ML072710574
| ML072710574 | |
| Person / Time | |
|---|---|
| Site: | Seabrook  |
| Issue date: | 02/06/2006 |
| From: | - No Known Affiliation |
| To: | Office of Nuclear Reactor Regulation |
| References | |
| FOIA/PA-2007-0255 | |
| Download: ML072710574 (9) | |
Text
Attachment Meeting at Alden Research Laboratory, Inc. On January 17 and 18, 2006 to Review Application of the Caldon Checkplus Ultrasonic Flow Meter (UFM) for a Power Uprate at the Seabrook Nuclear Power Plant 1
SUMMARY
AND CONCLUSIONS The subject meeting was held to obtain additional information as discussed during the Reference A and B meeting held at NRC's Headquarters on December 16, 2005. Attendance is identified in Attachments I and 2.Jim Nystrom, the Senior Vice President of Alden Labs, provided an analysis of the facility that supported an uncertainty of 0.088 percent in determination of the volume of water collected during each test run (Reference C). I believe the result is consistent with the National Institute of Standards and Technology (NIST) guidance that is applied to operation of the Alden Facility and I conclude this uncertainty is appropriate for use in evaluation of the Caldon Checkplus UFM.A total of 175 tests were conducted with about 2000 to 6000 data points per test, depending upon flow rate. The tests covered a wide range of configurations and flow rates designed to bracket conditions that might hle encountered in the Seabrook application.
This included flow rates significantly below what ý,ould be encountered at Seabrook, but the maximum flow rate did not achieve a value equal to Sdabrook operation at full power because this was outside the range capability of Alden Laboratory.
Interestingly, there was a test series where swirl was deliberately introduced.
The poorest correlation in all tests required a correction factor of 1.0048 to obtain agreement between the Checkplus flowrate prediction and the flowrate determined during the Alden tests. This close agreement between indicated and actual flow rate places less demand upon Checkplus correction than would be the case if the Checkplus was less accurate.Caldon uses the Checkplus output to compensate for flowrate changes and swirl and no longer uses a Reynolds number extrapolation to extrapolate UFM correction factor to plant conditions.
This approach appears promising but I have not completed my evaluation of these processes and have not reached any findings.
Consequently, there is little coverage of these topics in this report.I have identified no significant problems with use of Checkplus at Seabrook.
We tentatively scheduled a final meeting at on March 21 and 22, 2006, should an additional meeting be necessary.
Our schedule is to complete the safety evaluation report by May 31, 2006. This is contingent upon receiving the Alden test report with associated analyses, and any additional information that is found to beineeded during our ongoing review, by March 31, 2006. If such material is received earlier than anticipated, then we will attempt to complete our review earlier.2 THE ALDEN RESEARCH LABORATORY TEST FACILITY Figure 1 is a generic sketch of the test facility used for the Seabrook UFM testing. Flow in the test sections starts with a pair of pumps in the lower left, passes through the test sections, through the breakdown (throttle) valves, through the switchway (diverter), and either into the weighing.:: !. :..: ...: .... : " * .. .. : : : .. i ....... " : : ..... ... ..... .. ....... ....: "..: ::...: :~ ii .i .' .: .K I ; ' ! k i .k : ..; ...r~ I.. ....~ ..... .....Figure 1. :Alden Test Facility,:
- : ... : .tank or into the sump. The outlet from the breakdown valves is at atmospheric pressure and activities in the switchway have no influence on flow rate. The switchway consists of a manifold where water drops vertically onto a knife edge diverter plate that sends flow to the weighing tank or to the sump. During steady state operation, the knife edge is out of the flow stream. During switching, the knife edge accelerates to a constant rate prior to entering the flow stream and decelerates after leaving the flo 'w stream as illustrated in Figure 2. This restricts diverter interaction with the flowstreamn during diverter movement to times when diverter movement is constant.
Hence, by recording time in the center of the diverter movement for both initiation and termination of weighing tank fill, in the ideal sense, error due to switching is eliminated as is illustrated in Figure 2. In practice, the standard deviation due to the diverter is 0.0100 percent, a small contribution to the overall standard deviation of 0.044 percent (uncertainty of 0.088 percent or two standard deviations).
The other contributors to uncertainty are routine such as due to mass, time, water density, buoyancy due to air, and traceability to standards, and are not addressed further here.Flow from the pumps into the manifold'and from the manifold into the two test sections occurs over a short path with abrupt turns. This introduces the possibility of noise, poorly developed flow profiles, and swirl in the test sections; all items of concern with UFMs but that are of lesser concern with testing other devices.
Figufe 2. Dkerter Operation' : C .: : " .:. "s *. ....:- -... ..............> ...... ..U1I:::::::::-
- .: ........ ...... .. -.-.'"...." .........
..........
..."-:* ...... ..*"'..... .....' : : Fiur 3. SebokTetCnigrto
- ..".....: .: : : : : .: : : : ..: ::': ..: : ..: .. .: .: ...: , .. ..:: ....:: ..:: : : .: : ... .: ... ....r:::s:: ::t: : :: : : :: : : :: : ::;s: :.r.y. : : .: ... ....: ..1 ..: : ...1: : : : : : : : : .: ..: ...* ... .: : : : : : .: : : : ...: .: : : ..: : : : ...: ...: .......: : .: : : : : : : : ..:* .: : .: : : ' ..: : ..: : .: ..: * * : ....:. ..: : : : .:. : ..: :. .: : .:: : .: :: : : .* : ..: : .: : : ..... : -:.: : : :...C ................... ......... ..............................3 .. .::... :::- TH E: .SEA BR O O K i.:::." ::':::.. :: .!:..!!! :X:: :: ::.: !::The
- Alente t.onig ra io for... th Seabrook..
- test : isi : shown in.:, Fig.:.. re: 3.: Flow:: ::.. is:.. from--:.::.
right: to.: .:.:: left.. with ... two .24 inch.... pipe connecting4 to ..:.: ::::::::::::::
- a:36.inch.pipe that.contains the.hec lu :JF.::.:::..
- ..Figu..re .Se .r.o Test :. /:- Configuration:.
u::. ..;::::::. :
- ....' .b ! :: : .: i i il ' : " ': " .: : :. .....: " .b ... ' .: .: ! .. :! .!. ..: ..: : : .' : :F"'1. : ..::' .* .:7 ' i .
-" " ' " :-f..J :.t F , u d .:: : :'ii" ::s:I:.i! : i ::::iii: .: ::........ ... .4 ......... .. .. .. ......: ..... .....:: : : : : .:: ...: : .: .." , .......: : : i :: i ! i i :i i i : " .: .u : :: .: : ! : ::: :: :::: : :: : ::V .: i .: .: : .* U ..:.:i : Y ..I / : : : : ..::. : : : : :: :: :: ::: ..i : ii!:ii..ii. : :: :£ :.* .., " 4: 3 :... ..:,. :4 -/ : :. :,¢ ...,.,..,. ,.. ,.:-.......
..!:!...: .:::..\ .........` ` :`f ...:!i !..!:
The preliminary test results are summarized in the following table, with five individual tests conducted for each test entry, and each individual test consisting of at least 2000 flow rate measurements':
Test Description Flow Split Approx Flow Min / Max_I I IRate, gpm Correction Factor A-1 Straight 36 inch pipe run N/A 2010 1.0024/1.0041 of about 49 feet Line 1 6470 1.0017/1.0041 omitted. 11020 1.0029/1.0042 15440 1.0031/1.0048 19980 1.0029/1.0040 B-i Plant Model 50/50 1990 1.0003/1.0029 6510 1.0021/1.0034 11080 1.0023/1.0037 15520 1.0027/1.0037 19610 1.0029/1.0037 B-2 Flow Split A 25/75 2006 1.0029/1.0035 6520 1.0021/1.0033 10985 1.0029/1.0045 15550 1.0029/1.0040 19960 1.0028/1.0037 B-3 Flow 'Split B 75/25 1949 1.0003/1.0014 6509 1.0018/1.0034 11065 1.0023/1.0034 15530 1.0021/1.0037 18880 1.0024/1.0039 B-4 Remove Mitsubishi 50/50 1990 1.0023/1.0027 Flow Conditioner 6510 1.0030/1.0033 10965 1.0029/1.0036 15530 1.0033/1.0044 19770 1.0024/1.0035
'More than 2000 measurements are obtained at the highest flow rate. The number of measurements is a direct function of the time it takes to fill the weigh tank, a time that increases as the flow rate decreases.
Thus, -he results of each test run are based on a good statistical sample.
Test Description Flow Split Approx Flow Min / Max I I Rate, gpm I Correction Factor Place half-moon plate 50/50 2030 1.0005/1.0025 covering bottom half of 6540 1.0017/1.0041 pipe in Flow ' 11040 1.0018/1.0044 Conditioner location 15360 1.0017/1.0029 19300 1.0022/1.0036 Full flow 11070 1.0004/1.0013 through half- 15560 1.0011/1.0025 moon 17080 0.9992/1.0014 Replace two See A-I N/A 15580 1.0032 / 1.0046 Transducers 20050 1.0026 / 1.0037 A cursory review of the configurations, data, and supporting information lead to the following preliminary conclusions:
- 1. Seabrook's specifications for the Checkplus are 16,400,000 lbs/hr @ 447 'F (Reference D). This corresponds to close to 40,000 gpm or 20,000 gpm per feedwater train. With the exception of Test B-4 with full flow through the half-moon, the B series tests are with flow rates of roughly half what Seabrook specified for total flowrate capability.
This is an open item that we are reviewing.
- 2. Water temperature during the tests was approximately 95 'F. The operational specification is 447 'F. This is an open item that we are reviewing.
- 3. Test A-I covers the extreme condition of feedwater limited to one feedwater path. The full flow test through the half-moon covers the extreme condition of feedwater from the other feedwater path with the additional complication of introduction of swirl. These bracket the operational configurations that will be encountered during plant operation.
- 4. The tests provide Checkplus characteristics over a wide range of flow rates.5. The poorest correlation in all tests requires a correction factor of 1.0048. For this case, the uncorrected Checkplus error is 0.48 percent if the test uncertainty of 0.088 percent is neglected.
This close agreement between indicated and actual flow rate places less demand upon Checkplus correction than would be the case if the Checkplus was less accurate.6. The Checkplus is relatively insensitive to change in flow rate over a wide range.7. The Checkplus is relatively insensitive to swirl. (Some swirl is introduced at the entrance to the test sections due to the test facility.
Quantitative information is expected.)
- 8. The Checkplus is relatively insensitive to changes in feedwater configuration.
- 9. There is essentially no change in Checkplus indication when two transducers are replaced.(There were only two transducers on hand during the tests. More data are to be provided that addresses transducer changes.)The Caldon UFM provides either four or eight average velocities as illustrated in Figure 4 for the .four velocity design. Caldon then defines a "Flatness
.Ratio" (FR) as the sum of the Center velocities (V 2 , and V 3) divided by the sum of the outer velocities 1 (V, and V 4) for the four path UFM, with a similar i definition for the Checkplus eight path UFM. It uses this parameter to assess UFM behavior and to address characteristics under different flow i .conditions.
With this approach, for example, .z-Reynolds number extrapolations are no longer used.As a further example, preliminary test results are provided in Figure 5 that show correction factor (that Caldon defines as MF) as a function of FR for a range of conditions.
Further, Caldon has also used Checkplus results to determine the influence of swirl on correction factor.* ...: ..: ~ ... _ .. ...... .: ...*. ....... ............ .....: :. .: ........._ ..: .. ..- ..: ... ..... .. .. ... .: .: ...: -.: : .: :. .-..- ..: :. ::. : :.::.: :.1'e 0. 03 09 0 1 1 i i i : i : : : : : i / : i :: : i : / / / ~ ? : i L : :: : : : :: : : : : : ...: : : :: : :: : : O: :: : : :: ::: : :: ?i L : :: :: : ::: ::: :: :: 1 oO:i , : : :: i : : : :::: : : : i~ ! : : : : : : : : : : : : : i :: :: : : : : : : : : : : : : :. ....: : : : : : : ..0 0 DB:. : .ii : : i : i i i i .: : ....:: : i: :: : i : : :::: !: :: ? ? :i:: i :: : :: : : :: : : : :: : : : ::: :i .:~ l~ i i : i: .: : :: ::i i i i :: .:: :i: :.: :: ? : i: ! :: ! i 4 .:/ i, ! .:. .! : i .! .: ' i ' : : I / ' : :: " : i .: .: : : : : : : : : : : : ! ! i i : i i : .: : i i i .~ : i : ..i : -.: : : i : : i .: ' i .: : : : .: i :~ i : i i : : i i i : i~ i i : i : .i : ::::: : o: ... ........8 088: 0.BI fl: O.:::::::1 0.93 0 :: : !!:~iii~ii~~~.1i~
- . : :: : .: : .:. : : : .? !i : : ! !. : !! : :i ! :! : " ..: ! : :: .: i :: ' ! !: ! !: ! !! ! !! ? !! ! !! ! !! ! !! ! .! / L .!.My tit3 A! i~ i ?! ! !? .? !
REFERENCES A "Summary.of December 16, 2005, Meeting with FPL Energy Seabrook, LLC to Discuss Measurement Uncertainty Recapture Power Uprate Licnse Amendment Request (TAC No. MC8434," NPRC, ADAMS Number ML060030107 (Slides used at the meeting are ML060130031 and Meeting Attendance is ML060130033), January 11, 2006.B Hauser, Ernie, "Trip Report #1770, NRC / Seabrook Meeting of December 16, 2005," Caldon, January 5, 2005.C "Uncertainty Analysis of Flow Measurement, 100,000 Lb Weigh Tank," Presentation At Alden Research Laboratory, Inc., January 18, 2006.D "Caldon Proprietary Information Package for Seabrook/NRC Meeting, December 16, 2005," Tab 10, FP&L Purchase Specification S-X-I-E-0139 Rev. 00, "Ultrasonic Feedwater Flow Metering System."
ATTACHMENT
- 1. ATTENDANCE SHEET ON JANUARY 17, 2006[ Name F Organization Position Gregg Sessler FPLE FW System Eng Kerry Walton Southern Nuclear UFM Proj Eng, Vogtle Upgrade Kenji Tominaga Hitachi / Canada C&IE Eng Warren Lyon USNRC Sr React Eng Bob Dean FPLE Cognizant Design Engineer Ian Watters FPLE Project Engineer, Power Uprate Don Nowicki FPLE Oversight, Sr Engineer Mike McMahon CEG-Corporate G-S Corp Eng Engineering Mike O'Keefe FPLE Reg Comp Supv Howard Onorano FPLE Power Uprate Engineer Herb Estrada Caldon Chief Engr Don Augenstew Caldon Mgr Engr Ernie Hauser Caldon VP Nuclear Also present, but did not sign attendance list James Nystrom Alden Sr Vice President ATTACHMENT