ML20034C951
| ML20034C951 | |
| Person / Time | |
|---|---|
| Site: | Oyster Creek |
| Issue date: | 08/14/1989 |
| From: | GENERAL PUBLIC UTILITIES CORP. |
| To: | |
| Shared Package | |
| ML20034C950 | List: |
| References | |
| ES-002, ES-2, NUDOCS 9006040122 | |
| Download: ML20034C951 (14) | |
Text
.
nowrcAL nmcrrms Num=r Nuclear-orv1Szm Es-002 Title-Revision No.
INS'!RMNT ERIM CM.fUIATION AND SEITODff IETERMDOCION 2
Applicability / Scope Responsible office
'Ndinical Functicms Division EP&S 5110
'Diis document is within % plan scope 1Yes _No -
Effective Data Safety Reviews Requiraci Yes X No 8-14-89 List of Effective Pages
,Page Revision 1.0 2
la 2
T k %qm-2.0 2
3.0 1
4.0 1
5.0 1
6.0 1
7.0 1
8.0 1
9.0 1
10.0 1
11.0 1
12.0 1
13.0 1
'Iha pages affected by this minor revision are shown above.
Sianature Gamsina Orgertizaticmal Elenent Date or:ainator
/3/
Encrineer, EP&I
-4/4/89 W u.AA.6&uu Dy
/3/
F.d m.ssw. hw. 4 M45 ruumm.mr
- /M/JT
'/s/
Manaaer, Elec. I'cwor & Instr.
4719/89
'/s/
Plant Systans Duector 4/20/89-Approved By
/M Director, Encrineerim & Design 4/24/89 f,@f0$0f3bkb0 P
ENuclear W2 m m n u mCuS DIVISICH Title. Instnment Error calculation ard Setpoint Determination REV St.291Mtr OF OpM2 APPflDVAL DATE.
CRIGINAL ISSUE EATE: 1-1-88 1-mis rwision prwides clarity to Satpoint and T.S. Allowable Limits (Section 5.2.2),.
Setpoint Upper /Iower Limits (Section 5.2.4),
and Acceptable-as-Found Limit (Section 5.2.5);
i plus, adds surveillanos Test 'Iblerance Error to ocnsider under Normk1 Accuracy errtxe (Secticm 6.3.3), and delstas ocabination method requirements (Section 6.4) (5/22/89).
l l
2
'Ihis minor rwision clarifies that the 0 /fQ standard applies to safety related and BOP calculations.
1
'I l
la
Nuclear-DIVISION IS-002
= tc - Io"S
""**r Title Revision No.
Instrunent Error Calculation and Setpoint Determination 2
1.0 PURPOSE AND SCDPE 1his design standard describes a uniform method for establishing setpoints for instrumentation channels. Setpoints are intended to initiaita ~4ve or protactive actions in a timely manner before the safety of pm:ple and/or equipment are endargared. 1herefore, those factors that influence the accur cf of the instruments initiatirg the setpoint function nust be ocnsidered to assure that the instrument trip signal will be generated when it is required. This stardard addresses the factors that nust be considered when establishing an instrument setpoint and how these factors are ocabined. Included are the rela-tionships of the instrument setpoint to the various limits of normal operaticri, design basis event envirtrinental ocriditions, and calibration and marvat11ance tasting. The scope of this standard does not include the methods for deem m Ming equipment protective values, process safety limits, safety facters that shculd be utilized, or design basis event analysis that apply. Factors affecting instrument acx:uracy are de-scribed, along with methods for calculating their inpact cri overall loop accuracy, and setpoint.
i 2.0 APPIJCABIIJW
'Ihis standard shall apply to new designs, modifications of design, and for new calculations required for safety systemt instruments that are provided with setpoints more specific acticms arm either initiated, terminated or prohibitad. 'Ihe methods in this standard shall apply to both safety related and Balance of Plant (BOP) calculations.
Error calculations that utilize this standard shall address every item in the Rapirements 'ar+4rwn 5.0.
If a recpirement of this standard is evaluated as not Widle, then the calculation shall provide a statament of the reasons why it is not included.
3.0. REFERENCCS 1.
ISAH551.1 (1979) Process Instrument Terminology.
2.
ISA-S67.04 (1982) Setpoints for Nuclear Safety-Related Instrumentation Used In Nuclear Power Plants.
3.
Regulatory Gaide 1.105 revisicn 2 (February 1986) Instrumentation Setpoints for Safety-Related Systems.
4.
NUREG/CR-3691 ('=5+W 1984) An Asemasment of Terminal Blocks in the Nuclear sower Industry.
2.0
_ ~. _ _ _._ __ _ _ _ _ _ _ _..._. _. _ _ _ _ _._.. ~. _ _..._. _.
Nuclear TEDINICE FUNCI' IONS Number DIVISION iS-oo2 Title Revision No.
Instrument Error calculation and Setpoint Determination 1
4.0 DD"INITICNS 4.1 Accuracy - In process instrumentation, degree of conformity of an indicated value to a recognized W -4 standard value, or ideal value (Ref.1).
4.2 Accuracy. Measurud
'Ihe nadam positive and negative deviation aboarved in test:.ng a device uru$er specific corniitions and by a specified r h.
(Ref. 1)
Note 1: It is usually sensured as innocuracy and expressed as accuracy.
Note 2: It is typically expressed in terms of the =maatted variable, percent of span, percent of upper rangWue, percent of scale length or Hit of actual cutput reading.
4.3 Accuracy Ratina - In process instrianentation, a manbar or quantity that defines a limit that errors will not exceed when a device is used under specified operating ocnditions.
(Ref. 1)
Note 1: ition operating conditions are not specified, design operating conditions shall be aseLaned.
Note 2: ' As a performance specification, accuracy (or reference accuracy) shall be=====4 to mean accuracy reting of the device, when used at reference operating conditions.
Nota 3: Accuracy rating includes the cambined effects of conformity, hysteresis, dead bord and repeatability errors. 'Ihm
-t units being used are to be stated 5icitly. It is preferred that a + sign precede the runsbar or quantity. 'Ihm absence of a sign irdicatas a + armi a - sign.
4.4 Desian Basis Event Analysis
'! hat analysis used to deMmine safety system reopenses to design basis events.
4.5 g3 - An undesired ctiaruja in the output over a period of time.
'Ih;a change is unrelated to the input, envi,. hit, or load.
(Ref. 1) 4.6
_Eggg - In process instrumentation, the algebraic difference between the indication and the ideal trum value of the *mmasured signal. It is the quantity whicts algebraically subtracted frca the indication gives the ideal value.
Nota: A positive error denotas that the irsiication of the instrument is greater than the ideal value.
Error = Irriication - Ideal value (Paf.1) 3.0
0 le WINuclear
c= =n
- =
DIVISIs ES-002 Title Revision No.
Instrument Ertur calculation ard Setpoint Datarminatica 1
4.7 Renar, Erwi.
-.M - Errcr canaged by a mange in a specified aparating condition fram refersnos cparating omniition.
4.8 E..-,
i.-_-- - Mansurement (smii - Procaes errors that include thmaa ii.%..i in the mensarument taenique, for example fluid stratification affects on tamparature measurements, or the effect of fluid density manges en level measurement.
4.9 Errer Position - The change in cutput resulting frtan mounting or setting an instrument in a position different frta that at which it was calibrated.
(Raf. 1) 4.10 Eter kndas - In this design standard, all errors that are not systsematic errors. Error &as to no kncadn cause, usually the not result of a large number of semil effects. ' Die value tends to cluster about a central point with a fairly agaal spread above and balcat the central value.
4.11 Error Systauntic - An arrer WicA in the course of a ran=har of maasairements anda under the smaa omnditions of the same value of a given gaantity, either remains ccristant in absoluta value and sign ce varias accending to a definita law een the conditions mange.
4.12 L=6
. Ca. 1 - An a..-w.; of acaponents and arrkilas that ganarata a single protactive action signal tan required by a ganarating station condition. A chamal icons its identity whers single protective action signals are acabined.
(Raf. 2) 4.13 P.edve Act.cn - The initiation of a signal or aparation of equigment with..n the protection system or protactive action syrtam H
to N ah a protactive function in respanas to a ganarating station condition having rem &md a limit specified in the danign basis.
(Raf. 2) 4.14 Safety Limit - A limit plaond on important procons variablas that are rimamanary to renacnably protect the integrity of certain physical barriers guartiing against the uncontrolled releans of radioactivity.
(Raf. 3) 4.15
- An irgut variable that sets the desired value of the ed variable.
4.16
- the algebraic difference betwaan the upper and icwor ranga uns.
4.0 I
l
l o.
l l
Nuclear 2==
2
- =
l DIVISIN IS-002 1
1 Title Asvision No.
i l
Inste.amnt Error calculation and Setpoint Detaminatica 1
l l
l 5.0
=r=e. w 5.1 Detatninetion of Setuoints Setpoints shall be seleccad to prwide sufficient maririn between the udve action'astpoint and the systaa protmet: cn limits to r
account for all the inacmuracy inherent in the instruent icop.
Stis inaccuracy may be due to instrument inaccuracy, locp alibra-ticm tolerance, innocuracy of the test equipment, procons measure-ment innocuracy, effects of transient cwershoot, effects of tima xtepcmas ciaractaristics, envirtrumental effects, instramnt drift, or the effects of normal process transiente/ upsets. Detailed restirements for the instrueant setpoint relationships are delin-ested in this section and illustrated by Figure 1.
i 5.2 Allowable Value PMchnical Specificatica Limit) l l
The tactrtical specification limit shall be regarded as the oper-ational allcmable value. cperatics within the allowable value
+
ahall prwide assurance that autmatic protmetive actice will correct the must severs abnormal situatien ar*4eie= tad before a safety limit is M.
5.2.1 The allowances between the allowable value (tadrtial specification limit) and the safety limit shall include the following itmos unless they are included in the 2
determinatics of the process safety limit.
a.
1ha effects of potential transient cworehoot as datamined by the design basis event analysis.
b.
De effects of the time respones dm.cactaristics of the total instrueant channel, including the sensor, c.
Enviremental effects cm instraunt accuracy or time rempanos charactaristics caused by anticipated opersticmal occurrences or design basis events for those instruments required to mitigate the ocnsequences of such events.
5.2.2 Setpoints and Technical Specification A11cadable Limits the estpoint value to initiate rMdve action, cembined with the instrument loop inaccuracies at normal aparating ocrr*1tions, shall not em:med the Allowable value (Technical specificatien Limit). Stis margin between setpoint arti allowable values shall prwide allowance for:
l l
7 5.0 j
4 1
rl j
13OntImL 70tcITcts mauber J
DU1 Nuclear i
nu.
movistcri so.
Instrument Drur calculation arti setpoint Detasmination 1
)
)
l a.
The toleranos specified in the instruent laap onlibra-1 l
tien procesare. This value is considered a systamatic etter for annearvatism.
b.
All instrument rendem and systeentic innocuracies and drift that occur å normal envircrumental operating l
omnditicris.
c.
Accuracy of the test equipment used for surveillance tasting the instrument channel.
d.
Process mammarement accuracy sudt as the effects of fluid stratification and changing fluid density.
i 5.2.3 W ints and operational Limits Refering to figure 1, the normal desired aparation refer-ence line represents the naminal value the aperstor de-sires for the process. 1his ruzainal value of the process i
varies slightly with aparational transients that are in an acceptable range to the operator and this is represented
+
by the operational Limit reference line. She instraent trip estpoint specified mast pewide sufficient entgin to 4
- that ir==rucy and rum: mal drift of the instruments do not cause the instament to trip cut process values with-in its operational Limit.
i 5.2.4 Setpoint Upper /Inwar Limits (As Inft Calibration Telerance)
This is the toleranos specified in the surveillanos test preendure that is==+*1e without reenlibration. The band between the setpoint and its upper or lower limit shall be broad enouqtt to minimise the need for frequent instament adjustment. 1he setpoint upper and lower limit (amlibestian tole:nnae) specified cri the test rh any be derived frun historical amintanence data or other engineering sources. For new designs, the test tolerance shall be calculated to incluis the statistical cambination of the two sigen randen accuracy ratings of all the instru-monts that oceprise the laap, and the accuracy ratings of the test equipment used to ruHhrate thmen instruments.
This value is expressed with a plus and miraas sign that defines the upper and lower setpoint limits. Per new designs, the set: point is dstaruined by algebrair=11y adding the calibration r h toleranos to the calcu-lated loop a Tor. This is very ocrimervative. Historical performance of the instrument loop may permit a reduction in this ocriservatima and F-M:- A. setpoint readjustment.
6.0 i
. _ _ - ~. -. -
i TEC39fICAL FTJNCTIc35 Rober P"d71 Nuclear
=*
=
title Revision No.
Inst:nment Erzar calculation and Setpoint Detaminatiert 1
5.2.5 h-j *le-es-Pbund Limit the _W=hle-es-Feund limitj also known as the surveillance test Acomptance critaria at 1MI-1, shall be used for the evaluaticrt of surveillance tast as-found data. 1his band defines W= hie drift limits of the instaments included in the surveillanon tast, and is used to outfirm that the instament loop has not drifted boycrd an W ahle predicted value.
I As-found limit shall include errors due to drift, tapar-ature affect, and power supply affect in adiition to thces errors used to dotarmins the ugger/ lower set point limit.
i
(
Pbr acas M= ting instnsent loaps at 1MI-l the surveillance test %
Critaria (Wahle as found I
limit) is the amma value as the estpoint upper / lower limit value. Fbe these instnannt loops the values were either derived fra historical instnamnt performance data, or f'
were specified by~the mangfacturer that prwided the instmaant loop.
Errors that are not mamawable during a survalliance test are not included in the. & le-es-found limit.
(e.g.
Transmittar error idien the transmittar is not part of the
(-
surveillanca test.)
l l
1 4
4 7.0 w-~sv
{
Nuclear
=c=. ruerIo
=
- u DIvISIaN rs-o02 Title Revision No.
Instrument Erme Calculation and Setpoint betamination
,1
- NDttt ' mis figure prwides Figure 1.
Instrument Setpoint Ralationship
- relative position anly & is not
- intended to imply direction.
SML'N LDET o Design Basis Event Analysis o Safety Facter PROCESS SAFEIY LIMIT o Transient overshoot o Accident Dwironmental Effects on Instrument Accuracy o Any Errte Facter Not Incluled Balow -
o Time Response Characteristica
__antwamn LDET (TEGORCAL LDGT) j g See M im 5.2 o Preema Measuremet Accuracy o Instrument Accuracy of Any campanent Not In 7 test o Accuracyw YM Used in Surveillance Test AC3PIMEE-AS-FU.34D q y o Accuracy retirg of Instnsumts Tested calibration l
o 'Nuperature Erme Effect j
Racplired Sea Section 5.2.5 o Pawar Supply Error Effect I
o Drift of Instruments Ttustad
] f UPPER SEPPOINT LDET See Section 5.2.4 o 7blerance allowed for astpoint by j g No calibra-calibration y h tion Required SETPOINF IDIdER SITPODC LDCT See Section 5.2.3 OPERATIG4hL LDET (CESIRED) o Normal Operational Transients NCEMAL (CESIRED OPERATION) 8.0
e, Nuclear
== rumca
==
DIVISICH IS-002 i
Title Devision No.
Instzsment Errte calculation and Setpoint Detaminaticm.
1*
6.0 GMl11LATICIS 6.1 IN AccJrw/ Baseline 04%p i
6.1.1 Vendor pec&act specifications an(/or Envirtruental Quali-fication (EQ) files shall be used to obtain accuracy data for each instrument for norum1 and harsh envircruental ocmditions. It is usual practice for instrument suppliers to express accuracy data as an upper limit based en test data innocuracy values that none of their instrtaments will a mmad (the ccrifidence level is greater than 99.7%). It is ocnservative to mesmo that the vender accuracy data unless 1
otherwise specified, represents a probability distribution of three sigun with a ocnfidence level of 99.74.
l 6.1.2 Ftr accuracy miculations the IOC staff has accepted 95.5%
i (two sigen value) as a penbebility limit for errors (Daf.
L 3). 'marefore all calculations shall utilize all baseline l
data of rendca values at two sigen to provide loop acxuracy and set point values at a two sipa probability.
6.1.3
'Dires signs beenline values can be onnvertad to two sigen j
by, ultiplying by t6.
(
6.1'. 4 In an accuracy alculation of an instrument string of two or more components convert all indivi&aal omsponent errors 1
into the ccamen unit for the variable of omnoorn. For example, a string oansisting of a differential transmittar, an analytical unit and a matar any have their indiviimi accurecies expressed as % full range (psi), 4 of output (sv), and % of scale (ft. of water). If the variable of l
ozicorn is ft. of level, the accuracy of eacts ocuponent.
l should be convertad to fast of level for the calibrated range of the appliaticm.
l 6.2 ciabinim Act;uracy trTors
' Dip accuracy error values of the instriment loop are ccabined using the followirs; methods:
a.
All errors identified as random errors will be acabined usirq the square root of the sum of the squares methodb.
b.
Systematic arrers will be crabined by algebraic summation.
c.
'Ihe total innocuracy is the algebraic sum of the total rarxica error and total systematic error.
9.0
(
4 TfDMIchL FtNCnCNS Amber i
DTifT1 Nuclear i
o-Title Revision No.
i Instrument Emr Calculatim and Setpoint Deemminatica 1
l 6.3 Discussion of Accuracy calculatiers Fellowing is a discussion of various errtse to be ocmsidered.
6.3.1 Process Mansurement Prooses errore in mamaniemnt are caused by process variables other than that being measured. ncnr and level measurumore.s using differential pressure trenmaittare are matrject to errore chas to changes in the taperature and the i
static pressure of the process fluid.
'Naperature and pressure change influmnos the density of the fluid. mis danges the output of prienry ncat elements.
It also changes the pressure proikaned by a given level of nuid in a vammel or a level reference leg. Differential pressure transmittare are calibrated to namours ficet er level at a particular fluid pressure and temperature. Any change frta this conditicm prothzus an error. !he error sust be ocmputed for the fluid pressure and tampatiiture dange of interest and eNpressed as a WA. of the differ
- ential pressure span to netich the transmittar is m1thrstad.
l 6.3.2 Static Pressure Errors i
Differential pressure trenamittare are calibrated to correct fcr the syntamatic error at a particular static t
preemme. Any deviation frcel this normal static pressure l
conued by transient produces a new static error of the trenamitter span and zero. These new static errors are not i
corrected in calibratica and must be calculated for the pressure of interest. Tranmaittar marufacturere ganarally provide the arror egeticms for their instrsaments.
l 6.3.3 Nereal Accuracy h miculatiens are W^ 4 taking into acocunt normal (ncWdant) ccmditicms of laamidity, tamperature, etc.,
errors of the indivichaal mockLles that oceprise the loop considering the following.
(1) Module Accuracy Rating
'!he accuracy of the module under oansideration include the combined effects of probable error intrtxhaced as a result of dead band, repeatability, hysteresis, etc.
(2) calibratica loop tolerance - mis error is specified in the instrument loop surveillance test rh.
For new designs this errtT is calculated by statisti-cally cxnbining the two signa values of the accuracy 10.0 4
0
e _.
j TEORGChL FtNCTIONB Maber i
Flil71 Nuclear o="
Title Asvisicn No.
Instzumerit Erzw omicnalation and Setpoint Dstemnineticrt 1
ratings of all the names tus comprise the arveil-lanas test locp and the amourecies of the test equipment used Ardisy calibraticrt.
(3) 7tuperature Erzw - This erzw is caused by changes in i
the moMe ambient taperature frca cellbration ounditicms over the desicm ranas.
(4) Drift - this artw is caused by a change in accuracy within the design range acruiitions over a period of time.
(5)
Power Source - 1hese are arttuts introtkaned as'a result of fluctuatiano in the power apply feta c=1thruticm canditiens over the h.
(6) 1tst Instnament Errce - 1his erzw is the acusaracy of i
the test equipment used in the celibration or sur-veillance tasting of the instnannt. m ore the test equipment tolerance is expressed as plus or airsas % of reading a ocrieervative (manciaman) erzw value would be i
bened on the typer (1004) anlibrutad open reading of the instmaannt.
a.
Calibration test agalpment acuzaracias are used in calculating the celibraticri loop tolerance error (6. 3. 3. (2) ).
b.
Surveillance test equipment accuracies are used in calculating the total loop error (5.2.2.C).
l (7) leinnallariacus Errces - EtTcts that usy be P44** to a per*4 M me ac M e should be included, i.e., self heating.
l (8) Wirs Drop Errces - Errors intrtukaned as a result of i
)
l line deep, as signal current passes thmaagh dropping resisters, tarsinatiano, etc.
(9) Radiation Exposure 6.3.4 Accident Accuracy 1.
Accident accuracy calculations are performed only for thces instruments loops that have any component physically loostad in a harsh envircrament, only thces humidity, tamperature, and radiation erTers das to the amidarit erwircreant are czabined to define the Accident Accuracy.
l 11.0 p
.--S. -
_ _.. _ _ _ -. _ _ - - _ _.... _ _. _,.. _,.. _ _ ~.. _.. _, _, _. _ _ - -.. _ - _, _ _ _....
o T
I
'NotcchL FtMtTIG5 maber Nuclear om,=
m-*2 Title Asvision 13o.
Instnannt Error calcallaticm and Setpoint Detaminatica 1
Instuments are cpaalified to operate in the worst erwirtsment ar*ie4 = tad at its location. However, the actmaracy caloalaticms are perftsmed for the intended l
factim, and any utillas erwircremental limits that scist ears the intended function is accumplished.
1 2.
'me functicmal significanos of the varicus accident atTors may not be applicable een specific circuit functions ard applications are reviewed. Per instance, the instnment may have performed its protective function before it is exposed to anxima harsh accident erwironmental czmditiens.
'mie is very cumacn een cumsidering cirtzait Insalaticm resistanos errar. Good praction is to identify the errar and state clearly the roastms or==r*% for not in-cluding the arrue in the accddent calculaticut.
)
Mhare anxism==-% temperature and radiation errare do not occur annczarrently, perfces tae esparate calculatiens.
and identify them by the tian after the accident. Use the worst (greatest) error value of the tze calculations to define the W % Accazrecy.
In those instances Were the instament pro &act data shasts or a test reports provide error values fear erwiremental omnditisms auch more severs than the design omnditisms, and the arrer calculaticms are n, basis event ot realistic, J
one of the following authods any be T !
- 1=:
1 a.
Raquest the lastrument aarnatacture provide specific accuracy at the desired omnditions and document the respanas with IDsce anycr m files.
)
b.
Uma straight line interpo1Atim if the resulta vill be l
l omnservative.
l l
If the arrar is non-linear with the data points given, N
and is same unknown exponential function og =ill provide e,
for N>0 than a straight line interpolation w i
conservative arrar values ketman the extreme given values.
l 3.
Lar insulation resistanos values &as.to ocmdensaticri at i
tassire.tions (e.g. taminal blocks) or chas to cabl being in a hi$1 tauperature ambient can cause sigrtificant stmsment signal errar. A method to derive that arror is presented.
12.0
..mm.
c,..m,,,,
,y,,-,..,,.,%--
n-e-,
,e__
.=- m
7 I
l
=
8 l
7/
1 I
TE39tIchL PDICTIas Ambar P"51Nuolear l
Title Revision No.
l Instruusr* Error Calculaticm ard Setpoint Ostaminaticrt 1
l Effectively, an instnaert traramitter partenza as if it is a variable resister witts the resistanos value rWdcral to the proomes carditim.
Sandia Naticym1 Imbaratories, in NUIW2/CR-3691 (Raf. 4) described a ressemable basis for detaming icw IR induced errar. The following equivalent eqastian fcr exprinsing l
instnannt error is pewidad i
v - R,IT e=
s 2 TN - 8e) wiere:
e =
Error l
V3=
Sauros Voltage (Volts) j R, =
Etplivalent Series Resistanos of Icop Instnament and cable to low IR Region (Chus)
IT = - Transahttar output current (Amps)
Parallel insulaticrt resistance of cable /
R
=
i Sh ce splica (Including Adjustment for langth and circuit onnfiguratie) (Chas) 6.3.5 Time Response Time respcmas calculatims are dotarmined by alW.cially adding L, Os the respcrime tian of ancis instnannt in instnaurt diannal. Reaperine tian is prwided on the i
samfactures preckact specificaticn. Use the absoluta value pewidad, or the value repaired to obtain enn time constant.
Die tian response is analysed to insure the integrity of the. redvs actim is maintained. Diat the prccess safety limit is not e daring the tian required for the instmaane cnannel to initiate the protective action signal.
n 13.0
..u.
.