ML18024B185
| ML18024B185 | |
| Person / Time | |
|---|---|
| Site: | Browns Ferry, Oconee  |
| Issue date: | 11/16/1979 |
| From: | Clark K, White T TENNESSEE VALLEY AUTHORITY |
| To: | |
| Shared Package | |
| ML18024B184 | List: |
| References | |
| NUDOCS 7912270262 | |
| Download: ML18024B185 (74) | |
Text
NUCLEAR CONTAINMENT TESTIQ FOR TVA NUCLEAR POWER PLANTS Kenneth H. Clark Mechanical Engineer Tennessee Valley Authority Chattanooga, Tennessee Timothy J. White Mechanical Engineer Tennessee Valley Authority Chattanooga, Tennessee Presented to the American Nuclear Society Winter Meeting November 11-36, 3979
TABLE OF COHTENTS Fsge Xntroduction Test Obgectives Test Criteria 5'0 Techniques of Analysis Data Acquisition and Reduction Systems Instrumentation Techniques 13 System Software Discussion of Test Resu1ts 19 Conclusions 22
INTRODUCTION A significant part of the surveillance requirements f'r a nuclear power plant involves the assurance of'solation of radioactive contaminants from the environment in the event of a radiological accident. 'The primary containment serves as the final barrier of isolation in an accident. General Design Criteria 54 and. 56 of Title 10 Code of Federal Regulations, Part 50 (10 CFR 50),
specify design provisions for the reactor building primary containment.
Appendix J to 10 CFR 50 defines the basis for a surveil1ance program to ensure that the primary containment will perform as designed for the life of the plant.
I The most most severe significant test prescribed by Appendix J, the reactor building containment integrated leak rate postulated accident.
test, involves simulating The leakage as close as practical the predicted conditions within the primary containment after the of air from the primary is containment to the environment is measured to demonstrate that offsite exposure to postulated radioactive contaminants will not exceed 10 CFR 50 guidelines, as implemented by the plant technical specifications.
Since the publication of Appendix J to 10 CFR 50, it has been customary to conduct reactor building containment leak rate tests (CILRT's) for at least, 24 hours. This practice originated from experience gained in the OHNL-AEC containment proof program. The current national standard for the conduct of the CILBT, ANSI 45.4-1972, recommends tests be conducted. for ". . .not less t
than twenty four hours of retained pressure. . ." This arbitrary test duration was set as a means to ensure the primary containment leakage would be accurately measured, with the instrumentation typically in use when the standard was prepared.
Experience gained by the Tennessee Valley Authority in the conduct of CllHT's has demonstrated that the primary containment leak rate may be accurate~ measured for tests conducted for considerably less than 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.
The purpose of this presentation is to discuss the techniques, equipment, and method of ana+sis TVA proposes to use to conduct future CXLRT's of shorter duration than current practice. Data collected. from two CILRT's conducted for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> with the techniques and equipment described by this paper are discussed.
TEST OMECTIVES A. General The reactor building primary containment is designed to prevent the release of'adioactive contaminants to the environment either in normal operation of a nuclear power plant or as the consequence of an accident.
Plant site meteorological conditions determine from the guidelines presented in 10 CFR 100 a maximum amount of radioactive contaminants that may be released to the environment.
Various plant design and reactor specific features determine a predicted maximum pressure expected to exist within the primary containment under accident conditions and a maximum rate of release of radioactive contaminants to the environment. Appendix J requires that the plant operator periodically demonstrate the ability of the primary containment to limit the release of contaminants below the calculated, maximum.
The CXLHT measures the rate of release, or the leak rate, of'he primary containment atmosphere to the environment at a test pressure of'ither one-half or equal to the calculated peak pressure expected for the most severe accident. Lines that penetrate the primary containment are aligned with the configuration assumed. automatically after an accident. Lines postulated to rupture inside the primary containment are drained, to the extent practical of fluid and vented to the containment atmosphere for
'the duration of the test. Lines postulated to rupture outside the primary containment are drained to the extent practical of fluid and vented to the environment.
Before a nuclear power plant may return to operation, the CILRT must 0 demonstrate that of the design this maximum.
measured The rate of leakage is less than 75 percent 25-percent margin provides assurance that, with unforeseen degradation of performance, the maximum leakage will not be exceeded.
B. Specific Objectives The specific objectives of the CILRT are:
- 1. Accurately measure the actual rate of primary containment atmosphere leakage under conditions close to those predicted for the most severe postulated accident.
- 2. Demonstrate that the primary containment leak rate has been accurately measured by the CILRT by a subsequent verification test.
Demonstrate that the measured rate of leakage is less than 75 percent of the design maximum before the nuclear plant may return to power operation.
- 4. Demonstrate that no potential means for the release of primary containment atmosphere has arisen since the previous CILRT.,
- 5. Provide a statistical statement of the validity of the measured, leak rate of the primary containment by calculating the confidence interval of the results.
TEST CRITERIA A. Primary Containment Atmosphere Stabilization During the pressurization of the primary containment, the containment atmosphere temperature wiU. significant3g increase. This heating, due to the work required to pressuri.ze the air, can introduce instabilities of the containment atmosphere that may preclude the accurate measurement of leak rate. In a similar manner, the operation of large equipment withi.n the containment can cause the apparent leak rate to change during the CILRT.
Appendix J requires that the primary containment atmosphere be allowed to stabilize at least 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> after the end of pressurization. This arbitrary requirement can prove of insufficient duration particularly when appU.ed to high-pressure, smaU.-volume containments. From the experience gained-in the conduct of six CILRT's, the foUowing guidelines were prepared to supplement the Appendix J requirement:
- 1. The average primary containment atmosphere temperature change should be less than loF per hour before starting the CILRT.
- 2. A time versus temperature plot for the stabilization period should be approximately linear by the start of the CILRT.
- 3. Heat-producing equipment located within the primary containment should only be operated to maintain the safety of the reactor.
- 4. Any air circulation equipment operated during the CILRT should be operated continuously since i.ntermittent operation could. disturb the containment air temperature distribution.
e 5. Water levels within the reactor and, any other vessel within the primary containment should be held as constant as is possible.
required level changes should be made slowly.
Any B. Accuracy of the Measured Leak Rate Since any measurement has some degree of uncertainty associated with random and systematic errors, the reported measured. leak rate of the primary containment atmosphere is only an approximation to the "true" value. A statement of the goodness or degree of confidence of the CILRT results is necessary to provide assurance that the primary containment functions as designed. Follows.ng general testing practice, TVA reports a 95-percent upper confidence level for the reported leak rate.
h CILRT is 'considered satisfactory if the measured leak rate is less than 75 percent of the design maximum. To ensure adequate confidence in this leak rate, TVA further requires that the 95-percent upper confidence level be less than 75 percent of the design maximum leak rate.
e A. Containment Modeling TECHNIQUES OF ANALYSIS The accurate measurement of primary containment leak rate pivots on the precise measurement of temperature, pressure, and vapor pressure. The primary containment is not constructed. as a single homogenous pressure vessel but as a series of interconnected compartments. Although all compartments forming the primary containment are vented to each other for the CILRT, the flow of containment atmosphere may be restricted.
Pressure suppression containment designs incorporate special compartments that may have significantly different temperature and vapor pressure conditions fxom the rest of the primary containment. A boiling water reactor pressure suppression chamber is characterized by humidity approachinG the saturation point. The ice condenser for a pressurized water reactor employs two large compartments far below the fxeezing point of'ater. Since a substantial portion of the primary containment fxee air volume is contained within these pressurization suppression compartments for both reactor designs, significant errors may result in the calculation of the leak rate if the containment atmosphere conditions are not correctly considered by the analysis.
To compensate for the compartmental construction of the primary containment, the leak rate is calculated fxom a model in which the containment is a multiple element system. Temperature, pressure, and vapor pxessure are measured for each compartment. The mass of the air in each compartment is calculated from these measurements. The primary containment leak rate is calculated from the sum of the compartment air masses. Temperature, vapor 0
pressure, and pressure measurements are individually assigned volumetric weighting or influence factors determined by the relative volume each sensor represents within the compartment.
A primary containment model is developed. from information provided in section 6.2 of the Final Safety Analysis Report. Any compartment that represents more than 10 percent of the containment free air volume is considered a compartment for the CILBT.
H. l"lethod of. Leak Rate Calculation Several techniques have been used previousIy to calculate the primary containment leak rate. ANS 45.2-1972 recognizes the absolute and the reference vessel methods. The proposed. standard for containment testing, ANSI 56,8, recognizes the seme techniques, We have found, the absolute, or mass loss, method yields the most accurate measurement of the primary containment leak rate.
The primary containment leak rate is calculated by the application of the ideal gas law. 17uring the CILHT, the mass of the air in the containment is calculated. periodically. The leak rate is computed from the slope of the least squares fit line to these data. The uncertainty of the measured leak rate is estimated by calculating the deviation of the individual mass points from the least squares fit line, with adjustments for the sample si,ze ~
C. Instrumentation Selection Guide The, accurate determination of leak rate by the absolute method requires the precise measurement of primary containment atmospheric temperature, vapor pressure, and total pressure. Since any measurement will include some error, the accuracy of these measurements determine the accuracy of the measured primary containment leak rate. Prior to the performance of
the CILRT, the number of temperature, vapor pressure, and total pressure sensors required to accurately determine the leak rate must be estimated.
Based on the expected leak rate and the anticipated conditions encountered 1
in r
the test, this instrumentation selection guide determines the minimum instrumentation necessary to conduct the CILRT.
The basic criteria TVA uses for the selection of minimum CILRT instrumen-tation is that the primary containment leak rate should be accurately measured within the first 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> of data collection with an assumed, leak rate equal to 25 percent of the maximum allowed under technical specifica-tions. In addition, 'no temperature measurement may represent more than 10 percent of the containment free air volume. Appendix A presents an example of the estimation of sensors required for a typical boiling water reactor CILHT.
<<10-MTA ACQUISITION AND HEDUCTION SYSTEMS The precise measurement of many test variables is required to accurately calculate the primary containment leak rate. CILRT test data must~ therefore, be acquired and analyzed rapidly. TVA has developed a leak rate measurement system that acquires and reduces test data automatica~. The principal advantages afforded by this automatic system are highly accurate, reliable results and data collection speed. The purpose of this section is to describe the principal functions and features of the automatic data acquisition and.
reduction systems.
A. Data Acquisition System The principal function of the data acquisition system is to periodica1lg measure the test variables. A microprocessor controls the timing of periodic acquisition, the conversion from analog to digital values, and the transmission of data to the'data reduction system. The microprocessor will periodically collect data at a set interval or, at the discretion of the test director, can be demanded to acquire data within the selected.
interval. A log of all collected data is printed f'r permanent records.
Table 1 lists typical data collected. for a boiling water reactor and a pressurized water containment. The data acquisition system is designed, to allow for any combination of temperature, pressure, and vapor pressure measurements. Figure 1 depicts the components that form the data acquisition system.
The principal feature of the data acquisition system is the accurate, rapid measurement of test variables. In CILHT's previously conducted by TVA without the automatic data acquisition system, data could not be collected more frequently than once per hour. Even at this slow rate of
collection, mistakes by test personnel in the measurement of test variables degraded the results. For a typical ice condenser pressurized. water reactor, the data acquisition can collect up to 20 samples of the test variables per hour. The significant increase in the volume of collected, data improves the confidence of test results.
B. Eata Reduction System The primary purpose of the data reduction system is to accurate+ perform the necessary calculations to compute the primary containment leak rate.
The central element of the data reduction system is, a minicomputer system directly connected to the data acquisition system. All raw data collected by the data acquisition system is transmitted to the minicomputer and stored on flexible disks. These data are subsequently'orrected according to each sensor's calibration data. The leak rate is automatically calculated and results are printed on a local printer. The system is designed to be tolerant of power failure. Figure 2 depicts the data'eduction system.
Several features are included in the design of the data reduction system.
The most significant is that the reliability of field test results is significantly enhanced because no manual data entry or calculations are required,. The speed of data reduction is significantly increased. For a typical ice condenser, pressurized water reactor data can be coU.ected by the acquisition system, stored, reduced, and the leak rate calculated, in less than 2 minutes.
In addition to speed, the minicomputer offers several features to enhance test performance. Test variables or results may be automatically plotted.
by the minicomputer any time during the CILRT. The test engineer may also choose to redefine the time of the test start to any previously collected
II sample while the test is conducted. This "base reset" feature allows the field evaluation of the effect of prolonging test duration.
INSTRUlKNTATION TECHNI UES A. Temperature Measurement Four-wire resistance temperature detectors (RTD's) are used by the leak rate measurement system to monitor primary containment atmosphere temperature. Before and after the performance of each CILRT, each RTD is individualIy compared. with a standard certified by the National Bureau of Standards over a temperature range of 0-150 F. The uncertainty of the temperature standard is better than 0.005 F. A unique temperature as a function of resistance calibration curve is calculated. for each RTD from this comparison.
When installed in the primary containment, each RTD is connected to a separate excitation bridge (wheatstone) by quick disconnect extension cables. Systematic errors due to lead length resistance, excitation bridge nonlinearity, and analog to digital conversion repeatable offset error are measured by substituting precision resistors in place of the RTD at the end of the extension cable., A unique resistance as a function of measured bridge output calibration curve is calculated for each measurement channel. The minicomputer, automatical+ calculates and stores 4
each calibration curve. For each temperature measurement, measured bridge voltage is first converted to resistance. The minicomputer then uses the individual RTD calibration curve to calculate the equiva1ent temperature from this resistance.
Tests have been conducted to determine the accuracy of temperature measurements by the integrated leak rate measurement system. Seven RTD's were compared. with a standard certified by the National Bureau of Standards at five temperatures. This standard. is certified. with a measurement
lip uncertainty of better Chan 0.005 F. Figure 3 depicts the difference between the temperature measured by the standard and the leak rate measurement system over the range of comparison. An~sis of the data indicates that the system uncertainty of temperatuxe by the leak rate measurement system is better than 0.0202 0F.
B. Vapor Pressure Measurement Lithium chloride dewcels are used. by the leak rate measurement system to monitor primary containment atmosphere moisture content. The principle of operation of a dewcel is that certain hygroscopic salt solutions will change the amount of water in the solution in relation to the moisture content of the air. The dewcel consists of a thin coating of lithium chloride between two gold wires. As the moisture content of the air changes, the salt solution will either absorb or liberate water. This change in moisture content of the salt solution changes the solution resistance proportional+. Passing a constant voltage through the two wires and the solution causes resistance heating. An RTD embedded in the support bobbin measures the induced heating. Since the temperature of the solution is direct~ xelated to the solution resistance, and hence the moistuxe content of the salt solution and the air, it is necessary only to measure this temperature to measure atmosphere moisture content.
Three-wire RTD's monitor the salt solution temperature. Before and after each CILRT, each dewcel RTD is individually compared with a standard certified to the National Bureau of Standards over a temperature range equiva3.ent to a dewpoint from 0 0F to 0 100 F. A unique temperature as a function of resistance caU.bxation curve is calculated fox each dewcel RTD from this comparison.
Each dewcel is connected to a separate excitation bridge (wheatstone) and constant voltage power supply by quickie disconnect extension cables.
As in the discussion of the air temperature measurement, a calibration curve of resistance as a function of measured bridge output is calculated by the substitution of precision resistors for the dewcel. Each dewpoint is first converted to equivalent resistance. The minicomputer then calculates the salt solution temperature from the dewcel's unique element temperature as a function of resistance curve. Equivalent dewpoint is calculated from data tabulated. by the National Bureau of Standards.
C. Pressure Measurement Precision quartz bourdon tube manometers were selected for containment total pressure measurement. Prior to the CILRT, a pressure cell is selected, so that the rated pressure is gust above the expected test pressure. Each manometer and, cell is compared with a standard certified'y the National Bureau of Standards before and. after each CILRT over the range of the pressure cell. Proper selection of the pressure cell ensures the highest possible sensitivity to sma11 changes of the primary containment pressure. The pressure measured by the quartz manometer is converted internally to digital values by a special encoder. The rated cell pressure corresponds to a digital output of four hundred thousand counts, with a resolution of one count.
To convert the digital signal acquired from the quartz manometer to pressure, the minicomputer linearly interpolates the true pressure from the pressure ceU. calibration data. This technique yields a certified system accuracy of better than 0.015 percent of reading.
t D. Calibration after of'est All instruments Instruments included in the leak rate measurement system are compared with standards traceable to the National Bureau of Standards prior to to be out of tolerance in the and each CILRT. Any instrument found, range of measurement for the CILHT is re)ected from consideration by eliminating all data collected from the sensor. Influence or volume weight factors are a@usted for the remaining sensors to conrpensate for
,the failure.
-l7-0 SXSTFA SOFTWARE As a minicomputer performs all calculations required to determine the primary containment leak rate, the computer software system represents a complex element of the leak rate measurement system. This section describes the purpose and features of the software required. to conduct the CILRT. Three basic tasks are performed by the software programs of the leak rate measure-ment system. First, before the CILRT, model definition, calibration data, and, channel r
repeatable error correction data must be stored in the minicomputer.
I Secondly, software programs acquire the test data and perform the leak rate calculations during the CILHT. Fina11y, raw and corrected data must be t
summarized after the test for plant records.
A. I'rior to the CILHT Several proGrams are used to define the model of the primary containment before the ClLHT is conducted. Based upon the number of temperature, vapor pressure, and. pressure sensors, the minicomputer aU.ocates storage space for the test data. In addition, the calibration data for each sensor must be stored prior to the test. Several programs are available to check various parts of the data entry process. The most significant is CHECK, which allows the computer to instantaneous+ compare the temperature of an installed RTD with a precision temperature standard.
Table II lists and summarizes all software required in the preparation for the CILRT.
H. During the ClLBT As the CILRT is conducted, the raw data must be stored, corrected according to the calibration factors, the results calculated.
0 and. The
primary program, FORE, receives data from the acquisition system, corrects according to the sensor calibrat1on factors, computes, stores, and displays the primary containment leak rate. Several other programs (HASE, TALLY, and LIST) provide the ability to change the sample considered the start of the test, provide statistical confidence intervals, and tabulate the test results.
Several unique features are included. to prevent the loss of data and enhance the information provided. to the test engineer. The most significant feature ensures that any time the data acquisition is prepared to transmit data, the minicomputer stops all activities so that the main data co33.ection program, FORE, may execute. When these data have been rece1ved and results printed, the minicomputer completes the task interrupted by the acquis1tion of data. All programs are designed to be tolerant of power failure. No previous data is lost when power is restored. Table III lists and summarizes all software programs required during the CILRT.
C. After the CILRT After the CILRT is completed, test data can be corrected. for any 1nstrument failure and arranged for inclusion in the permanent test record. Several software programs provide the ability to list all raw and corrected data, final test results, and, calibration constants. Table IV lists and summarizes the software programs used after the CILRT 1s complete.
0 Two DISCUSSION OF TEST RESULTS CILHT's have been conducted with the equipment and techniques described in this poper. Each type represents an extreme of conditions typically expected during the CIMT--smaU. volume with moderately high pressure and low pressure with moderate volume. Both tests were conducted for at least 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, with data collected at least every 15 minutes. This section presents a summary of the CILRT results. Complete reports have been filed with the NBC's Division of Operating Heactors.
A. Hrowns Ferry Nuclear Plant Unit 2, Conducted June 1978 Browns Ferry unit 2 is a boiling water reactor employing a steel pressure suppression Mark I containment. The maximum leak rate at a reduced pressure of 25 psig is limited by technical specification 4.7.a.2 to less than 0.04437 percentage per hour of containment air mass. The containment was modeled as two compartments the pressure suppression chamber and the drywell. Twenty-nine temperature sensors, six humidity sensors, and. two pressure gauges were used to measure the primary containment leak rate. ~
The free air primary containment volume is approximotely 300,000 cubic feet.
A 24-hour CILHT and a 12-hour verification test were conducted. June 13-16, 1978. The final measured leak, rate was 0.00949 percentage of containment air mass per hour. The observed 95-percent upper confidence limit for this measured. leak rate was 0.00994 percentage of containment air mass.
The mass leak rate calculated during this test is depicted. in figure 4.
Table 5 compares test duration with leak rate and upper confidence limit.
Clearly, the primary containment leak rate was accurate+ determined within the first 4 hours of the test. Figure 4 indicates that data collected. beyond the fourth hour of the test served on+ to improve the
upper confidence limN of the leak rate. Figure g depicts the upper confidence interval as a function of the time of data collection. The rapid approach to the asymptotic limit demonstrates the value of proper instrument selection. Complete summaries of the calcu1ated, test results are included in appendix B.
B. Sequoyah Nuclear Plant Unit 1, Conducted March 1979 Sequoyah unit 1 is a pressurized water reactor employing an ice condenser prcssure suppression primary containment. The maximum leakage of air at a test pressure of 12 psig is limited, by technical specification 4.6.1.2 to less than 0.0078 percentage per hour of containment air mass. The primary containment contains four compartments--the lower ice condenser compartment which houses the energy absorbing ice beds, the upper ice condenser compartment which encloses support equipment for the ice condenser system, the lower compartment which encloses the reactor and main piping systems, and the upper compartment which encloses the refueling work area. The free air mass was calculated separately for each compartment, with the calculated, leak rate derived from the sum of the compartment air masses. Based upon the instrument selection guide, 46 RTD's were used for containment atmosphere temperature measurement, 10 humidity sensors were used to monitor the containment atmosphere moisture content, and four quartz manometers monitored the total pressure. Total free air volume for the primary containment is approximately 1.19 million cubic feet.
A 24-hour CILRT and a 4-hour verification test were conducted March 13-16, 1979. The final measured leak rate was 0.00011 percentage of containment air mass. The observed 9g-percent upper confidence limit was 0.00024 percentage of the containment air mass. The mass leak rate calculated. is
depicted in figure 6. Table 5 compares test results with the duration of data collection. Clearly, the primary'containment leak rate was accurately determined within. the first 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> of data collection.
Figure 7 depicts the upper confidence interva1 as a function of the time of data collection. Complete summaries of the calculated test results are included in appendix C.
CONCLUSIONS CILBT's conducted by TVA on a high-pressure boiling water reactor containment and a 1'-pressure ice condenser pressurized water reactor containment verify that the leak rate measurement system used with the techniques outlined in this paper measured the primary containment leak rate in far less than the 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> the tests were conducted. An analysis of the 95-percent upper confidence limit of the measured leak rate indicates that the primary conlainment leak rate was accurate+ detexmined with a high level of confidence within the first 4 hours of data collection.
To consistently achieve this accuracy for future CILBT's, this paper has outlined several key techniques. The model used to calculate the primary containment leak rate must compensate for areas of varying temperature, pressure, and moisture content. The test instrumentation must be capable of extremely accurate and repeatable measurement of the containment atmosphere conditions.
Collected test data must be acquired quickly with reliable equipment. The test director must be provided with accurate results during the test.
TVA wiU. conduct future CILRT's in accordance with the techniques described, in this paper. Each CILRT will be conducted for at least 4 hours and extend until adequate confidence in the accuracy of the measured leak rate is-achieved.
FIGURES
ATMOSPHERIC PRESS.
JES $ !i'.E TA-'E hlP.
l I serial "antre! c~ JA:
TVA REMOTE Ai 'TODA~~ DIGITAL iocnar. Was 3IGIITA-COMPUTER COMMAND ~ gib ,'Ni">T lose M ItlTERcA"..E at MPXR '!i ~I < r FREDATA PLOTT E< THUM BYr".'.EE'SS'Y O
O 0 V C3 MENSOR VENSOR ME .83P IOIOO- 00 "IOIOO-GOI "10 -"0 00; VA JQ ~ characters each OUTSIDE CONT" INivlENT INS! DE CON i AiNIii'ENT TVA J.G
. 8 control UP TC 6I6 SCAiVNER ac RTD'S VIS I G IZ data IOO CHANNELS e IO TEN-CHANNEL control UNITS 6I6 SCANNER analog data IP TO power J. B.
DEYi(CELLS Ioo channels 2 sIgna SYSTEM LAYOUT TVA INTEGRATED LEAKAGE TEST SYSTEM DE'I'i'CEI L POWER SUPPLIES
LV~PATED ~ HATE HEASUBElEiiT SYS~i'1 DATA iKDUCTlON SYSTEM Dual Terminal ll/D4 Minicourputer Plexible Disk Printer Drives Data Acquisition Digital System Plotter
TENNESSEE VALLEY AUTHORITY COMPARISON OF LAB AND FIELD TEMP. MEASUREMENTS 4
CD is Cl bl V)
OCl ZZ QJK s
~CO QJD s ti O 4 K H l-.
Cl Z
pv is O
N Ql S 4 5 0 JÃ TU45 7 li Ii N STANDARD TEMP. DEG. F
0, TENNESSEE VALLEY AUTHORITY BROWNS FERRY NUCf EAR PLANT MASS LEAK RATE PLOT CZS PS16 TESTO g
8 ld CL Q)
M I -2
-6 w7 9
8 -$ 8 12 H lb TZMECHRS.O
TENNESSEE YALLEY AUTHORITY BROWS FERRY NUCLEAR PLANT UCL INTERVAL l:2S PSI8 TEST PRESSURD K
0 2S 1o K
<e Q p (fj $
14 Ia fg H
e 4
O o o , 1S lZ l4 10 lo & 2R Rf TINE PHRS)
TENNESSEE VALLEY AUTHORITY SEQUOYAH NUCLEAR PLANT HASS LEAK RATE PLOT C ) 2 PS18 TEST PRESSUREO 184 CY D
EE CL l
18 8 18 21 24 27 8 12 16
~
Tj:NECHRS.O
TENNESSEE VALLEY AUTHORITY SEQUOYAH NUCLEAR PLANT UCL INTERVAL (12 PSIG TEST PRESSURE) 4 N f8 10 O'I TINE (HRS)
TABLES
-$ 0-0 TABLE I DATA COLLECTED BY AUTOMATIC ACQUISITION SYSTEM
- 1. 13oiling Water Reactor, Pressure Suppression Containment
~uunntit Function 29 Resistance temperature detectors (RTD's) for containment atmospheric temperature measurement Lithium chloride dewcels for containment atmospheric vapor pressure measurement Precision quartz manometers for containment atmospheric total pressure measurement RTD's for containment vessel metal temperature and. test station temperature Mass f1owmeter for measurement of induced leak required for the verification test Precision quartz manometer for atmospheric pressure
\
Suppression chamber water level Reactor vessel water level
- 2. Pressurized Water Reactor, Ice Condenser Suppression Containment
~gunntiu Function HTD's for containment atmospheric temperature measurement 10 Lithium chloride dewcels for containment atmospheric vapor pressure measurement Precision quartz manometers for containment atmospheric total pressure measurement HTD's for containment vessel metal temperature and test station temperature measurement Mass flowmeter for measurement of induced leak required for the verification test Precision quartz manometer for atmospheric pressure
0
-3l-TABJZ II SYSTEM SOFTWARE REQUIIRD PRIOR TO THE CILRT Progrsm
~Name a Descri tion Sl Define the integrated leak rate system parameters: number of RTD's, dewcels, pressure gauges, analog inputs~ and local, RTD's. Create the xequired. system files required to store the test data.
C RM4 Define the sensor calibration data and volume weights. Requires EHTAM ENTYM calibration reports on all dewcels and RTD's that may be used for the CIST.
Measure the integrated leak rate system analog to digita1 repeatable offset. Requires all temporary cables to be installed and integrated leak rate system to be operational.
STARTH Define the calibxation data fox the quartz manometer pressure gauges and any plant process instrumentation, e.g., suppression chamber and reactor level transmitters.
CHECK Verify in-place system temperature or dewpoint measurements.
A standard for comparison is required for this program.
Print all stored calibration constants required to conduct the CILRT.
TABLE III SYSTEM SOFTWARE REQUlRED DURING THE CILRT Program
~Name s Descri tion Acquire containment data from the data acquisition system, store, correct raw data, and calculate leak rate.
LIST Print a summary of measured leak rate. Drive an online digital plotter to produce graphs of principal test results.
TALLY Calculate confidence limits of the calculated. leak rate.
BASE Redefine the sample considered. the start of the CILRT.
0 TABLE IV SYSTEM SOFTWARE REQUIRED AFTER THE CILHT Program
~Name s Descri tion Measure the integrated. leak rate system analog to digital repeatable offset after test is completed.
DUMDEV Print all raw and corrected test data.
AIIUSSS Print a compartment summary of the measured temperature, vapor pressure, pressure, and, air mass. Correct the test results for any sensor found out of calibration.
34 e TABLE V CILRT RESUJTS AS A FUNCTION OF TEST DURATION Brow>s Ferr Nuclear Plant Unit 2 colours PTP Leak+ UCL PTf~ Mass Leak UCL Mass CILRT Duration Number of Rate Leak Rate Rate Leak Rate Mass S les ~Per Hour ~Per Hour ~Per Hour ~Per Hour 33 0.00527 0.01693 0.00855 0.01036 49 0.00798 0.02318 0.00785 0.00893 24 97 0.00506 0.01921 0.00949 0.00994-Se uo Nuclear Plant Unit 1 PTP Leak+ UCL PTE+ Mass Leak UCL Mass CXLRT Duration Number of Hate Leak Rate Rate Leak Rate Mass Samples ~Per H'our ~Per Hour ~Per Hour ~Per Hour 6 25 0.00456 0.00470 0.00193 0.00238 34 0.00323 0.00336 0.00159 0.00188 10 42 0.00254 0.00265 0.00190 0.00211 51 0.00296 0.00307 0.00178 0.00193 0.00248 0.00258 0.00162 0.00168
+As defined in ANS-274 (draft)
4 I APPENDIX A
INSTRUMENT SELECTION GUIDE The containment air mass is calculated by the application of the ideal gas law:
144xVx(P-P)
RT i)y the mass point method the primary containment leak rate is the normalized slope of the mass loss curve:
M = At + 8, LR d x x 100 The tota1 differential of the calculated is mass is; dW-"144 V R
dP T
-~+~
dP T
dT
~Z Therefore, dP dPy + dT ~(P P
dt dt dt T The error in measurement of the independent variables, pressure, vapor pressure, and temperature determine the error in the leak rate. In general, an upper bound on the error in measurement of an independent variable X is:
hn upper bound on the error in a dependent variable Y, as determined by the measurement of a set of independent variables Xl, X2, . . . X is:
X Therefore, the upper bound of error of the measured leak rate can be expressed as:
The minimum change that may be reliably detected in the measurement of an independent variable is determined by the error of measurement.
In general, I/2 dX ' ( X~
2 (8) nx h
Substituting for each independent variable differential in equation (7) yields:
LR = -p dt +
~dt + ~dt x T x 100 (9)
In the paper, "Describing The Uncertainties In Single Sample Experi-ments," by McClintock et. al., it was shown the contribution of the measurement of each independent variable should be equal for an optimal instrumentation system. Therefore, equation (9) may be I
rewritten:
L< E dt 2+~x (P-P)
T 100 If a bound on the error in leak rate is assumed, the error in the measurement of an independent variable can be bounded.
TVA selects test instrumentation so that 25 percent of the maximum allowable leak rate can be measured within 8 hours. It is assumed that data is collected every 30 minutes. Therefore, equation (11) is rewritten:
E = ( 5)(.25 x L x .75) T
~
100 2T + (P Pq)
Substituting into equation (8) and solving for the number of instruments yields:
(13)
llLE>nJ c.>on ol ~>m>0>s e - Abs<>lu<,e <.rr<>r of th< measure of a vari<>ble A'bsol >I c < rror ol Lhr Indi< ation of thc measure of a variable I'.
'lI
- Rrl<>r lv< ~
< rr>)>: of a v>>rI<>blc I. - Abn<>lut r rror
< <>f leak mt.e, I>crccnt of <:o<>I,nin>>>cnt, ni.r mass per hour Number of r<~l>iicatious of a m< nsurement N - Number nf i>>dep< ndcnt m<.asurcmrnts I' At>solute pr<:.>>>r<,
R - Univcrsnl gas const.nnt S l)>.vlation from the mean of a population T Im<'f. sample
<<ss Stu<lent>s t
- l. - Lo"cr bound II - I!pprr bo>>nd
- 6. 159 0?.'?983 '3<.8578 57442.87 .0.0)9955'? 0.0473484 0,0)84184 T.PCS 82.9899 39i8503 57439.53 9. 4232Z94 I.II78766 I I485228
- 15. 15R 02.7573 3').'(! 96 5'/4P2.79 -S.S759592 1.9975599 I,897057P Ib.bee 02,?697 39,0261 57417.03 . I e48232 I, S.HI56958 I. 80/ill 9 16 '59 0).1/91 39.0165 5/370.92 I.2714910 I,RI96833 I. 997'9172 I/i.559 02.704I 39.0190 5/371.47 S. I I S I757 e.d89806H 8.9961108 I/..159 0?.?Obe 39.0I98 57304 '3 -P.5529!26 I.HH09634 I, ~ 902279 I?.era 0?.1'/45 3'r.P?dl .57306.59 - S. I I 15175 Q.9806624 P.I882954 11 '/'"() )I?./959 39.8173 57312.55 S.P9785i's P.R99'953P d.bd64?P9 I l .ird )I2.055? 'I9.0106 573'?2.13 .S,I365516 B,QH78634 P.4834615 I/./>0 02."ulh 39.)?167 57390.6/ .P,9454654 P.587!I32 S.PS63095
- 10. RIP 02.0'353 39.0?h? 57396.35 C.elbbbIZ 8.8972312 P. P I83311 10.."58 0?,879? 39 0)55 57382.52 I.S963606 4.8964564 P.II83738 I)I. SPH 02.90RI 39(.0? H") 57368.52 -S. S417698 I, HS17786 8.8863632 IO ~ 75R 02.92/i0 3'!.02/9 57379,11 5 '655357 e.dsBshlh P.9804457 19.950 'V,9'?I9 3>.0319 51393, "8 -H,S9942S2 I.BS71289 S.f f83439 I'(.258 03.SSZI 39.0?11 57)68.45 P.9920617 8,58624R6 , H.IPB3641 I ~. M4 03.115'? 39.0361 57357.46 H.1574594 9,elllseh S. SI05247 I'?. 759 03.1738 39.8CI3 '7359.34
- l. 006 53931.'l 91146.9 79'?1.4 16947.7 S.es2.4s 8.88326 H,HPZZS 1.2$ 0 ".3933.3 91140,9 1897.6 16947.1 -0,0$ 720 S.88764 9.88223 I ~ $ 09 $ 3934.1 91149.7 7181.3 16946.6 ~
- 5. V'70 5'3931.4 9II53.6 7692.6 16944.3 H.PS9PI s.se240 e.celeb
- 21. I'36 53143 ~ Ir 91101.2 7971.9 16370.4 -$ ,80033 $ .08179 9.08171 21.!,06 5 94?.4 91 1 06.3 7S18.7 15077.3 S.00299 e,eslse H.QPI7e 2'..936 53944,7 9I IS 1.9 7879.8 16076.7 -S.es?SS B.SB169 B.HHI69 2?.1"5 53945.1 '11104.1 7970.0 16076 ' -$ .08399 e.ee163 Q.eelbs ZZ. I36 53917.3 91187.3 7078.6 16076.8 -S.BIBZS 9.8$ 15H B.HPI66 22.60! 539IS.S 91100.9 7S?S.S 1601 S.6 -8,086H3 S.HS141 Q.P$ 163 2c..9>6 53949 0~ 91191.7 1068.0 16075.1 8.81568 H.$ 6156 s,eel62 23.1<<<<6 53949.6 91191.5 7962.2 16073.5 S.92$ 26 9.$ 9176 B.SB162 23.436 53947,3 91109.3 7667.1 16071.9 S.SB281 H.HSI77 S.SSI62 Z3.606 53945.6 91107.3 707B.S 16079.3 S.SB356 $ .$ 0179 S.S6162 23.936 53946.8 91103,I 7874,0 1606'1.3 S.08181 B,PBI79 S.S$ 162 21.106 53946.9 91105.4 7974.0 16060.3 -e.seesl e.e9177 S.SS162 Z4.436 53946.6 91101.3 7913.8 16867.5 -e,s8392 e.sp171 S.BH162 24.6M 53946.9 9110'/.9 7974,0 16066.6 e.see?2 e;B&169 s.eslbl