ML072710606
| ML072710606 | |
| Person / Time | |
|---|---|
| Site: | Calvert Cliffs, Seabrook, Fort Calhoun  |
| Issue date: | 06/05/2006 |
| From: | Lyon W, John Nakoski NRC/NRO/DCIP/CQVB, NRC/NRR/ADES/DSS |
| To: | |
| References | |
| FOIA/PA-2007-0255 | |
| Download: ML072710606 (18) | |
Text
/r \-% PefA(0 USE OF ULTRASONIC FLOW MEASUREMENT TO DETERMINE REACTOR POWER
- PURPOSE OF THIS PRESENTATION
- To inform the ET and LT of the status of the staffs review of the application of ultrasonic flow meters (UFMs) in determining reactor power
- MEASURE FOR SUCCESS OF THIS PRESENTATION
+ ET and LT are aware of staff activities
+ ET and LT are advised of the potential need for regulatory action to address deficiencies in the application of UFMs PRESENTERS: Warren Lyon Presentation Date: June 8, 2006 John Nakoski
"-, 1
OVERVIEW
/" Caldon Check and CheckPlus UFMs are acceptable
/ Westinghouse/Advanced Measurement and Analysis Group (W/AMAG) review is ongoing. Anticipated staff finding is unacceptable
/" External UFMs for power recovery under 50.59 have not been shown to be acceptable 2
PLANT-SPECIFIC STATUS - RECENT LARs
/ Seabrook (Caldon) UFM is acceptable (ML061360034, May 22, 2006)
/ Calvert Cliffs (W/AMAG) is on hold - An amended LAR is anticipated v/ Ft. Calhoun (W/AMAG) is on hold pending generic review completion 3
ACCEPTANCE CRITERIA
- Traceability - Relating a measurement to a standard Standard maintained by a national laboratory - National Institute of Standards and Technology (NIST)
Each step between measurement and standard - Clearly defined and no unverified assumptions Unbroken path between measurement and standard Total measurement uncertainty reflects aggregate uncertainties of each step
- Applicability - Provide accurate information over range of use
- Sound basis (Theoretical understanding or equivalent)
Calibration is constant or change is fully understood, predictable, and verifiable 4
ACCEPTANCE CRITERIA ASSESSMENT
/ Caldon Check and CheckPlus meet all criteria
/I W/AMAG Crossflow has not been shown to meet any criteria but we have not completed review of all submitted information 5
CALDON CHECK AND CHECKPLUS ASSESSMENT 0 Theoretical and analytical descriptions cover all test, installation, and operational conditions
- Instrument provides complete assessment of hydraulic conditions and operation within uncertainty bounds 0 Every instrument is tested at Alden Research Laboratory in a full-scale mockup of the installation configuration over a wide range of flow rates
- Alden Research Laboratory results are controlled by NIST-certified standards. Test flow rate uncertainty is typically -- 0.1%.
- Seabrook tests included deliberate introduction of distorted flow profile far beyond what would be anticipated in plant
- 175 tests with 2000 to 6000 data points per test were run 0 Worst uncalibrated CheckPlus flow rate error was < 0.5%
6
CALDON CHECK AND CHECKPLUS ASSESSMENT (cont)
- Check and CheckPlus provide the velocity profile
- Alden flow rates were less than Seabrook's and tests were conducted at room temperature
- Consequently, Reynolds Number was about a factor of five less than for plant operation
- Check and CheckPlus address this by using instrument output that is consistent with theoretical considerations
- A Reynolds Number extrapolation is not required
- Similar tests have been conducted for 35 CheckPlus and 9 Check UFM flow elements 7
CALDON CHECK AND CHECKPLUS ASSESSMENT (CONCLUSION)
/ There is a substantial theoretical, analytical, and empirical basis for understanding how the device works
,/ There is substantial data that supports operation
/ Meter self-assessment has been demonstrated
/There is an unbroken path from calibration used during meter operation to reference standards
/" Check and Checkplus UFMs provide the claimed precision 8
W/AMAG CROSSFLOW ASSESSMENT Assessment is continuing. The following are based on material reviewed to date and are preliminary and subject to change.
Theoretical and analytical descriptions do not cover test, installation, and operational conditions Crossflow appears to be incapable of providing a complete assessment of hydraulic conditions Crossflow is tested at a full scale facility only if the application appears to cause a flow profile that differs from previous installations
- Generic Alden Research Laboratory uncertainty of 0.25% is used. Iftest data agree to within 0.25%, then the claim is made that one cannot do better and no further test data are obtained. This is an incorrect, unacceptable use of statistics.
9
W/AMAG CROSSFLOW ASSESSMENT (cont)
Initial full scale tests were limited to what was believed to be fully developed flow conditions
- 8 tests with 30 to 40 data points per test were run 4 tests were later rejected essentially as outliers. Still later, W/AMAG found the reason for poor data was noise contamination.
- Uncalibrated Crossflow flow rate error was - 7%
Calibration factor obtained from these 4 tests is used as the calibration for what W/AMAG claim is "stable flow" 10
W/AMAG CROSSFLOW ASSESSMENT (cont)
Am additional correction factor is obtained from tests for single and multiple elbows, and for a few plant-specific configurations
- There are sometimes insufficient data to acceptably substantiate the claimed representation. Further, W/AMAG use the 0.25% uncertainty argument to claim converged conditions have been achieved.
- Further tests have been run at facilities other than Alden
- These are claimed to substantiate crossflow calibration. These claims are not always valid.
11
WIAMAG CROSSFLOW ASSESSMENT (cont)
Crossflow provides a representative velocity along a path that is perpendicular to the pipe axis
- Velocity location along the path is unknown
- Indicated velocity may be a function of velocity profile Correlation of indicated velocity to flow rate, and hence calibration, are unknown functions of flow profile
- Alden test data are at room temperature
- Reynolds Number is about a factor of five less than for plant operation 12
WIAMAG CROSSFLOW ASSESSMENT (cont)
W/AMAG extrapolates Reynolds Number to obtain plant operation characteristics
- This appears to break the calibration reference to a standard
- Our assessment of the extrapolation validity is incomplete W/AMAG assumes Alden test results and calibration may be used without an uncertainty or bias impact when crossflow is installed consistent with practice with venturis 13
WIAMAG CROSSFLOW ASSESSMENT (cont)
This is unacceptable:
Swirl causes a venturi to over-predict flow rate but causes Crossflow to under-predict flow rate
- Venturis are less sensitive to swirl than Crossflow
- .There is no proof the installations are identical
- Calibration is no longer referenced to a standard 14
W/AMAG CROSSFLOW ASSESSMENT (cont)
Calvert Cliffs has attempted to bypass the above problem by using chemical tracer tests for full power calibration in the plant Ft. Calhoun appears to be using the same approach with recently calibrated venturis
- These appear to be acceptable but uncertainty has not been determined But the calibration applies only at the time of calibration. Changes in plant condition may later invalidate the calibration.
15
W/AMAG CROSSFLOW ASSESSMENT (cont)
W/AMAG apply multiple procedures and tests to circumvent Crossflow sensitivity and inability to independently identify flow profile changes
- Power levei and feedwater configuration are constrained to fixed limits Data are compiled and processed in many ways to determine acceptability Comparisons are sometimes made with all other available plant parameters that provide insight
- Staff is continuing to evaluate these activities 16
W/AMAG CROSSFLOW ASSESSMENT (cont)
Potentially effected measurement uncertainty power uprates (MURs):
Plant % Uprate MWt Plant % Uprate MWt Increase Increase Salem 1&2 1.4 48 x 2 = 96 Pilgrim 1.5 30 Hope Creek 1.4 46 Hatch 1&2 1.4 48 x 2 = 96 SONGS2&3 1.42 48 x 2 = 96 Kewaunee 1.4 23 STP 1&2 1.4 53 x 2 = 106 Palisades 1.4 35.4
- Total MWt increase from Crossflow based MURs = 528 MWt.
- This equates to about 185 MWe.
17
WlAMAG CROSSFLOW ASSESSMENT (CONCLUSION)
/I Theoretical, analytical, and understanding basis is weak
/ Empirical data basis is weak
/ Meter self-assessment does not appear viable
/ An unbroken path from calibration used during meter operation to reference standards has not been achieved
/ The NRC-approved topical report is no longer valid
/ NRC review is nearing completion although much work remains
/ Regulatory action may be required to address the use of crossflow and other external UFMs for MURs and power recovery applications 18