ML17209A183

Review of Temp Sensor Response Time Measurement Using Loop Current Step Response Technique.

ML17209A183
Person / Time
Site:	Saint Lucie
Issue date:	08/16/1978
From:	Ackermann N TECHNOLOGY FOR ENERGY CORP.
To:
References
NUDOCS 8010060267
Download: ML17209A183 (31)

Text

REGULAT~ INFORMATION DISTRIBUTIO YSTEM (RIDS)

ACCESSION NfkR:,BQjl0060267 DOC ~ DATE; 78/06/16 NOTARIZED: NO DOCKET" FACIL:50-355't. Lucia Plantg Unit lg Florida Power 8 Light Co. 4'5000335 AUTH NAME

~ AUTHOR AFF'IL'IAT ION ACKERMANN<N.J. Technology for Energy Corp.

REC IP. NAME RECIPIENT AFF IL'IATION

SUBJECT:

"Review of Temp Sensor'esponse Time Measurement" Using Loop Current Step Response Technique."

DISTRIBUTION CODE: Hoo 1 s COPIES RKCEI VED: LTR ENCL~ SIZE':

TITLE: DCD Materi al NoTEsi @DDo', Qoogg Qg@Qpp RECIPIENT COPIES RECIPIENT'D COPIES ID CODE/NAME LiTTR ENCL CODE/NAME LTTR ENCL INTERNAL: NRC PDR 02 1 1 EG FILE 01 1 1 EXTERNAL: LPDR 03 1 1 NSIC 04 1 1 ocr 8 @

TOTAL NUMBER OF~ COPIES REQUIRED! LTTR 4 ENCL'I'

l ) f L c >t 4

O

( p:I I t 1 C L I

'h "p l.

S A

4 t I R

C ~ +Q(~g, h V8" w I ,t

,l L'

r II

REVI EH OF TEMPERATURE SENSOR RESPONSE TIME MEASUREMENT USING LOOP CURRENT STEP RESPONSE TECHNIQUE 4

FOR NUCLEAR REGULATORY COMMISSION BY NORBERT J ACKERNANNg JR JULIAN E MOTT AuGUsT 16,'978 801006o QQ,7

'o 1

I' VI g ~ H I

C

1 l

Customer Technology Ior Energy Corporation 10431 Lexington Drive Bulletin Knoxville, Tenneaeee 37922

<e1s) ssasase SURVEILLANCE QF SAFETY/RELIEF VALVES CAN IMPROVE PLANT AVAILABILITY r

Displays hourly RMS Alert light indicates

!'ic

• ~

levels and variance for lg CIA%

g channel trip

>1h Qp-

%1 each channel on request ,I w ~ tl l

and automatically if

~

v ~

~ ap>> Keypad selection of tripped VE A ~lr ~

~ ~ ~ ~

~

QvE

~ W channel and function. for VI ~u ,< ~ v viewing TEC 1400 program-controllable 16 channel TEC 910/911 signal multiplexer conditioning amplifier system with charge-TEC 1408 programmable converter drivers filters: 1-15 kHz Microcomputer uses Over 100 days of data, TEC's custom software, including all trip records for safety/relief, valve for all channels, stored on monitoring one diskette A TEC Model 1400 microprocessor-controlled surveillance system is being used to continuously monitor and process the acoustic noise within BWR safety/relief valves, so that valve leakage can be recognized long before a malfunction results. A commonly used safety/relief valve in BWR plants is a pilot-operated type, having a two-stage pilot valve section and a main valve section. The pilot-disc seat is vulnerable to deteriorating processes which lead to increasing leakage across the seat into the chamber of the second stage. A small leak has no immediateadverse effect, because the pilot valve discharge line can accomodate the flow without substantial pressure buildup. However, as a small leak grows, the second-stage disc opens; even though the pilot disc was not lifted. Since valve malfunction does not occur until a leak worsens to that critical leak rate, early detection of leakage can allow corrective measures to be scheduled, thereby reducing the loss of plant availability. Also, the surveillance data provide criteria for selecting valves which should be inspected during the next outage.

'I Serving the Energy Production Industry

Several types of surveillance techniques have been considered for detection of the valve leakage.

Sensing the temperature near the discharge line is one possibility. This approach is generally considered to have poor sensitivity. Direct measurement of the pressure in the second stage chamber could be accomplished by fitting the valve with a pressure transducer. Although this invasive technique can be accommodated in a laboratory experiment, its implementation for plant use would require substantial additional safety analysis and testing of the valve's integrity. The monitoring of the leak noise transmitted to the external wall of the pilot stage is a highly sensitive non-invasive method. A valve having a leaking pilot-disc seat was tested under laboratory conditions in order to determine the correlation of the leak noise with leak rate or second-stage, pressure. As the second stage pressure increased from 40 to 200 psi, the leak rate increased 59%,

the temperature increased 9% and the acoustic RMS voltage increased 370%. Therefore, the sensitivity of the acoustic detection is excellent. Furthermore, if the leak were intermittent, the, acoustic detector would respond accordingly while the thermocouple would not. These measurements provided additional encouragement for utilization of the acoustic detection method.

The basic components of the TEC Model 1400 surveillance system are a multiplexer, signal conditioner, microprocessor and display. Under a combination of pushbutton and software control, the system displays data recalled from computer memory through front-panel keypad commands. Additional data handling and processing are provided through an optional floppy-disc data storage device, terminal and hard copy unit.

The RMS values of each input signal are sampled at a typical rate of 1 per second. Each sample consists of two values: namely, the RMS value (x) in bandwidth A and the RMS value (y) in bandwidth B. The bandwidths of the programmable filters are selectable over the range 1 kHz-15 kHz via TTY or terminal keyboard control. The computer memory stores the last 60 values of x and y for each channel, as well as the last 100 hourly averages of x, y, and the variance (a') in x. In addition, every hourly average value is preserved on the disk for every channel. One disk can store up to 120 days worth of hourly records.

Using the keypad, the last one-minute or 100 hour0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> history of x, y, and a'or any channel can be displayed on the system's built-in CRT. Also, whenever an x value exceeds a preselected warning level, an automatic display mode is activated. Ordihary operation is returned only upon acknowledging the warning display from the keypad.

Detailed technical specifications and prices are available upon request.

FOR FURTHER INFORMATION CONTACT'echnology for Energ'y Corporation 10431 Lexington Drive Knoxville, Tennessee 37922 (615) 966-5856

~ ~

Customer Technology for Energy Corporation 10431 Lexington Drive Bulletin Knoxville, Tennessee 37922 (615) 966-5856 TEC 900/901 SIGNAL CONDITIONING AMPLIFIER SYSTEM i ~ y5.

~ TEC ~ TEC ~

TEC ~ TEC ~ TEC

~ ,

4

~:g.

.4

.4$

FEATURfS.

~ Ultra-Low Noise, Ultra-Low Drift

~ Differential or Single Ended Input

~ Bandwidth DC to 50 kHz

~ Total Gain 1 to 100,000

~ AC Coupling from .001 Hz

~ Variable DC Offset The TEC 901 amplifier, designed to satisfy a wide range of linear amplifications, provides high or low amplification of either dc or ac signals with excellent linearity, low drift, and fast recovery. Control features on the front panel include differential or single-ended operation, first- and second-stage gain, bandwidth control, and overload indications. The modular design using high-quality integrated circuits makes the TEC-901 a low cost, compact, and highly versatile signal-conditioning amplifier. lp The TEC 900 Pow'er Bin is designed to accommodate eight (8) TEC 901 plug-in amplifiers.

The bin measures 5/4I'igh, 19" wide, 12" deep. Power is supplied to reserved connector pins from three separate regulated supplies sharing a common source.

An.optional eight (8) coaxial cable signal harness that connects directly to the power bin front panel and terminates with BNC connectors is available. The standard harness is ten feet Different lengths are available. in'ength.

'I Serving the Energy Production Industry

TEC 901 AMIQFIER SPECIFICATIONS ~

Inputs Differential or Single-ended TEC MODEi

~ Impedance Single-ended: 1 megohm (min)

Differential: 2 megohms (min) 901 LOW-PASS Maximum Input Signal Single-ended: +15 V Differential: +10 V OVL Common Mode Rejection 80 dB (min), 90 dB (typ) at 1st stage gain = 10 5'UTPUT Settling Time 30usec N Gain = 100 for+10 V Gain 2nd GAIN 1st Stage: 1, 2, 5, 10, 20, 50, & 100 TAIM 2nd Stage: 1, 2, 5, 10, 20, 50, 100, 200, 500, &

0 1000 Accuracy better than 1%

~ Frequency Response DC: to 50,000 Hz AC: .001 to 50,000 Hz e.. Filters High Pass: 0.001, 0.01, 0.1, 1, 10, & 100 Hz (switch selectabie on board)

I Low Pass: 10, 20, 50, 100, 200, 500, 1,000, 2,000, 5,000, 10,000, 20,000, & 50,000 Hz

+ CPUT Output Linear: 110 V Impedance: c10 ohm Overload Indications Adjustable from 1 to 10 V TEC 900 POWER BIN SPECIFICATIONS 1st Stage Noise Output Voltage. +15 VDC G = 100 (referred to input)

Output Voltage Accuracy ........a1% 0.01 to 10 Hz: 2 Ijv p-p (max)

Line Regulation ...........................02% (max) 10 to 10,000 Hz: 3 Ijv rms Output Current ............................ 600 ma Output Ripple 2.0 mv rms (max) G = 1 (referred to output)

Temp Coefficient ........................ 0.2%'C 0.01 to 10 Hz: 40 pv p-p (max)

Operating Temp ..........................-15'C, + 71'C

'0 to 10,000 Hz: 50 pv rms Source Voltage ............................105 to 125 or 205 to 240 VAC 2nd Stage Noise 50 to 400 Hz G = 100 (referred to input)

Input Power Cable...................... Standard 3 Prong 0.01 to 10 Hz: 40 pv p-p (max)

Plug 10 to 10,000 Hz: 3 pv rms FOR FURTHER INFORMATION, CONTACT:

Marketing Manager Technology for Energy Corporation 10431 Lexington Drive Knoxville, Tennessee 37922 (615) 966-5856

0 Customer Technology for Energy Corporation 10431 Lexington Drive Bulletin Knoxville, Tennessee 37922 (B15) 958-585B THE TEC MODEL 1430 LOOSE-PART DETECTION SYSTEM Th'e TEC Model 1430 Loose-Part Detection System (LPDS) was developed not only to meet the requirements of NRC Regulatory Guide 1.133, but also to provide for special diagnostic proce-dures which enable confident evaluation of data. The Model 1430 Loose-Part Detection System is based upon the TEC Model 1400 Surveillance System pictured below. The modular system con-tains complete signal conditioning and data processing instrumentation needed for a quality-assured Loose-Part Detection Program.

Meets NRC I hhil ilil:I hhil W f Microprocessor Controlled hhl1 il;h Trend Analysis Requirements h v'

nlhh hhh I 4

A~ ~ Ay Vi VV VVVV Automatic Display of 4 Minimizes False Alarms Pertinent Data Data Records on Special Diagnostics Floppy Disk Available 11-CHANNEL VERSION OF..TEC MODEL 1400 SURVEILLANCE SYSTEM In addition to supplying the hardware and software that comprises the 1430 system, TEC offers consulting and field measurement services to assist the customer in implementing the full Loose-Part Detection Program.

Serving the Energy Production Industry

FEATURES:

~ Monitoring of Event Related Parameters The peak signal amplitude and the rms signal level for each channel are monitored, read into the microprocessor, statistically correlated, and stored on disk for subsequent trend analysis. This eliminates the need for analog-tape recording of signals.

~ Sensor Channel Performance Checks Microprocessor controlled signal trend analysis is used to test for and identify degraded channel performance. This approach greatly simplifies the "once every 12 hour1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> channel checks" required by NRC Regulatory Guide 1.133.

~ System Sensitivity The system sensitivity is commensurate with the goals established by the NRC Regulatory Guide 1.133.

~ Alert Level The alert level is optimized for each channel by digital analysis of the signal history at all oper-ating conditions. The system, therefore, remains operable during start-up periods when loose-part impacting is most likely to occur. The signal histories aid the operator in meeting the "at least once per 92 days" requirement for verifying that the alert-level is consistent with the "normal" background noise as called for in Regulatory Guide 1.133.

~ False Alarm Minimization A Deliberate-Plant-Maneuver Monitor informs the microprocessor that plant operating condi-tions are changing and momentarily disables the alarm. Because the system remains active, detection of impacts resulting from these maneuvers serve as an automatic check of the sys-tem's operability.

~ Records All data records (time, events, peak amplitudes, rms levels, etc.) are stored on a floppy-disk that can be periodically removed, filed, and replaced with a new disk. By using this simple and inexpensive method, a permanent history of the plant's LPDP is maintained in a compact form.

~ Visual Display of Recorded Data A visual display of the recorded data for each channel is obtained by entering the channel number and file record of interest on the front panel keyboard. The information is automatically displayed on the CRT.

~ Options a) A switch selectable audio monitor allows the operator'to listen to the output of any channel. 'i')

Custom'oftware for additional diagnostic capability.

c) Automatic impacting devices for system checkout and calibration.

FOR FURTHER INFORMATION CONTACT'echnology for Energy Corporation 10431 Lexington Drive Knoxville, Tennessee 37922 (615) 966-5856

Customer Technology for Energy Corporation 10431 Lexington Orive Bulletin Knoxville, Tennessee 37922 (615) 966-5856 t

Sensor Time Response Testing Time response testing is required by the Nuclear Regulatory Commission through Regulatory Guide 1.118 for all components in reactor protection systems including the plant sensors. Technology for Energy Corporation (TEC) offers",complete consulting engineering services, field measurement services, and special instrumentation to assist thc electric utilities in meeting these testing requirements.

TEC is the nuclear reactor industry leader in the implementation of advanced measurement technology, particularly in th<<surv<<illanc<<and diagnosis ol'plant anomalous performance and component malfunction.

From this broad experiential base, we have developed a complementary set of measurement techniques and instrum<<ntatinn for pcrfnrming low cost quality-assured sensor time response testing on RTD's and pressure sensors in nuclear power reactors.

RTD We have implemented three measurement Self-Heating. This test is just the steady-state techniques and the related instrumentation for version of the loop current step response method.

in-situ RTD time rcsponsc tcstin in a>>u<<l<<ar

~ Thc current in the RTD is externally increased, power plant: noise analysis, loop current step but in this case measurements arcn't made until response, and self-heating. Hence, we can help steady state is attained. Using this procedure, you choose thc measur<<m<<nt technology that thc change in thc RTD's resistance is measured best matches your nccds l'rom the standpoint such that a characteristic resistance-versus-cur-design, instrumentation channel <<rrangc-ol'ensor rent curve can be constructed whose slope is ment, and operations philosop)>>. Th<<basic related to the sensor time response.

principles of these measurement techniques arc as follows.

Pressure Sensor Noise Analysis. Noise analysis uses the We have implemented an extension of the naturally occurring random fluctuations from the pressure sensor testing technique that was sensor to determine its response time. These developed under EPRI auspices. This technique fluctuations are spectrally analyzed to construct involves the introduction of an externally applied thepower spectral density (PSD) which is related pressure perturbation by an hydraulic signal to the sensor transfer function. The PSD is then generator. Following this perturbation, the transformed by autoregression analysis to give response of the pressure sensor under scrutiny is the equivalent "plunge test" result. intercompared with that of a reference fast-Loop Current Step Response. The loop cur- response pressure sensor to confirm adequate rent step response technique, developed under response time..

EPRI auspices, involves the measurement of the RTD output when its current is cxtcrnally For furth'er sensor time response testing increased rapidly from its normal value of a few information or for details about other TEC milliamperes to a level of tens of milliamperes. technical services or instrumentation, contact:

I. Norbert J. Ackermann, Jr.

The resulting resistive heating in the RTD causes a temperature rise of a few degrees. The RTD Technology for Energy Corporation time response may be 'etermined then by 10431 Lexington Drive observing the response of the sensor to this Knoxville, Tennessee 37922 perturbation and transforming the results to 61 5/966-5856 an equivalent "plunge test" result.

Serving the Energy Production Industry

Customer Technology for Energy Corporation 10770 Dutchtown Aoad Bulletin Knoxville, Tennessee 37922 (615) 966-5856 IN-SITU RESPONSE TIME TESTING OF TEMPERATURE SENSORS

>EC qgTEC ~ QT EC iQ TEC MODEL 1100 LOOP CURRENT STEP RESPONSE MEASUREMENT SYSTEM FEATURES:

~ Determine response time without removing RTD

~ Meets recommendations of NRC Regulatory Guide 1.118

~ Constant current source cannot damage RTD

~ Adjustable for any RTD resistance

~ Automatic or manual testing capability TEC offers the Model 1100 system to utilities making RTD sensor response time measurements in accordance with NRC Regulatory Guide 1.118. The Loop Current Step Response (LCSR) technique uses technology which was developed under DOE funding at the Oak Ridge National Laboratory and further refined under EPRI Research Project 503-3. The LCSR method is the measurement of an RTD's response when its current is externally stepped from its normal value of a few milliamperes to a level of tens of milliamperes. The resulting resistive heating in the RTD causes a temperature rise of a few degrees. The RTD response time is determined by observing the response of the RTD to this external current and transforming the observed result to that of an equivalent plunge test.

TEC offers a full set of instrumentation, test procedures, and analysis software to accurately and efficiently perform and interpret the LCSR measurement. TEC offers supporting field measurement and consulting engineering services to further assist the utility in planning, performing, interpreting, and reporting the measurements.

Serving tbe Energy Production Industry

MODEL 1100 SYSTEM DESCRIPTION The Model 1100 system is housed in a bin which measures 5/~" high, 19" wide and 18" deep. The LCSR testing unit is incorporated in the four right hand modules. Blank panels cover positions that are available for other TEC instrumentation modules. A half wide bin is also available which contains only the LCSR testing unit.

Model 1101 Floating Wheatstone Bridge/Amplifier Module. The bridge/amplifier is balanced via front panel controls and has selectable amplification of the true bridge imbalance voltage of 20, 50, 100, 200, and 500.

TEC's design has several notable features. The most important is that self-heating of the LCSR instrument does not affect the measurement accuracy. A unique design approach resulted in a bridge resistive temperature coefficient less than 0.1 ppm/mw-'C. The bridge is floated above ground so that inadvertent disconnecting of the RTD at any stage in the test will not damage the bridge amplifier. For RTDs in the 102 ohm to 547 ohm range (expected range for reactor applications) the pre-test nulling of the bridge has been simplified by a 31 position coarse adjustment coupled with a 10 turn, locking fine adjustment. Another feature is that TEC includes with each bridge the calibration data necessary for determining in the test the actual resistance of the sensor. This is accomplished by nulling the bridge, reading the coarse and fine settings, and consulting the calibration data.

Model 1131 Constant Current Source Module. The Model 1100 uses a constant current supply because a variable voltage supply could lead to oversupply of power and possibly damage the RTD when testing it cold. The constant current source has an output range that is fully adjustable from less than1 ma to 100 ma and is preselectable at both low and high levels to provide a consistent, well-defined step function.

Model 1121 Control Module. The control module can be operated in a manual mode via front panel switch-ing or an automatic mode via remote TTL compatible logic from a computer or signal generator. In the automatic mode, the test is initiated by providing a TTL logical high level (+5V) at the "Auto Step" input terminal on the front panel. This effectively closes a relay which increases the current to the bridge from a few milliamperes to some desired current level (typically tens of milliamperes) chosen by the user and dependent upon the design limitations of the sensor being tested. The return of the voltage level at the "Auto Step" input to a logical low(<0.8V) terminates the test. The manual mode performs the same func-tions, but the test is initiated via the front panel switch. In either mode, an adjustable output (0 to+5 VDC) is provided by the control unit when the test is initiated and is removed when the test is terminated. This output signal feature allows the user to identify when the test started and stopped; it also provides a means of remotely controlling digital analysis equipment if so chosen. If a recording of the LCSR test results is made, this output signal would also be recorded to give a timing reference for the initiation and the termin-ation of the test.

Model 934 Power Bin Module. Three power supplies provide separate power to each of the above modules.

1 FOR FURTHER INFORMATION, CONTACT:

Marketing Manager Technology for Energy Corporation 10770 Dutchtown Road Knoxville, Tennessee 37922 966-5856 '615)

LCSR HISTORY

1. BASIC DEVELOPMENT ORNL/DOE R. L. Shepard R. M. Carroll

2. LWR APPLICATION RESEARCH EPRI/UT T. W. Kerlin

3. COMMERCIAL APPLICATION TEC J. E. Mott J. C. Robinson J. E. Jones R, K. Fisher M. V. Mathis

A N

8O D C7 D

0 l 6'0 8 0 6'0 S'0 Q 0  %'0 8 0 E 0 T'0 0 0 38ALU83dH31

T.

i Ae t + Ae -A t

+

T l 1 2

2

~1+ ltd A2, X , etc. are the same for Plunge & LCSR A's are different To determine plunge time constant from LCSR test we must (l) Get good estimates of Al X2 from test results, (2) Determine Al.

TIME CONSTANT DEPENDENCE ON hR k

T-T.i -A t Ae 1 + Ae2 2 +

T f T.i hR (t/t )

k (1/~1) 0 1.00

.Ol 1.008 0.1 1.024 1.0 1.169 0.27 10.0 1.460 0.82 50.0 1.471 0 '6 1.471 NOTE: Combustion Sensor h > 2400 ~ hR > 50 k

SENSOR TESTING RESULTS Yew sensor and well plunged 180'F, 3 Ft/Sec 2.89 Sensor installed in plant Good Fit {6 = 0) Poor Fit (8 = .001") Damaged Sensor {6. 0) v ~ 1.70 sec. v = 4.80 sec. v = 2.80 sec.

Six months later 2.88 sec. x = 5.59 sec. ~ = 3.70 sec.

Back in Lab with mismatched well t = 6.8 sec. v = 6.8 sec. 9.6 sec.

air gap'latinua

" sensing element oxide sheath and well

COMPARISON OF PLUNGE SENSOR TESTING METHODS hR/k for Various Test Conditions Reactor 3 Ft/Sec 1 I t/Sec Sensor Conditions 180'F Water 500 F Lead- Comments Combustion 300 27 115 Internal Resistance dominates for all tests.

Westinghouse 3.8 0. 34 1.5 3 Ft/Sec has wrong mechanism dominating, even liquid metal does not simulate well.

Q t\ \~ ~ ~ ~ ~0 ~ ~ ~ ~ 0 1~ ~ ~ ~ ~ g' ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

u ~ 0 ~ ~ 0 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~, ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~ ~

4

~ too ~ ~ o oo ~ e ~

JJ lr 5 ~ ~ ~ ~ ~ 0 ~ ~ ~ ~ ~ ~ ~ ~

Jl O

u'4 ~ 0 ~ ~ ~ ~ t ~

0 3 ~ tt ~ 0~

+oooo ~ o~o ~ OO \ ~ ~ ~ ~ ~

0 10 20 30 40 50 60 Time in Seconds FiS ~ 5.l Typical Output of an RTD Follovire a Sten Change in Current from 2.2 ma to 35 ma.

h

o~ tet to t t o t t o 0 t ~ ~ester as ~p r ~ ~ ~ s sss e t t p ~ ~ ~ 0 t s ~ ~ s ~ ~ ~ ~ t ~ r e ~ ~ ~ ~ o ~ \ s s e ~ ~ s ~ ~ t r s r o t ~ s Q s s r te r e r o t

5 s ~ store ) ~ ~ ~ ~ ~ t ~ ~ s ~ ~ p s ~

) ~ ~ ss s ~ ) ~ ~ ~ ~ sts ~ j ~ ~ ~ e ~ ~ ss fto ~ ~ oteot attest

~

C r t te e

eosoto ~ ~~~ ~ so ~ bs r ~ A s pt ~ ~ s ~ ~ o ~ s ~ ~ o ~ ~ s ~ sr ~ ~ et ~ too o io t r or o oe 0 ee stet e

o 4

>>st ~ ~ p ~ ~ s s~~~s t ~ e ~ ~ \~~ ~ ~ ~s ~ ~ ~ ~ t~ Q ~roe e

sot ~ )~~e ~o~ ~~ i-- ~s s ~ ~ s qs ~ s ~ ~ s ~ r ys ~ eoo~s s p s ~oroe s s j s s s ~ ~~~ ~

t ~ s e o ~ o e o g e o o ~ o r o o Q aa oOoo O s r

s s IQ 3 tote rsfssooeeos leo o o o o o s J>> s ~ s s s s o A p s o t s ~ ~ l ~ ~ ~ sss s ~ ~ s ~ ~ ~ ~ rs ~ s ~ p ~ooe s o

s s ~oooos f>> s ao ~ o s s e

~o ~ 0 t et ~ ~ t ~ ~ ~ ~ ~ ~~~ r o ~ ~ ~ ~ ~ ~ ~ s 0 o ~ ~ s s t ~ s ~ ~ o r ~ ~ ~ s s t ~ o ~ s ~ s ~ s ~ s ~ ~ r ~ p e ~ ts ~ s o ~ t o ~ Shet o~ '

t e P e e

o t otto) ~ ~~~ oooo P o o t tt t Q t ~ s ~ s \ t Q s ~ \ ~ s r s ef ~ ~ s ~ s~ ~s f ~

s e ~ e ~s ~~ ~~ t o o ~ o t t p o t t o t e t Q aea t s o e

~ ~

t topee)seor e

~ oo ~ fso steers+so e

~ ~~~

~.

spot ~ ~ t ~ ~ s) ~ ~ sot ~ p ~

} sop ~~~ ss J t

~~ tsetse frosts ~ to ft teeter P

0 12 15 18 21 24 27 30 Time in Seconds Fig. 5.2 Average II LCSR Test Results for 14 steps.

LUMPED PAIUQKTER TRANSFORMATION

1. Requires No Knowledge of Sensor

2. Has Non-Conservative Bias i.e. One Eigenvalue A ~ 1 1

Two Eigenvalues 1

1 1-Al Eigenvalues Gives exact value of Al for center located sensing element

3. Does not give credit for sensing element near surface

>1

'I l Il

~ >> ~ ~ I f I

~ ~

I III Il!

I>>

'>> ~ I

>>:l>>

'r I>

I I

fiI I

I I:, 1

! I)II ip~ r I

r ll

>>I I i;Il I!,I[

j! I)

>>'/'li! I,'>>l

.II i!!l

CONTINUOUS MODEL (1) Requires Knowledge of (a) Thermal properties of sheath and well (b) Sensor dimensions and corresponding uncertainties (2) Requires Two Eigenvalues (3) No Bias Error (4) Calculable Error (5) Accounts for Sensing Element Location

TABLE 5.1 EFFECT OF 'to ( INITIAL T IME I FOR ANALYS S ) ON THE T IME CONSTANT T

True = 5.25 sec true T trans g lpp to (sec) / Error <true 2nd ORD ER FIT 0 41.3

.25 21.2

.50 10.2

.75 4.4 1.00 2.0 1.25 0.2 1.50 -0.4 1.75 -0.4 3rd ORD ER FIT 0 13.5

.25 1.4

.50 .5

.75 .6

~ -(o) yr

~ 3'

I P

O nO r'NITE STATES NUCLEAR REGULATORY COMMISSION

+n i

ill

)i t

~O WASHINGTON, D. C. 20555 e/~k

+%*5" y'W co gJ~ dC 4C dCCr ~~

+o -~z>

h C

L

'l~

5i

('v