours (based on a 1.0 million-hour operating The vendor noted that the 0.8 failures per million operating hours q
sample).
was based on a group of potentiometers taken from the first production lot.
W 3-22 i
1
HEDO-21617-A The failures that occurred were only a shif t in resistance value which exceeded their specification of 110%; no open,.or.short circuit failures occurred.
per MIL-HNBK-217B, the failure rate for this potentiometer, when oper-(Note:
ating at 70*C and maximum power, is 12.95 failures per million operating hours.)
The chief contributors of the essential failure rates for the trip units are
.relatively few in number and consist of ' the operational amplifiers having indi-vidual l's calculated at 2.2305 and R17 (as mentioned in the above example),
whose 1 was calculated to be 7.344 at 35'c (the above mentioned 1 at 8.64 was calculated at 65'C). The pertinent results of this analysis are given along with the FMEA considerations in Subsection 3.4.4.
- 3. 4. 3 Failure Mode and Effects Analysis (FMEA)_
IEEE-352-1975 contains the following statement:
"Some of the fsilures thae occur in syatens annunciato themselves.
}
and the repair process may start immediately. Other failures are not
-)
self-annunciating and can be discovered only by periodic testing."*
it can be Since the repair time (replacement with spara module) is very short, assumed to be instantaneous compared to the test interval; hence, repair time -
l and self-annunciated failures may be neglected.in the determination of an l
appropriate 6 Additionally, anny of the remaining nonannunciated failures do not affect the essential part of the tandem chain, which generates the trip output.
Primarily for these reasons, the Failure Mode and Effects Analysis (TMEA) was This document provided required of each vendor in addition to the MTBF report.
the additional necessary information to determine the nondetectable fai. lure Each component within components and their relevance to the essential loop.
each piece of hardware was tabularized with data as suggested in IEEE-352-1975.
O eSee 1EEE-352-4975,,s.e 22.
I 3-23 W M O CO eQCQs* CP 7 e#EPpg=W
l l
Analysis Conclusion,of System Availability 3.4.4
)
FMEA analysis indicates that the most probable failures originating from the inverters, power supplies or transmitters capable of affecting the essential operation of the trip units are detected and annunciated by the trip units Therefore, only the failure rates within the trip units and trip themselves.
relays need be considered in determining test interval 9.
l Table 3-6 shows values for failure rate A, expressed as failures per million hours computed from MIL-HDBK-217B broken down as follows:
nonessential f ailures (do not affect trip output);
A
=
A
"***I B
essential n annunciated failures; and A
=
C t tal hardware failure race.
l A
=
T Table 3-6 FAII.URE RATES (A) AS COMPUTED PER MIL-HDBK-217B FOR 35 j
4 i
A A
Hardware A
B C
T Master Trip Unit 31.06 5.56 19.50 56.12
- Slave Trip Unit 20.61 11.52 3.0.96 63.09 N/A N/A 1.0 1.0 l
Trip Relay
- Includes series failure affect of master analog output to slave.
As explained in Subsection 3.4.3, each channel A required for determining over-the trip unit and A f the trip all system test interval 0 is the sum of A C
C 19.50 + 1.0 = 20.50 failures Thus, for a master trip output, A master =
relay.
= 30.96 + 1.0 = 31.96 fail-per 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br />, and for a slave trip output, A,1,y, 6
ures per 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br />.
s>
d' t
3-24 1
1
-m<
NEDO-21617-A The equation (3-2) for (1/2) x 2 logic systems for both simultaneous and per-factly staggered test intervals, was plotted for various values of availability When implementing the above values of A on the log-log graph of Figure 3-12.
to the curves, it can be seen that an availability of better than 0.999
~ ) is easily maintained with a monthly test interval for (unavailability = 10
~
If exceptional availability of 0.9999 (E = 10
) is both masters and slaves.
desired, the curve suggests masters may be tested " perfectly staggered" (one channel out of four per week) every month and the slave tests staggered over approximately three weeks.
It is concluded from this analysis that adequate availability is maintained by scheduling a regular monthly test interval (perfectly staggered) for master and It is also believed that life experience data will show that slave trip units.
MIL-HDBK-217B is conservative by' a factor of at least one decade on the log The failure rate 1 of the principal contributors, the operational plot.
amplifiers and R17 is expected to be much love'r than the calculations predict.
For example, if the vendor data for R17, alone, were substituted into the Rose-the 1 f Table 3-6 would be decreased mount analysis for the master trip unit, C
This translates to an availability improvement of about one-half by 34%.
decade in the system eq'uations.
e 4
1 e
3-25 1
P t
NEDO-28617-A 3-
\\
'l j
1 3
'o I
I I
1 "S
8 a"
i S
p 7
~
m I
O.
E
~
4
$' 4
<!~
i.
-l 2
4 k.
h 3 j a
E g
. l.
~
D.
5 l
~
lV.
j f
f I
f I f I
i 1 1 1
l l l
1
~
't "S
"S S
]
b' l
i 3-26 l
_a
_ QUALIFICATION TESTS 4.
The vendors of each component in the _ analog chain were required to qualify a For plants in prototype sample in accordance with the purchase specification.
Phase A (as defined in Subsection 1.4), the only instrument mounted within the
[
The inverter, power supply, high stress containment area is the transmitter.
f trip units, and trip relays are all to be mounted in control room cabinets.
j However, if plants in Phase C elect to purchase the retrofit option, they may '
desire to install the trip unit cabinets near the location of the instrument racks housing the transmitters. For this reason, the trip units were qualified
~
These
. l in accordsuce with BWR/4 abno.rmal condition outside primary containment.
conditions, given in Table 5.(sheet 7) of Reference 1. match the requirement l
The inverter, power supply'and trip relays have been of GE specifications.
qualified for only control room environment at thu present time; however, l
since the inverter remains in the control room for all applications, only the l
relays and power supplies require the additional higher stress certification 0
l if the trip unit cabinets are. installed adjacent to the instrument racks. The
~
(Qualifi-higher stress for the power supply and relay is shown in Table 4-1.
cation testing will include testing the maximum stress levels plus appropriate I
margins stated in IEEE-323-1974.):
l Table 4-1 HIGH STRESS REQUIREMENTS FOR THE POWER SUPPLY AND TRIP RELAY Temperature 40*F Minimum 70 to 104'F Normal.
j 156*F Maximum Humidity 20% RH 40 to 90% RH Normal 00% RH Maximuss l
Radiation Dose Rate (R/hr)
-3 15 x 10 Normal 2
6.5 x 10 Maximum Exposure Integrate.d (R) 3 5.3 x 10 5
Normal 1.7 x 10 Maximum 4-1 I
l
- .
- a-e s==ei r ves a-er wa a w * ** * *- :
_..-e.
.. I wm c 7 =. _
NEDO-21617-A 4.1 ENVIRONMENT The term " environment" is defined to include projected temperature, humidity Those which may be locally mounted and pressure extremes for all components.
also include radiation exposure as applicable (Table 4-2).
4.2, SEISMIC AND FRAGILITY The seismic test requires the sample part under test to function and remain The functional' throughout the limits of the specification during the test.
g-level specifications are primarily determined by the " gross seismic limit",
which is defined as the accumulated effect censed by the device at its final This accounts for the acceleration amplification from the mounting fixture.
ref erence point to the panel location and the transmissibility of the panel to the device.
All seismic tests were done in accordance with IEEE-344-1975, testing three Lower g levels were used to find resonant frequencies", 'if
}
axes independently.
they existed, then resonant points were driven to-the specified g level for If no resonance was found, dwell was usually done dwell times of 30 to 35 sec.
The fragility test is a nonoperational test conducted to deter-at 30 to 33 Hz.
mine if the unit is still functional af ter exposure to excessive g levels which (Table 4-3).
The may be several times higher than those of the seismic test sample did not have to be vibrated to the specified fragility level, but only until a nondestructive malfunction was observed.
If no malfunction was l
observed, the test continued
- to the specified limit.
4.3 POWER SUPPLY REGULATION While the various components were being qualification tested, the input power supply voltages were continually varied over the range required by the purchase This was done to assure that the components would function l
specif ications.
acceptably during adverse power supply conditions concurrent with abnormal supply variation limits of each series element was specified with conservative N
overlap. For example, the power supply is required to produce nominal output the inverter supplying the with input voltage variation of 102 to 127 Vac, but 4-2
. A A
t 1
/
/
/
se x
x N
N N
- T 9
n 5
o 1
i ta i
5 5
S d
0 0
1 N
a 1
O R
I c
x x
R R
R e
T e
I p
7 7
0 0
0 r
5 5
5 o
D S
1 1
3 3
3 m
)
N e
O 2
3 C
h
(
c t
E R
r U
o o
g f
S i
t s
s s
s s
O K
s p
o o
o o
d o
P E
e m
m m
m e
eT 0
t t
t t
i L
r 6
A A
A A
f a
i A
u
~
l C
C C
C a
M s
O e
g W
W W
W u
R s
r i
q o
NB P c s
A e
p n
n n
n e
p i
i i
i re 1
7 1
1 1
w L
S 5
A T
t N
u b
E )T t
m 0 S s
a
- 0 E ye e
9 0
0 0
2 0
t T t
9 9
9 9
6
/
4 R T I
L i
S R
e V S d
W
- ~
n l N V l
B a
m b E m c a
T M E H e e
0 0
0 0
r a
C i
U P p
t 0
9 9
9 o
f 1
M S S
S m
I
(
l, X
o A
e t M
r s 3
1 2
5 0
o u e 0
7 3
3 2
r G
t T 3
1 1
1 1
'l s N
a}
ot I
r rn O
Pb 2
8 0'
0 0
t e W
e H
e e 1
4 2
2 2
nm 1
1 oe S
e p 2
1 1
cr S
i Y
eu t
t hq R
M n
n m
m m
t e A
e e
o o
o r
U n
m em o
o o
n M
S o
n d n R
R R
it n in i
ii N
t o ea sa l
l l
d e O
ai dt t t o
o o
em I
ct i n un r
r r
sn T
i a so O o t
t t
ui A
f c t C C
n n
n oa C
io u
4 o
o o
ht I
lL Oy
/ y C
C C
n eo F
a r
R r rc I
u 6 a W a 6
6 6
L Q
/ m B m
/
/
/
a A
Ri i
R R
R 4
U Wr r
W W
W s/
Q BP P
B E
B tR iWmB r
l t
y e
t t
. p y
pn e
s p
a ie c
t i
u r
l rg i
i n
S e
e t n t
R i
v m
U er r
r D
n p
e e
p S
ht e
s a
i w
v i
E T s r
r o
n r
T
)
T T
P I
T O
1 N
(
mL l
l[
00
,,!*.FI
]
115 Vac tS% (109.25 to 120.75 Vac). Therefore, the power supply met ou tpu t power supply output should be unaf fected even with a drastic out-of-spec abnor-This concept was expanded to include all the components mality of the inverter.
illustrated in Figure 4-1.
3 18 127 2s
}
~ 12o.M 26 5 137.5 125 VOC bus I
Y INVERTER
?
POWER SUPPLY rRIP UNITS 109.25 23 5#
112.5 d
ab 16 L
n 22 i6 102 1o0
,[+10.4
+16.7%
+ 10.M
+12%
l j-2pm l-11.3%
l-2.0%
l-4.3%
Figure 4-1.
Voltage Regulation Conservatism of Power Chain Tabla 4-3 QUALIFICATION
SUMMARY
SHOWING MAXIMUM SEISMIC AND FRAGILITY EXPOSURE CONDITIONS (SPEC VS TEST)
Seismic g'c Frequency (Hz)
Fragility g's Device Spec Tes t Spec Test Spec Test 5-100(I)
None N/A
~
Transmitter 3.0 3.0 3-30 5-33( )
20.0 15.0(2)
Trip Units 10.0 11.0 1-33 8.5( }
Power Supply i..
5.5 1-33 1-33 10.0 I)
Inverter 5.0 5.3 1-33 5-33(l) 10.0 9.0 17.0(2) 5-50( }
17.0 Trip Re)r.y
$.7 6.7( }
1-33 17.0 17.0 N'JTES :
(1) Low frequency limited by capacity of test machine.
( A nondestructive f ailure occurred at this point, which defines the fragility l
limit of the device.
(3)The 6.7 g's is with relay de-energized and 17 g's applies to an energized J
,s relay.
4-4 i
. - -. =
---.----,m__--
v NEDO-21617-A 1
4.4 ELECTROMAGNETIC INTERFERENCE (EMI)
\\
The Electromagnetic Interference (EMI) susceptibility tests for instrumentation 1
were established as a result of worst-case transient and Radio Frequency (RF) conditions measured in actual field tests. These EMI measurements were con-ducted in and around nuclear reactor control rooms and were the consequence of operating various EMI generating sources. Typical items associated with a
. nuclear power art:1on which generate EMI include many inductive components as well as inoustrial electronic and electrical devices.
D 4.4.1 EMI Transients O
EMI transients typically are impressed on 110-Vac power lines by de-energizing-g an inductive load (i.e., relay, solenoid, electric motor) and are generally a D
100 to 500 kHz damped oscillatory wave of six to seven cycles with a 300V maximum peak-to-peak suplitude and a characteristic impedance of 150 ohns.
- p 4.4.2 Radio Frequency EMI lI lt Radio Frequesey EN! produced by arcing contacts, fluorescent lights, SCR con-l o trolled circuits, etc., is generally between 500 kHz and 100 Wz sine wave that amplitude or frequency modulated or a combination thereof, which I
is continuous, can produce 5 maximum peak-to-peak amplitude of SV and currents up to 100 mA.
Inductive or capacitive coupled EMI from radiated electromagnetic fields are g
limited only to near-fields because the distance from the interfering source is usually less than 1/28, where 1 is the wavelength of the interference signal.
g in The following type of D(I susceptibility tests were conducted on each unit t
the tandem arrangement:
Conducted EMI transients, 100 to 500 kHz, 300 Vac peak-to-peak or 8
(1) i
!5.0V (24 Vde)
(2) Conducted RF EMI, 0.5 to 100 MHz, SV peak-to-peak Radiated transient Dil fields,100 to 500 kHz, 300 Vac or 25.0V (3)
(24 Vde)
U'* *
(4) Radiated RF DiI Fields, 0.5 to 100 mz, SV peak-to-peak
(
4-5 i
C
_--~
NED0-21617-A Table 4-4 shows the highest amplitude (in Vac peak-to-peak) and f requency (Hz)
')
for which the unit successfully performed without adverse effects.
The notes j
l clarify justifications for those which were less than the specification l
i l
required.
l 4.5 MARGIN l
l
~
Section 6.3.1.5, specified the following margin extensions to be IEEE-323-1974, placed on existing specification:
Temperature:
+15'F Pressure:
+10% of gauge
+10% (on accident dose)
Radiation:
Voltage:
110% of rated value i
25% of rated value Frequency:
l
+10% of period of time the equipment is required to be,
)
Time:
operational following the design basis event.
i 1
i Environmental initial transient and dwell at peak temperature shall be Transients:
applied at least twice.
i
+10% g level at mounting point.
l Vibration:
f Table 4-5 shows that margin tests have been performed for almost all applicable 1
parameters on all components.within the tandem chain.
)
s l
l t
s.) ]
4-6 I
i
Me E u z
o o
o o i.
eo q
H t o t
t t
0 d t o
F e
(
t 8 F0 u
w 5 5 R
d t
Rr
- 5. 0 5 t
A a
1 i e F
d 08 0 0 0 /
e
,1 c
"s l
N r n p u e
a on md e
t e i a a e e
ad o
i ust) eth r
s d t t k
h i t F
t d R e ailka Rl oae d e b
p pV eP 53 5 5 5 n ar f
o i
d m( P o
o o
nm l
l.
A eI u
d t
0 A
o M
M M M u
.u 6 w mt y
5 5
5 5 t
p idt y
0i c
t n
0 0
0 0 l nu l
s 2
p pao a
e
)
I u
z o
o o o mb p
o e u sq H
t t
t t
a t
l t
Mt rc s
e
(
A k oe t
4 1
1 1
1 nr
/
arf r
e a e d eF ei 0
0 0 0 N erf pea s w
hh S
t s n
e o t
R ane
'o o%t g t
p T
d rust) t i
a E
i ad 1 o u e a
t s
E aTt t k
0 0
0 0 n
M R
ilka 0
0 0 0 ke e
nh s A
l oae 3
3 3 3 ahl d et y h t
a t
etl p
t t l R
pV eP p
i nw o aa e ob A
m( P r sa P
A Vi o
l s d n S hk p l ep s T
M M t a S
MM MMM 0 0 hie e o u ei d T
c 00 050 0 0 t wph o o s r re E
y t l w
t a
I n 80 680 1
1 i
c wd o s
4 M
E u z
o oo o o et n r t ad I
Me
)
1 1
'W 4
i e i n
s E
q H
t o t t o t
t
%b -
rk t
l a th 0
A 1 oad t
ea i
1 e
F e
(
t t
- 5. 0 5.6 5 5 5 n
e i rn l L R r
/
s e~t m ugs ue b U F
a F d
08 0 8 0 0 N fb p
T S e
oa n s sir V e n sse p
S t e yyS m a a t
i E
cd o
cl e r F t r
C d t t k
55 ald l t
oRi t
C uust) t m
runp S
ol oae 53 532 5 S uf am l
- V s e U
nilka cs i l dS n h y x
C pV eP cs) a e-a t
T m( P aez e t ar S
A cH r r s t
n h
i E
ncM e o e
, h M
M M M i u w
f t s H
us t
as4 G
I y
5 5
5 5 s
t ir aniwe t
7 s d m
H cn 0
0 0 0 ns1 n e ia o i w s
a d
l e
)
u z
o o
o o awoi f
.t H
s t
t t
m t t i y msq
.e u c l sp n A
td2 a e pl i ge
. m_
t e
(
1 1
1
. /
o u3 c p par ih d nr 1
t i 0
0 0 0 N nt1 e s unt eeF e
i( r si cs dl p s me d une l pe i rrh d ad o
umgl eet ee oTt t k
0 0
0 0 oana e wt nrust)
C ilka 0
1 0 0 c
an l o
r V ro b pt o we l oae 3
3 3 s3 i a uf t
c cp pV eP r
,"e i
d n) m( P e
A t rid d Vi m en t ekd e u
i vl a d 4 sm mea l
2 t i ie y
swc g e i x no-n i e a
e p
y aHei h h nm Fi r
l e
t t
p a
r i w S To( Rf c
t i
u r l t
k o i
i n
S e e
.l l
ez al
)
)
v m
U t R a b h H "w o r r S
TM f
(
(
e s
D n
p e e p E
a i
w v i T
mb r
r o n r O
T T
P I T N
_ _w i
l[!! i '
{
7 13 h
l l
ai t ei rt 3 na acl zp l t fi w
o ri He p
n ogih bht i
Mc re mu l nt c iwa x
ec e
iau l
i 5 e ws h p yvrm asf 8
o u t i rib cro s
ps r
eris e
h oi ot vdl a at s.
t h t s t
a l rs T
rs ne g) ce di ot 0
oe ih nn v
nf t s 6
hl t
i ni a
ce s
g t 8 er Cat rt h
no a( hd sLf of ec i t e
W t
h fi h u m
rA d
it nc h
T m on c3 l
nuon u
u pi e V s co e
,t so nt b t
. r Id l ut c pi a 5
3 n sa e
eoi i
ue (Ds nb ns rsnh d
oo ns
- s aaue t aet
- i e
)
t p il ya p
i ht l e np s pl e
t o e
n d p n
a eiip et at u
ei sa I
h rp ri t
xr up tM tdt u h n1d i
n at b
uE a
s t u( eDr l
e o
no0 C
ee ll d
il7 r gp,a o
p rl eg re h
e ni b
nn et t mgw i rsm
(
S ea ai.
t c i uno nt uo l
R bw pss l u wmip i
hc s i E
o r
e id iv ahT p
T ol end f n ynie t c s o E
tl ti u o
l i rh na a o dd M
a nht oc pmdt oe
. s
,d A
d iti t
p c
at r aa R
an in t
ul),
A ha h wg di saDd enon e i c 4 P p
t a em c
e l ol e t
t s i sm ii ri4d ii s
t i u
4 T s
we fl ep(d f t et i ne S
ef ll i wy a
ih r m ih e E t o nia cy otA dd t t s
l T
of n el p
6 e rd s n b
F%
i g pi A
r aaon a et a
I R6 tdi ss e
ra c
ti r
T M 0
ars a
l e
n t
vr eIdd E
s1 ca ee g.
t.
i ch r nsoe n
ee l L
i0 l
g ed i e w
o t t l
nn a e pdi wl snno F
ues pnh u
iip yAeu U
n S
h n aa so el l3 so r S
pt o e eh h
d y n
em o v E
i s rye vs t raa o1 r -
f C
rna ol s l
eol pd C
ti e f p es mwl e nt
.l d
U ar po si oo r
out e e me rpm oii l
S et d ut mh hng es eT e f
ud eb n f.
f t in d y h c
V mn d au i
S ai nrl t
r d - ia a
n c T
mf pi e
E rmw eee
.u e4 x H
o o
t wk s ro r2 asy oi p
f ol noi ees e mel e rr s tl i pl i t we ie rd t u G
l rr ohll v ea c e d I
o rf sns pee pt abi woh c b ii H
n o
i e aes pvw cmt l c o td e wee uno eriec s a
l arh wl sip r opse t l ed l
af ysp sa l
i cot nt i
oee rrV t as et h u w f nf ibb eo-ss e
t ot o i io t
w 4
tdAt r ti h e
c a) d os2 i a u
t ws l
eze c gl pe no o, ia b
)
pH s i nu ne ul
.dA n
l s a d sMu f oo cih sa5 egee c r e
a il w dl t pet e3 nl m u e0 c c V
i cir pil i d d n d0 e et d
- ch rnn d i t at e n rnr d
u1 b pf n 4 ag t a u,
i t
t s
a 2V u d
sa t eae l
n i od 0
o yept spe e o l t e e2l e0 r npi iA ee r i amr n2 h r nh h h p C p i
h< mh2h m5f T(
T1 t Mit u1 T pi t S
(
a8i t
)
ers
)
)
)
d S hou a b
c
(
E Tf j
(
(
(
TO}
N(
aE
__.gli i
l
c d
.n e
R e
V
.I0~ #n iC n A 0
n A 0 0 2 2
- W o /
5 o /
4 07 43 1
N N 3
N N 2
22
- +
0 1
1 A
A 37 R"
3
/
/
21 aO2
_+
N N
1 c
d V
n e
R e
- ! A 0 5 iC n A 0
n A 5
6 91 2 3 5 W o /
5 o /
2 64 22 1
91
- +
N N 3
N N I
1 1
1 a"
A A
2
/
2
/
'O N
1 a"N N
1 ca V
A$a
.[~
n e
R e
0 2 i C n A 0
n 'A 5
0 5 2 3 2
- W o /
5 o /
1 04 2 1
1 1
1 N N 3
N N 1
91 1
DR 88 A
A 1 2 O
R" 2
6
/
B N
1 E
3#2 1
5 R
- O 4 C c
S dV 88 28 e
n e
l T
9 2 E*8 #n 8
1 iC n A 7
4 90 46 b S a E 4 7 3
- W o /
1 1
0 2
1 3 1
1 2
7 N N T T N
I 1
G R"
0 4
3 0
8 0
R a#2 4
3 2
2 A
M g
g g c
c i
i i d) d) s s s p
p p 3
V n V n i
i 7
uOMeSn 2 3 0 8 5 m 2 m 5
0 5 2 3 9 1
6 4 1
5 4 1 (
1 (
3 1
0 1
)
R
)
)
0 F F 1
t
)
(
(
x R s
(
e c
t T
e s
)
t 0 p e F S T
- c t
s 1
g e
n
(
e s
i e e p
e T x
)
(
c w
r r n S
T u u i n
1 n
e S
t t
g e
e :
o (
o n p
a a r r
r n i
)
i i S
e r r a u
u i t
R t
g g
n e e M s
s g
a a r e
a i
p p s
s r i 5 i a g
t m m e e
e a d 0 d M a
l 1
a t
o e e r r
r M a
T T u P
P R
R n l
V M
t e
x o
o t
t a
t t
r t
(
t i V
m s s r s
s u s
s t
u e e e e
e s e c e a d
m h h p
h h s h e h i e
i g g m g
g e g p g d t
x l
i i e i
i r i S i a a
a H H T H
H P H
H R R
M
~
- i ;{f$j i; !ii.Ill l
i
.,l
- , i' !l !IL : js
iy$$[
i l
A A
s
-q_
R-A
/
/
8 r
i
/
1 t
3#a N
N N
t t
i6 m/
sR nW e4* &e e
A A
n A
A A
aB t
/
/
o
/
/
/ 7 0 r
n N
N N
N N
N 1
2 3 tf e
o m
w p
en i
A A
No u
A
/
/
0 i
q 3Oe R
1 t
e
/N N
N
.a Rc m
i o
f o
0i r
1 c e
e l
4%E3 A
A n
A A
A 5 5
'x p o
/
/ o
/
/
/
N N N N
N N
5 5 0 s
r 2
t w
n f e o
A A
on c
/
/
0 O-
'e s r
5 1
N N
si o
3#a oh f
1 dt d
nd e
oe r
)
i e i
d
$3u[~
e e
n A
A A
5 5 t c u
u 73 o
/
/
/
n 56 3 N
N N
N 5 5 0 ax q
ie e
d r
i ar t
ro
't n
A 03 o
a-04 K
0 o
3#a N
19 O
1 rt n
/
C oe
(
f e e
m r
D s
a 5 R
- A F
F
- lo r*
r*
rF lt A
h8 h2 h *4 a
C 4 OB cn O
e E 01* &e A
A 4
1 b O N
N 1 9 29 19 2
1 1
1 o
L l R
/
/
21 42 26 0
1 ni ot r
a C i a o
t c f
T S ai T
S g-A
%3 cf s
74 K
E 1 @
O 0
ii t
/
fl n
T 3pa N
ia e
cu i
N eq s
I F
F p
n C
F sd a
R A
r2 r3 r*
n r
M k3 gg:
A A
h1 h0 h1 l a t
/
/
2 3
9 3 3 0 a
N N
69 79 I 9 2
t n l
ng a
ei t
ms n
ne e
od m
s r
n n
)
)
t a
s s ig o
vn r
s r
e c
t
'g 'g ni i
~
T e
s T
eo v
(
(
p e
g n
l
~c t
)
er e
g S
T a
n t
e s s he td d
i n
n w
o o
n p e
'g n
n i
i e: S T
(
ou a
S t
t md
- a a
ne n n t
n c
n r r
om o o n y,
y n
i u u
rr i i i nl s o e
g D D
i o t t
g ot n i vf a a r ino t
q a e e
n r r r a sei a u
r ist r e
M m m
n E e b b M vea u i
i i
P i i F
y T T
g f
V V n erc d r
rpi r
o s o
l e
c m
n m m
a t
m m i wep m u
e u u
M r s u u t m
u m m
e e m m a S erp i i
q i i
e b T.
i i r E Naa T x
e x x
m m
x x b T
a r a a
i u
a a i O
(
(
M F M M
T N
M M V N
yg i!
1; t'i f
/lff
.1Ifi j 1-
p NEDO-21617-A j!
l
' I 5.
APPLICATION s' }
t il 5.1 OPERATING POWER PLANTS (PHASE C)
)
All BWR plants in operation use mechanical switches for sensor trip input.to f
engineered safeguards logic. The retrofit package suggests changing out a
, minimum of 90 of these switches in the RPS and ECCS systems in accordance with i
the variables. shown in Table 5-1.
Implementation of the new package involves removal of the existing pressure or temperature switches and replacing them g
with transmitter trip unit combinations.
l 5.1.1 Trip Unit Cabinet Assembly 1
The trip unit cabinets are modular panels prefabricated according to the
]
The basic panel dimensions and cutouts will i
requirements of each customer.
be standardized (Figure 5-1). Maximum capacity of each panel is 3 card files containing up to 12 trip unit cards each, 2 power supplies and 36 relays.
l Blank fronts may be substituted for those requiring less than 3 card files.
i
,)
J A
( l; pINTERFACE u
i RELAYSPACE l
l g
I TERMINAL BOARD ilit MAXIMUM) l f
J 4
I h
CARD FILE !
0 l
E p
CARD FILE l
P l CARD FILE L l
~
90 in, t OR 5pEl l
l )
l aa sn.
/a TRIP RELAY fPOWER SPACE
[
POWER SUPPLIES l(36 MAXIMUM) g j
hh I
i 36 in, y
y h 30 in.
9 Figure 5-1.
Standard Retrofit Panel 5-1
v F
)
Table 5-1 RECOMMENDED RETROFIT APPLICATION Number of Devices Transmitter Variable System _
Type N o._ _
Master Slave RPS dPT 4
4 0
Steamline Low Pressure RPS PT.
4 4
0 Low Condenser Vacuum RPS/RSCS PT 4
4 3
First Stage Turbine Pressure RPS PT 4'
4 0
Drywell Pressure
'RPS dPT 4
4 0
4 0
Vessel Level Isolation RPS PT 4
4 8
Vessel Pressure Scram / Isolation RPS dPT 16 16 0
Main Steamline Flow RPS' TE 18 18*
O Steam Tunnel Temperature Switches ECCS dPT 2
2 0
Vessel Level ADS Confirm ECCS dPT 4
4 8
Vessel Level ECCS Initial ECCS PT 4
4 0
Vessel Pressure ECCS dPT 2
2 0
Vessel Level Containment Spray ECCS dPT 2
2 0
RCIC Steamline Isolation (Flow)
RCIC Steamline Isolation (Pressure)
.PT 4
4 0
HPCI Steamline Isolation (Flow)
ECCS dPT 2
2 0
1 1
HPCI Steamline Isolation (Pressure)
4 0
FCCS PT 4
4 0
Drywell Pressure 1
l
- This trip unit has the RTD bridge at the input.
Control room environment is preferred for trip unit panels, if possible.
However, they are qualified for use in transmitter or other location if NOTE:
desired.
I' l
l
)
~
i 5-2 t
4
ti The components are arranged within the panel in accordance with their qualified f*
Since the card seismic g levels and convenience of operation and maintenance.
files contain the only external control and indicating functions, they are placed near eye level. The power supplies are mounted near the botcom of the The relays for the trip output are mounted within the lower half of cabinet.
Interf ace relays for the annunciator and other logic are l,
the enclosurs:.
. mounted near the top of the cabinet.
The power supplies and relays are not visible externally.
l 1
These annunciator interf ace relays are those which have divisional power at their coils f rom the card files but whose contacts are tied to the nondivisional l
annunciator or computer, such as the " trouble monitors" described in. Subsection' The Class 1E coil wiring for these relays will be tied into bundles flush 3.2.
with the inside wall like the rest of the Class 1E wiring for the card files.
l However, the nondivisional contact wiring goes directly back trip relays, etc.
f rom the relay terminal board in a direction perpendicular to the panel face, f
There, it is bundled separately on offset bar or strut and routed via separate raceway through fuses to the annunciator (Figure 5-2).
N, FUSES NON.lE WIRING
~_
=
/
O s
1
/
p p
IE TO NON lE INTERF ACE RELAYS y
/
Y FOR'NOklE WIRING i
1 5
7
~
=
,,e s
CLASS IE
/
i A
2
[
WIRING H-cf
_4 2-7 l
/
/
4 O
Perpendicular Separation Methnd frr Divisional Figure 5-2.
and Nordivisional Wire
{
5-3
v NEDO-21617-A Design for Reactor Protection System (RPS)
)
5.1.2 Trip Unit Cabinet The RFS system is a two-division system (A and B), where each division has to have four separate enclosures redundancy. This produces the requirement The four separate chan-to provide separation between the redundant hardware.
nels are identified as A1, A2, B1 and B2.
The hardware for each channel is However, two channels (A1 and B1) may be housed housed in separate cabinets.
in a single cabinet 1.f suitable barriers are provided to maintain separation The hardware housed in each trip unit cabinet are:
trip between the channels.
units, trip relays, annunciator' interf ace relays, and de power supplies.
)
Trip Unit Cabinet Design.for Emergency Core Cooling System (ECCS) 5.1.3 l
l f
The ECCS system is a two-division system with hardware mounted in two separate Each cabinet houses trip enclosures to provide separation between divisions.
E units and relays for the ECCS system and trip units and relays for Anticipated Transient Without Scram (ATWS) logic.
The h rdware housed in each trip unit trip units, trip relays, annunciator interf ace relays, ATWS inter-cabinet are:
face relays and de power supplies. The de to ac inverter is mounted in an existing ECCS logic cabinet in the control room.
Location and Interconnection of Trip Unit Cabinet and Existing Cabinets 5.1.4 The auxiliary room or control room environment is preferred for all new hardware except the transmitters for the following reasons:
Analog resdouts and calibration / test controls are more readily avail-f (1) able to operating personnel.
f
]
)
MTBF/FMEA analysis shows component f ailure rates roughly double as (2) i temperature doubles. Thus, equipment life expectancy and reliability are increased when installed within the cooler environment.
Existing power supplies and trip relays may be used "as is" without (3) requalifying and possibly modifying design to comply with high tem-
.s perature and radiation environment (see Section 4).
l l
5-4 i
-A
m NEDO-21617-A:
(4). Fewer panels are required because RPS and ECCS divisional systems can be consolidated.
However, it is recognized that control room space is limited 'if not nonexistent
~
Furthermore, since shielded cable is required between the in some plants.
transmitter and trip unit, the existing nonshielded conductors for the-pressure These factors switches have to be replaced with twisted pair shielded cable.
usually make it more convenient to install the new equipment locally near the Other location options may be an auxiliary relay room, cable transmit ters.
If it is necessary -
etc., as desired by the individual customer.
spreading room, to install the panels locally in the reactor. building, they can be arranged as Here the-cabinets are,
shown in Figure 5-3 for the suggested 90-loop changeout.
mounted close to the instrument racks so that existing pressure switch cables-from the instrument rack to the existing logic-cabinet in the control room can New raceways are required between the trip unit cabinet be used (Figure 5-4).
and the local instrument rack, but these are relatively short and may be in the form of finxible conduit.
N One of the major advantages in mounting the cabinets in the control-room or an auxiliary room is that fewer of them are. required for a given number of loops.
When locally mounted, the number of trip units and associated hardware within a panel is limited by the number of transmitters on the local rack'it serves However, when all trip units are placed at a central location, (Figure 5-3).
It is estimated they can be consolidated divisionally within' system panels.
the number of required enclosures can be that for the 90-loop changeout, Interconnections l
reduced from 10 to 6 when a common cabinet location is used.-
near the local instrument cabinets may be handled as shown in Figure 5-4 using the original junction boxes and raceways but. pulling in new twisted It is also.possible shielded pair wire as required for the transmitter loops.
to use only six cabinets mounted locally in the reactor building, if extensive Interconnections use of conduit is made for the transmitter to trip unit wiring.
made in the control room or auxiliary equipment room may be handled as shown in Figure 5-5.
5-5
'i
.l
. ~.-:~n -w:.mm :,-r wyy~. r vs.~, n.a.,, -..- a won..v en... wi r v... n...
m(g>
3C T
A C c I
v L
0 A '
Ch L
S S
A
- "u R
v A
Y N
R e
N A
E 5
f E
G T
2 N
G T
w e
T 1
A t
OE S
I Y
S E
E L
TC G M G
LA O
S L
L P
O AR L
N G
S P
T U
L A
A us UTi S A
N G Ru RuR t
NOR D
Oh Oh l
NO GNI ATT T
N T
E AT S
T P rW e
W I
I E
=
PS
=
A I Sa u G
R RO t,
RM E
ST S
TP A
T E L
P P
]
R-
' R - -
t S
S P
- k R
P pB, R
i' A
s
)
S s
I l
n P
e d
R R
S o
e E
G O'
N E
i N
Ne t
O 4
t R O L
t8 O a
4 L
E Re L a AL h
l uu O
t
=
Ts l
I u
Y E
F
=
a IT T
O L
Sl
=
,J m t
s P
N E E
~
t O R L P
l s
',4 s
U C E S n
T ut iS W Ut Ose
~
g 8
I 1-m' O Os RuR e
O RU E
T P s l
e T P i u p
G a
e 2 RO t
s t
=
c s vT VTP
[r W
R
=
OM c
O-I
= s S
t
~
O-TO o
CL Ai L
t8 Eu
(
O l
Re l
TA t
2 t
SC Vt v
V l
A K
K n
N O
O C
C e
U W
S i
A A
fe C
C R
R 9
m N
S S
2 e
C C
8 W
A W
g A
E E
1 O
O e
U 2
n A
L L
S a
I F
F P
r S
M M
R P
A i
1 r
A R
E A
Ry E
T T
G g
S S
S N
e 3
i' t
l PR' i-T e
7 l
S b
t g
l a
.\\
x g
E C
g 2
d V
SC n
s O
CA a
C T h
S E
C N
I (2
E t
C L
E A M aE I
h e
(2 n
SS b
SN a
EI C
s R
II t
M Y
E l
i O
R T
l n
R E g O
.I U
T E
R T
8 E
S I
[
V l
N B
2 T
A p
t.
[
A t
N K
K 2
i I
O Y
A C
j g
C S
r S
T A
S A
g A
P T
A S
R R
R C
P C
R T
T l
I 1
N N
G G
A G
E R
E N
L O
V R
NI u
e SC M
O M
I 3
E T
TS u
L O
CA T
CS R
I 5
S Y
S C
S R
AE T
X A
C ET t
T
_L EV S
E l
e N
x S
L C
L L
R E
.l E
N I
L W
IN E
E AIM I
E 2
r R
l L
V u
Tu E
E l
/_
NR IE V
g ET t
S S
O S
S S
i S
O E
S E
S S
N_
R V
C F
t S
V S
E R
L C
C L E
]
E E
C AT y
AT
- 5 E
II S
R NT A
NT G
T I
W GI y
E g
G GI N
l I M N
R l
h R
Y t
E T
E A
T GN LA T
V S
O A O A tx T
AR LR u
A N
v IM LR N G AT E
S
=
I E
AT GNI NO L
O A -
NO i
AT i
R r AT ST T
2 PS g
I I I
lv R'
l RX v
TE O
o O
S S
C P
^
S p
C C
R' C
E C
E pbt Il
[
lillll\\
j
(,--__-____-____ _ _ _ _ _ _ _
NEDO-21617-A f~ )
~@
i I
11 il 15 2
N u
b Y
N 5
o o
o o
g i
=
l o
u 1
w g
.gg i.
33 f
h "t'
\\
5; 50 a
2 3
"A l
g i
E oooo 3-5 y
=3
=
g 5.3 3
t;
=
.f g
, -)s y
c.4 O3
~
w i
)
x p(
0
)
i>
at e
g a5 ga Se gg
=
y v.
Eg 000Y
!gGB! If fE
.u.
el$
I! l
()
~
Es
!j j
lk mg r
ga 8
Jh
!l
/
I
.u
/
' n.
sa 1
l WL o
dBj l
N
/_
I!I l i i
i f
{
}
E
\\
b/
/
fia f
[
-e e
x E
35 U
W o
s o w
g g
e$
f O
ol 0
5-7
\\
i l
l
11
!Ii Il ll M /o;?-
4 IG y
)
G y
N C g
I I OTG S
TiOXL E
M YA O
O L
E R
lI; R
L L
O A
R N
TN GI O
t n
)
I[lu vPI I
S C
i o N
ni R
Ut T
T a
EN pl T
il l
I t
ra N
A Tt U
C
/ s P
64G I
T rn I
4 R
N eI 1
T U
t "J
t R
t m P
io 1
E IR mo TS T
sR A
n M
al ro Tr t
wn eo u
N (C n
D et E
en D
we L
t m
-)
Y E
A ep I
W H
Bi
'u E
S u
C D
nq A
E oE R
TE i
S t g R
G I
cn N
WlW ei I
T T
R nt S
WI ns i
E A oi X
NP cx E
rE e
X
.\\
td nn I a OB N
Y 5
L O
wi L
I 5
ER A
T s>
C RE C
~
UT O
e N
~,,
ST L
r U
J f
t G
RS O N g
N PN E
i T
W A TM F
I E
U S
R N
E IX N T I R BT E
S AS E
CN R
\\
I I
T D l
D W
\\
J N
/
E E
E E T V
T MN O
U RU b
TM UU M
C SH RO E
SC R
ET S
RI N
E P W I
B S
D L
O L
A T
O C
/
O L
~I V
I l'
i; 5.2 EARLY DESIGN PLANTS (PHASE B)
Plants which are under construction but not yet licensed for operation include l
BWR/4, 5 and 6' designs. Retrofit application for these plants is dependent on the appropriate design specification.
For BWR/4 plants, the application 'is New hard-identical to that 'given for' operating power plants in Subsection 5.1.
ware may be installed-at local racks, control room or other auxiliary rooms as desired, because the trip units were qualified for the BWR/4 local rack environ-Panel location options for BWR/5 and BWR/6 plants are limited to those ment.
areas where the environmental specification parameters do not exceed those of j
the BWR/4 abnormal condition outside primary containment (.i.e.',156*F, 991 1
1 5
humidity and 1.7 x 10 R integrated dose).
5.3 LATE DESIGN PLANTS (FHASE A) j The analog transmitter / trip unit sensing feature is incorporated in the basic 1
standard plant design released to manufacturing for all Phase A (DWR/6) appli-Transmitters are mounted locally and trip units, power supplies, cations.
inverters and trip relays are placed within the control room. - For the Reactor Protection System, all of the required control room hardware is housed in the In the Emergency Core Cooling Systems, each system contains its
.l RPS panels.
own trip relays, but the trip units, power supplies and inverters are centrally For located in a " boss" system panel designated according to power division.
example, in nonsolid-state BWR/6 plants, the " boss" system for division 1 is -
Each of the coils of all ECCS LPCS; division 2, RER; and division 3. HPCS.
trip relays is connected to trip units housed in one of these " boss" panels.
5.4 INTERFACES Each applicant that uses this topical report as licensing basis must provide application detail of the analog transmitter and trip unit hardware into The following information must be provided by the applicant to his plant.
the NRC.
\\. _,)
5-9 l
... - ~. -.... -..
r,
-y NEDO-21617.AJ
~
s I5.4.1 Specific Instrument Loops i
Supply information for each instrument loop that will be converted to the analog sensor system as' identified'below:
(1) Variable name (2)- Part number of device being deleted (3)
System involved (4) The' engineered safeguards division h
(!-) Model number and vendor of the transmitter or RTD 5.4.2 Trip Unit cabinet Supply information for each trip unit cabinet as identified below s,
.) ;
Cabinet la9out showing location' areas of'the power supplies, trip relays, (1) and trip units.
(2) Division to which the cabinet is assigned.
(3) Layout of each card file in the trip unit cabinet showing the trip l
variable for each card file slot.
5.4.3 Environmental Interface The environment at each location where the retrofit hardware will be located -
must be compared to the maximum environment as stated in the topical report' for the following factors:
(1)
Normal operation and post-accident temperature and humidity.
%i V
5-10 i_.__.--______---.- _
v
(
(~)
Comparison of the floor seismic response spectra of the cabinet mounting (2) location for the specific plant to seismic test envelope that the cabinet was tested to.
I (3)
If the. trip unit cabinets are not located in the preferred location as per paragraph 5.1.4, provide justification for the alternate selected j
location.
l 5.4.4 Specific-Plant Interconnections An interconnection diagram which shows the interconnections between the I
existing logic cabinets and i.strument cabinets and the new trip unit cab-n inets is to be provided the NRC. The content of the information is to be i
similar to the information shown on Figures 5-3, 5-4 and 5-5 as applicable.
l The detail of interconnection shown on the retrofit elementary and inter-connection block diagram should be sufficient.
l I
~
1
}
5.4.5 Field Calibration Rack 1
?
The design and operational information on the " Field Calibration Rack" is to be supplied to the NRC if such a device is purchased and used for transmitter I
calibration.
1
)
s 1
( /
5-11/5-12 i
mew ~.
- - = -. - - -.
,._-~_.,~n.-
is.-
.F\\
6.
SITE CALIBRATION AND SURVEILLANCE TESTING TECHNIQUES 6.1 TRANSMITTER CALIBRATION 6.1.1 Frequency of Calnration Test Interval All of the essential logic within RPS and ECCS contain redundant channels.such as required for (A1 + A2)(B1 + B2) logic inherent in RPS or (A1 A2) + (B1 B2) logic characteristic of ECCS. Each of these channels is designed with its own independent sensce and tandem elements (i.e., transmitter and master trip unit).
Each master trip unit provides continuous readout of the transmitter control current via the mater on its f ront, which is calibrated in terms of the process.
~
In addition, an output jack provides a precision 1 to 5 Vdc signal variable.
which is generated internally across a 0.05% accurate 250 0 resistor in the 4 to 20 mA control loop. Thus, each parameter being monitored has at least one other channel, and in most cases three channels, which should show an identical reading at an instant of time.
The operator is able to cross-check-his transmitter output currents by comparison and therefore determine if,one of the transmitters is malfunctioning.
If a transmitter completely fails such that the loop is shorted or opened, the gross failure detection system in the For these master trip unit will activate an annunciator in the control room.
reasons, along with the proven reliability of the transmitters, it is considered that an adequate surveillance test interval for the transmitter is once per operating cycle.
i 6.1.2 Calibration Procedure Details of internal adjustments and calibration procedure for transmitters are e
described in vendor's literature. When the transmitter is connected to the g
(Figure 6-1), a digital voltmeter may be plugged into J1, master trip unit 3
which is a 1 to 5 Vdc conversion of the 4 to 20 mA transmitter current across
?
an interval 0.05% precision 250 0 resistor.
Calibration check points may be made according to the table provided with Figure 6-1.
S i
6-1
-l 1
....... ~ _... -.
L
NEDO-21617-A w.)y C
uij l
iCA L18 R A1'lO SOURCE OPEN VALVE TWISTED SHIELO MASTER TRIP UNIT Palm WIRE POWER n
l l
(s
{l TRANSMITTER j
"T----q l
F 1
l V
j 4 20 fnA 14 VDC NIGH IMPf0ANCE J
v CLOSED V ALVE OlGITAL h
VOLTMETER 1
PERCENT OF SPAN OCmA DC VOLTS (Jil 4.00 1.00 0
26 8Do 2.00 PROCESS 50 12.00 3.00 3
INSTHUMENT 75 1600 4.00 i
100
- 20D0 sa 9
l q
Figure 6-1.
Transmitter Calibration Setup.
l i
1
)
j The mechanical valving and test source connections necessary to set up the calibration source with the transmitter can be enhanced through the use of These devices allow immediate snap-in quick-connects (Figures 6-2 and 6-3).
threading and wrench-connection of tubing from the calibration standard without 1
A field calibration rack, which is available as optional ing tubing fittings.
hardware through GE, is capable of producing calibration pressure for both pres-sure and dif ferential pressure transmitters through any range specified in This device uses the quick-connects, which provides much nuclear plant use.
greater speed and simplicity to the calibration procedure presently used for
^l More information is available from GE on request.
J.
pres.sure and AP transmitters.
6-2 l
l t
v
-i OUICK CONNECTS quick CONNECT BLEED V ALVE
@i i N_
~
~
-_=
~ ~=
PR ESSURE
[ TR ANSMITTER
/
DtFF ERENTIAL
- PRES 3URE TRANSMITTER O
0 N
i C
ZZ n"
(<
m
.=-:
i
- N
)21/5 in._
-~E]
')
Figure 6-2.
Differential Pressure Figure 6-3.
Pressure Transmitter
'5 Transmitter
- p 6.2 TRIP UNIT CALIBRATION When the trip units are installed in the cabinets, a locking bar is provided across the front of each card file, which prevents tampering with the setpoint 6
adjustment screws of the trip units. The bar also locks all of the cards in place within the file so they cannot be easily removed by unauthorized person-l A key is required to remove the bar for setpoint calibration or card nel.
e removal.
l l
6.2.1 Functional Surveillance Testing It may be desirable for the trip units to be simply " tested" to assure func-l tional ability and/or logic permissives in the downstream system without pre-cisely checking the trip setpoint value. This can be done at any time without necessity of the readout assembly by simply selecting the appropriate card of 3
\\ /
the file with the calibration unit selector switch and using the transient 6-3 b
l
--s
,;n.,-nu,.,,-~.--.-n-:...m-:--
I__.-,-.----n-_.-.c...-
....-,.-.,..,m,,.,
.i l
l
/
(see current supply as the calibration current source to the trip unit input for further details on operation of the calibrator and its func-Ref erence 1 Steps for this type of simplified test may be as follows:
tions).
l confirm that no other channels within the system are in calibration I
(1) or test; pull out and rotate the calibrate command / calibrate current selector l
(2) switch on the calibrator until it indicates the appropriate master card associated with the channel under test; trip unit push La the selector switch and note the calibration'" gross failure" (3) light and annunciator are activated; I
be sure the Transient Current' knob is pulled out and turn the Stable l
(4)
Current adjustment until the meter on the master trip unit reads midscale; select the desired position of the Transient Polarity switch such (5) that current will add C+) or subtract (-) with the preset stable current; l
push the Transient Current knob in and adjust' current until trip
(.6) output light comes on; confirm that trip relay has changed state and that its contact has l
(7) activated the channel permissive; the pull out selector switch to disengage calibrator and reconnect (8) transmitter to the trip unit; confirm that the annunciator and gross failure clears and that meter (9) indication has returned to its normal level (also assure trip light and logic has returned to their normal states); and t_:
(10) reset the logic if necessary.
6-4 I
r-1 NEDO-21617-A
- 6. 2. 2 ~ Setpoint Calibration u
l Detailed calibration procedure for the trip ' unit setpoints may be found in' This procedure should be followed at least four times per year.
Ref erence 1.
if However, until the new equipment has accumulated life data, a monthly test interval is recommended for each system (. Subsection 3.4.4).
If a monthly l
interval is required for the system logic as well, this calibration check may be performed in conjunction with the functional test described in Subsection l
6.2.1.
l l
The setpoint calibration test must be performed with the portable. readout l
assembly, which has previously been standardized with the Bench Test Facility (Reference 1).
The calibrate current ramp rate is limited internally within the calibrator to.
a rise time of no more.than 1.1 mA/sec.
Since the response.. time of the trip unit is less than 2 msec (Subsection 3.3.2), an error of no more than 1.1 x mA can occur between the actual trip point and the value dis-
,D (0.002) = 0.0022 played on the readout assembly. This represents a worst-casa error of only
')
of span and is not even detectable on the 4-digit readout assembly,-
l 0.01375%
which has resolution of 0.0;1.mA.
is not rate limited, but when it is engaged, the trip' The transient current
]
This current display on' the readout assembly is automatically blanked out.
prevents erroneous setpoint calibration by. attempting to use the wrong current The transient current source is designed to source from the calibration unit.
provide step current changes to check time response or functional ability of intended for use when. calibrating l
the trip units and downstream logic but is not l
the setpoint.
' o 6-5/6-6 ri --
-'~ m m,c,w :..
- F m m
NEDO-21617-A~
. 1 t
1
?
l7. CRITERION COMFORMANCE
'N It should be amphasized that.the analog transmitter / trip unit " system" described is only a more sophisticated set of replacement hardware for.
in this report mechanical sensor switches.. Implementation of this hardware into a logic system
'I Guides and f
only af fects that ' system at the sensor level, not the logic level.
standards which specify regulations and criteria at the system level are only applicable as they concern individual sensora within the system. - For example, j
the sections within IEEE-279, or IEEE-379 which ' deal with " Single-Failure criterion" or "Manu'al: Initiation of Protective Actions" address tha logic philosophy of essential systems,.which ~ remains unchanged with' this particular i
Similarly, a system's conformance with Regulatory Guide 1.75, I
modification.
" Physical Independence of Electrical Systems," remains of equal'statusLwhethe this hardware is used or mechanical switches are used as sensors in the logic -
h ht
. ' However,: all of these. examples and other standards documents t oug 3
circuits.
to have at least remote significance with the new hardware addition are addressed
)
individually in this section.
7.1 - IEEE STANDARDS l
\\
i Criteria for Protection Systems for Nuclear Power
7.1.1 IEEE-279-1971
/
Generatina-Stations j
Conformance to the criteria of this standard is met as detailed in Table 7-1.
7.1.2 IEEE-308-1971
' Clash 1E Electric Systems for Nuclear Power.
'1 Generating Stations-for the most part, not applicable The standard was reviewed and found to be, Some portions of the standard may be to this innovation of sensor input s.
l to the inverters and/or 24V power supplies.
deemed to be somewhat televant However, channel redundancy and FMEA/MTBF analysis requirements defined.in suffice to show the conformance criterion is met.. for Section 3 of this report these units.
cN 7-1
=
4 er a
6. w,.e
+
e e -ee - o o
,e
a, NED0-21617-A-4
7.1.3 IEEE-323-1974
-Qualifying class 1E Equipment for Nuclear Power y
i Generating Stations n
This standard was used as a primary guide throughout the qualification process of each new piece of hardware in this package and was emphasized 'on the purchase specification or drawing as a requirement to each vendor performing the tests.
It is. felt that conformance has been r.et co the maximum extent possible at the present thme.
Since operating experience is not' yet available for the trip units, power eup-plies or inverters (as specifically designed for General Electric) were quali-fied by a combination of both type testing, including margin (Section 4 of this -
Subsection 3.4 of this report report), and analysis. using MIL-HDBK-2178.
details how MTBF 'and PMEA analysis documents, required of all five vendors, were used in conjunction with IEEE-352-1975 :to determine a conservative c'est interval (B), which may be expanded as experience is gained.
7.1.4 IEEE-336-1971
Installation
- Inspection, and Testina Requirements
'v for Instrumentation and Electric Equipment Durina the Construction l
j of Nuclear Power Generatina Stations l
The hardware associated with this innovation' is _ subject to the same Quality 1
)e l
Assurance program within the General Electric Company as. has been applied to
. -other' Class 1E hardware associated with the BWR/6 safety essential systems.
1 l
l Conformance is therefore met for those portions of the-standard within. General 1
Electric Company scope of supgly.
l t
7.1.5L IEEE-338-1971:
IEEE Trial-Use Criteria for the Periodic Testing l
of Nuclear Power Generatina Station Protection Systems l
The hardware defined in this ' innovation conforms to all of the criteria within as shown in l
this standard that apply - to. the channel or sensor level equipment, Section 6 and Subsection 3.5 of this report.
5 e,%
1
- )
1, 1-7-2 l
i
-8 a
-v----____________
NEDO-21617-A IEEE Trial-Use Guide for Seismic Qualification of
7.1.6 IEEE-344-1975
for Nuclear Power Generating Stations Class 1 Electric Equipment the seismic qualification This standard was used as a primary guide throughout process of each new piece of hardware in this package and was emphasized on the purchase specification or drawing as a requirement to each vendor performing as shown in Subsection 4.2 the tests. Conformance to this standard has been met, of this report.
7.1.7 _IE_EE-379-197 2 :
IEEE Trial-Use Guide for the Application of the Single-Failure Criterion to Nuclear Power Generating Station Protection Systems II Conformance to this standard is met both before and af ter this innovation l's because of the multichannel logic already inherent in essential systes (Sub-The standard is not directly applicable to the individual new 1
lp section 3.1).
hardware components but to the already existing redundant channels in which l7 they are used.
I
- 7. J. 8 IEEE-384-1974:
IEEE Trial-Use Standard Criteria for Separation 5
of Class 1E Equipment and Circuits
!l l
All of the domestic BWR/6 RPS and ECCS essential Class 1E systems have conformed with this standard on a system basis with optical isolation and separation as Therefore, f
lg defined by this standard, in conjunction with Regulatory Guide 1.75.
this criterion is met for these plants regardless of the sensor innovation dis-f p
cussed in this report.
l f
Operating power plants and most of those nearing construction end have existing lq systems which are not subject to the criteria of this standard.
Nevertheless, l
the panels designed for retrofit with these systems will conform to the maximum practical extent, as shown in Section 5 of this report.
0 l
l I
i i
7-3 4
_,, e' I
e-a
l
'i NEDO-21617-A
]
l
-1
]-
7.2 NRC REGULATORY GUIDES Periodic Testing of Protection System Actuating 7.2.1 Mu_1atory Guide 1.22:
j Function this guide is applicable to the logic of the safety systems j
For the most part, l
themselves and to the sctuators from the systems, both of which are beyond the l
J
.However, conformance to.Section 3 is specifically met as l
. scope of this report; shown in Subsections 3.1.3 and 3.2.4 of this report.
a 7.2.2 Regulatory Guide 1.29:
Seismic Design Classification
.j
{
All hardware for this innovation' has been qualified in accordance to the Seis-Conformance.to IEEE-mic Category I classificatici as defined by this guide.
has been met, as shown in Subsection 4.2 of this report.
l 344-1975 l
Quality Assurance Requirements for the Installation.
7.2.3 Regulatory Guide 1.30:
Inspection and Testing of Insttgentation and Elejtric Equipment The hardware associated with this innovation is subjected' to the same quality assurance program within the General Electric Company as has been applied co-e other Class 1E hardware associated with the BWR/6 safety essential systems.
Conf ormance is therefore met for those portions of the. guide (specifically Sec-j tion 2.2 of IEEE-336-1971) applicable to General Electric Company scope of
.j supply.
J J
4 Use of IEEE-308-1971, '" Criteria for 1E Electric 7.2.4 Regulatory Guide 1.32:
'l Systems for Nuclear Power Generating Stations"
.1 as specified j
l The additional criteria to be added with that of IEEE-308-1971, These l
l by this guide, concern of fsite power and battery charger load demands.
have no relevance to the innovation defined within thrt. scope of this report.
l Therefore, this regulatory guide is not applicable.
l i
7-4 l
e 3
gyp
)
7.2.5 Regulatory Gu,i_da 1.47:
Bypassed and Inoperable Status Indication far m
Nuclear Power Plant faf ety Sys tems Additional criteria supplementing Section 4.13 of IEEE-279-1971, as specified by this guide, concern extension of bypass or inoperability annunciation to indicate such action at the system level.
Bypass of the sensor loops, as are available to system design appli-defined within the scope of this report, These stay be used in con-cations as shown in Subsection 3.2 of this report.
junction with system level annunciator where determined necessary by the system Confornance to this guide is therefore met as it applies to the
- designer, capability of the sensor loops.
7.2.6 Regulatory Guide 1.53 Application of the Single-Failure Criterion to Nuclear Power Plant Protection System,s, Additional criteria supplementing IEEE-379-1972 provided by this guide are. not directly applicable to the individual sensor hardware but to the already exist-j ing redundant. channels in which they ~are used.
(
o o
l 7.2.7 Regulatory Guide 1.62:
Manual Initi,ation of Protective Actions l
This guide applies to the system level and is not applicable to the individual sensor chain as defined by the scope of this report.
7.2.8 Regulatory Guide 1.75:
Fhysical Independence of Electric Systems This criterion has been met as indicated in Subsection 7.1.8 of this report.
7.2.9 Regulatory Guide 1.89:
Qualification of Class 1E Equipment for Nuclear _
Power Plants This criterion has been met as indicated in Subsection 7.1.3 of this report.
l l
l i
7-5
.,m
.m.,., n,,.o.,,,.,
v q
).
7,3 OTHER DOCUMENTS 50 Appendix A Title 10, Chapter 1, Code of Federal Regulations, Part 7.3.1 The following general design criteria (Table 7-1) are considered at least partially applicable to this innovation and applicable portions are assumed to have been met l
in c'onjunction with one or more of the requirements' imposed by previously discussed j
regulatory guides or IEEE atandards.
l GDC 13, 18, 20, 21, 22, 23,.29, 34, 35, 37, 46 Criterion numbers not shown above are considered not applicable to the innovation as described within the scope of this report.
I J
l l
l l
v
./ -
t 1
7-6 l
J
NEDO-21617-A Table 7-1 IEEE STANDARD 279-1971 (CRITERIA FOR PROTECTION SYSTEMS FOR NUCLEAR POWER GENERATING STATIONS)
Conformance Section of IEEE-279 4.1:
" Automatic Action" The new hardware meets this requirement (Reference 1).
4.2:
" Single-Failure Criterion" This requirement is met regardless of implementation of this hardware because j
of the multichannel logic already inherent
{
in essential systems (Subsection 3.1).
q t
This requirentant is met as shown in Sections
-l 4.3: " Quality"
)
~
3 and 4.
^
I This requirement is met as shown in Section l
1
4.4:
" Qualification" 4.
1 This requirement applies to the total pro-
'8 4.5:
" Channel Integrity" tection system channel, of which this hard-were is only a part; however, the hardware conforms as shown in Section 4.
I e
This requirement is met regardless of 17
("I
" Channel Independence" 4.6:
implementation of this hardware because of the multichannel logic already inherent in essential systems (Subsection 3.1).
l l
I I
This requirement is not applicable to the 4.7:
" Control and Protection System Interaction" new hardware because it is only used for l
protective acts.ons.
This requirement is met by virtue of trans-4.8:
" Direct Inputs" mitter operation which ditectly monitors the process variable.
i This requirement is met by both channel t
4.9:
" Sensor Checks" crosschecking and surveillance testing as shown in Subsection 6.1.1.
Conformance is improved with this innovation over the
" blind" sensor switch.
I 4.10:
" Capability for Test and This requirement is met once for fuel outage transmitter and monthly trip unit Calibration" tests as shown in Section 6.
O 7-7 4
\\
i, m., m,..
m
,,,,m e._m_,. m..,.
.,m. n,,.m.
r._,3
. Table 7-1 IEEE STANDARD 279-1971 (CRITERIA FOR PROTECTION SYSTDfS FOR NUCLEAR POWER GENERATING STATIONS)(Continued)-
Conformance-Seetion of IEEE-279 4.11:
" Channel Bypass or Removal" This requirement is met as shown in Sub-section 6.2.1 for (1/2) x 2 logic. Con-formance to this section is greatly improved
~
with this innovation because of the shortened bypass time over mechanical sensor switches.
(See justification Subsection 1.1.2, items A and F.)
This section is not applicable at the 4.12: " Operating Bypasses" individual sensor level, which involves the new hardware.
1 4.13: " Indication of Bypasses" This requirement is met as shown in Subsec -
tion 3.2.4.
Sensor bypass is manually achievable as 4.14:
" Bypass Access" indicated in Subsection 3.2.4.
Channel -
]~-
bypass downstream of the trip units is The beyond the scope of this report.
requirement is therefore ' met for appli-cable portions.
4.15: " Multiple Setpoints" This requirement is met (Subsection 6.2.2).
Conformance is improved by this innovation.
1 i
the 4.16: " Action Completion" This requirement is already met at l
system level and is' unaf f ected by this innovation of the sensors.
i l
4.17: " Manual Initiation" Same as above.
Conformance. to this requirment is improved 4.18:
"Setpoint Access" with this innovation because of easier access and better precision of setpoint adjustment (Subsection 6.2.2 and Reference 1).
This requiement is already met at the down-4.19:~ " Protection Action stream channel level within each system and Identification" is unaffected by this innovation of the' sensors.
4.20:
"Infomation Readout" This requirement is met as shown in Subsec-
.) !
tion 3.2 of this report.
b 7-8
NEDO-21617-A' f,
Table 7-1 IEEE STANDARD 279-1971 (CRITERIA FOR PROTECTION SYST1!MS FOR NUCLEAR POWER GENERATING STATIONS)(Continued)
Conformance Section of IEEE-279 1
Conformance to this requirement is improved
.I 4.21:
" System Repair" with this innovation as indicated in almost every section of this report (i.e., Sections 1, 2, 3, 5 'and 6).
4.22:
" Identification" All of the hardware pieces in the tandem chain are given master parts list numbers which are prefixed with the alphanumeric code numbers of the systems in which they belong.
For example, some. actual trip i
unit identifications may be seen on Tables 3-1 and 3-2.
The same codes are used for the transmitter, power supplies, inverters and trip relays. Thus, the requirements i
I of this section are met.
e f
(. h) la l
l O
4
)
7-9/7-10 eg me e s o m esummmmeenpapegsph
- de****
8.
00NCLUSION This report demonstrates the superiority of a new concept in sensor intelli-gence adopted by the General Electric Company in its standard product line for The mechanical " blind" switch used previously was found to essential systems.
have inherent drif t problems which were undetectable to operating personnel Undesirable inadver-except through cumbersome mechanical testing procedures.
cent activity has been attributed to errors made during such tests or to the switch drif e itself.
Replacement of the mechanical, switch with the analog transmitter / trip unit system has yielded the following valuable improvements:
Loop setpoint worst-case accuracy for normal plant conditions has e
narrowed from 22% to !0.35% of input span.
l l
Previously unknown values of logic sensor loop parameters are o
\\f available to be continuously monitored.
l Most failures in field cables and hardware are automatically l
e detected and annunciated.
Calibration and testing procedures are greatly simplified and more l
e efficient.
Unpredictable human error introduced in attempting to read moving l
e gauges when calibrating setpoints appears to have been completely eliminated.
Conformance to several regulatory guides and IEEE standards has been e
i considerably improved (specifically Regulatory Guide 1.22; IEEE-279,
- 4. 3, 4.5, 4. 8, 4.9, 4.10, 4.11, 4.15, 4.18, 4. 20, and 4. 21 ;
Sections and IEEE-323), while none has been compromised.
Possibility of inadvertent scram or ECCS initiation is greatly (y
e reduced because of improved calibration test frequency and technique.
8-1/8-2 I
s,..
_.-..,.,z.,,_.._x
. ' ~
9.
REFERENCE d
\\
' Rosemount Inc., Operations Manual - Trip / Calibration System - Modei 510DU, 1.
Instruction Manual 4247-1, Copyright 1976, i
4 l
1
.{
)
D l
l 1
1 9
k,, /
l
.i l
1 l
1 l
l 1
9-1/9-2 i
i l.
e w__
i
NEDO-21617-A-
-)
~1 APPENDIX A.
RESPONSES TO REQUESTS FOR ADDITIONAL INFORMATION 1
l 1
i 1
8 y
a 1
0 l
0
~
}O l
e l I i
- - - - - ' - - - " * ' ~ ~ * '
.4, 4
- -