ML080280400

Thermal Power Measurement

ML080280400
Person / Time
Site:	Byron
Issue date:	09/18/2003
From:	Exelon Generation Co, Exelon Nuclear
To:	Office of Nuclear Reactor Regulation
References
FOIA/PA-2008-0046
Download: ML080280400 (30)

Text

ExelknM Nuclear Byron Station Thermal Power. Measurement September 18, 2003 Enclosure 2

ExeI~n-.

Nuclear Agenda and Opening Remarks Jim Meister Vice President - Engineering Enclosure 2

Agenda Exe[ký"n,.

Nuclear

• Opening remarks

" Background on ultrasonic feedwater (FW) flow measurement

• Chronology of ultrasonic flow meter (UFM) implementation o UFM test plan and results

" Root cause and corrective actions

" Safety implications

" Future plans/actions

" Closing remarks 3

Enclosure 2

Opening Remarks ExeL n, Nuclear

• Several in-depth investigations performed, from 1999 to 2002, to determine reason for megawatt (MW) differences between Byron and Braidwood

• Test plan identified problem with UFM measurement of FW flow in August 2003

• Preliminary root cause determined to be UFM correction factor error caused by pressure pulses in FW piping which were caused by resonance

• Prior to 2003 test plan, multiple rigorous reviews conducted to evaluate all causes considered plausible

• Amount of reactor overpower initially determined acceptable with respect to safety analyses criteria

• Broad review of UFM decision making initiated 4

Enclosure 2

Exelkn,.

Nuclear Background on Ultrasonic FW Flow Measurement Bill Kouba Exelon Engineering Director Enclosure 2

Background on Ultrasonic FW Flow Measurement Exelnt .

Nuclear

" Purpose of UFMs

- More accurately measure FW flow

- Potentially recover MW lost due to FW venturi flow inaccuracies

" Installation was not part of a measurement uncertainty recapture (i.e., Appendix K) uprate

- No intent to raise power beyond existing licensed limit

- Five percent power uprate moved plants from Appendix K to Best Estimate loss of coolant accident (LOCA) basis 6

Enclosure 2

Background on Ultrasonic FW Flow Measurement (cont.) Exekrnu.

Nuclear

• UFM captures the signature at A and B and calculates the travel time L - 1 A Pt MR C/)

IA B Time 0

-)

2 At 7

Enclosure 2

Background on Ultrasonic FW Flow Measurement (cont.) Exek n Nuclear How the UFM Works

• Signal sent through each of the two sets of transducers is modified by eddies o Eddy modification "fingerprints" signal o Time delay for "fingerprint" movement is determined.by a statistical technique called'cross-correlation

• Plant UFM "correction factor" is calculated by dividing the UFM mass flow by that from the venturi 8

Enclosure 2

Background on Ultrasonic FW Flow Measurement (cont.) Exelkn.

Nuclear

" Installation verified by UFM vendor

- In accordance with vendor procedures

- NRC subsequently approved UFM technology in March 2000 for use in Appendix K uprates

- Vendor procedures consistent with NRC-approved topical report

• UFM installed on each FW branch line supplying steam generators

• UFMs installed in the same manner on the four Byron/Braidwood units

• Correction factors used in calorimetric calculation to correct FW flow

• Correction factors determined periodically, after a defined change in power (potential de-fouling event), or plant parameter trending 9

Enclosure 2

Exekýn,.

Nuclear Chronology of UFM Implementation Bill Kouba Enclosure 2

.Chronology of UFM Implementation Exelkn,.

Nuclear

• UFMs implemented at Braidwood - June 1999 o UFMs implemented at Byron- May 2000

• Electrical output differences identified between Braidwood and Byron

-. Upon initial installation

- Following five percent power uprate in 2001 o Multiple evaluations conducted from 1999 through 2002 to determine reason for differences in electrical output 11 Enclosure 2

Chronology of UFM Implementation Summary of Evaluations Nuclear Evaluations performed June 1999 to May 2000

- Dual instrument test with ultrasonic flow instruments, and UFM vendor review of Byron installation Additional validation testing at-Braidwood to verify data acquisition based on venturi cleaning methodologies Internal Exelon design review Secondary plant parameters, fuel utilization, heat rates, implementing procedures

• Evaluations concluded Byron UFM implementation was installed and operating within criteria established for UFM technology 12 Enclosure 2

Chronology of UFM Imp lementation Exelnm Summary of Evaluations (cont.)

Nuclear

• Independent Exelon review performed in February 2002

- Identifies fuel burn-up anomalies

- Recommends additional detailed evaluation

• Byron removes UFM venturi correction factors pending evaluation of fuel burn-up concern

• Exelon Nuclear Fuels organization determines fuel burn-up is within predicted range

- Byron reinstates UFM correction factors after determination 13 Enclosure 2

Chronology of UFM Implementation E Summary of Evaluations (cont.) ,xe,6n.

Nuclear UFM vendor, Corporate Engineering, and site review UFM implementation - March 2002

- Installation and operational criteria verified including piping, transducers, cables, software, and test procedures Comparison testing conducted between common FW header and individual FW lines Study concludes UFM measured flow per design and implemented properly Continuous data subsequently recorded in response to a recommendation from this study 14 Enclosure 2

Chronology of UFM Implementation Exeln,.

Nuclear

• NRC resident inspector issues unresolved item in inspection report

• NRC issues letter to Exelon concerning Byron Unit 1 thermal power level

• Exelon response concludes Byron Unit 1 UFMs installed consistent with NRC guidance, and that Unit 1 is operating within its licensed thermal power limit

• Exelon Engineering test plan initiated - March 2003 15 Enclosure 2

Exelkn.

Nuclear UFM Test Plan and Results Bill Kouba Enclosure 2

UFM Test Plan and Results Exelknt .

Nuclear.

• Purpose of test plan - identify reasons for electrical output differences between Byron/Braidwood stations

• Investigate long-term trends of specific plant parameters

- Monitor correction factor trends on Byron Unit 1 with continuous data link to UFM vendor

- Observe performance during steady state, power changes, pre/post-refueling outage

• Use additional UFM on common FW header upstream of UFMs on individual FW lines

- Check venturi flow sum and existing UFM flow sum

- Determine if difference between common header UFM and sum of individual FW line UFMs at Byron Unit 1 is within statistical allowance 17 Enclosure 2

UFM Test Plan and Results Results of Testing Exeltn Nuclear

• Braidwood Unit 1 comparison showed very close correlation between common header and individual line UFMs

• Comparison of common FW header UFM to sum of individual FW line UFMs on Byron Unit 1 was not within statistical allowance

" Signal noise observed on some individual FW line UFMs

" Common FW header UFMs had no signal noise

• Definite problem identified, decisions made to reduce power 18 Enclosure 2

UFM Test Plan and Results Power Reductions ixel n,m Nuclear

• Byron Unit 1 reduced power 32-MWe

- Based on correction factor differences between common FW header UFM and individual FW line UFMs

- Correction factors reset to 1

• .Signal noise anomalies, in conjunction with cormmon-header to individual line comparisons, were used to determine extent of condition for other units

• Byron Unit 2 reduced power 22 MWe

- Based on noise observed in one of four individual FW line UFM signals

- Correction factors conservatively reset to 1 19 Enclosure 2

UFM Test Plan and Results Power Reductions (cont.) Exelnt Nuclear

• Braidwood Unit 1 not affected because of very close correlation between common header UFM and sum of individual FW line UFMs

" Braidwood Unit 2 reduced power 11 MWe

- Based on noise observed in two of four individual FW line UFM signals

" ENS notifications made in accordance with license condition 20 Enclosure 2

Exeln,.

Nuclear Root Cause and Corrective Actions Brad Adams Site Engineering Director Enclosure 2

Root Cause and Corrective Actions e-M Overview n Nuclear

• Formed root cause analysis team to determine root cause-of UFM inaccurate FW flow measurements

• Preliminary root cause

- UFM correction factor error

-Correction factor error caused by noise impact on time delay Bias on calculated time delay (flow) varies at different power levels

- Resulting bias varies as a function of noise structure and intensity 22 Enclosure 2

Root Cause and Corrective Actions Ex n,.

Overview Nuclear Preliminary root cause (cont.)

- Presence of noise in individual loops' flow signal caused a non-linearity in calculatedventuri correction factor as a function of power level

- Noise caused by pressure pulses in FW piping

- Pressure pulses in FW piping caused by resonance

- Resonance in FW piping caused by a driver at the natural acoustic frequency of the piping 23 Enclosure 2

I I Root Cause and Corrective Actions Exektn.

Corrective Actions Nuclear

• Removed correction factors to eliminate error e Installation of common FW header UFM to provide for an alternate calculation of correction factor

• Revising appropriate site procedures to check UFM for noise

- Acceptance criteria established for correction factor determination

• Independent technical review

• Evaluation performed on portion of decision making regarding use of U FMs, team has been chartered to comprehensively evaluate decision making on a broader level and over the life span of this issue 24 Enclosure 2

., I " I Exelkn,.

Nuclear Safety Implications Brad Adams Enclosure 2

, .1 .1 1 Safety Implications ExekIna, Nuclear

" Byron Unit 1 overpower limited to 101.6%, Byron Unit 2 overpower limited to 100.4%

" Braidwood Unit 2 overpower limited to 100.3%

" Evaluations of LOCA, non-LOCA, containment, and dose analyses being performed

" Evaluations being performed with conservative assumptions that envelope historical power levels

" Preliminary results indicate applicable safety analyses acceptance criteria were met 26 Enclosure 2

.11 Exelk nsm Nuclear Future Plans/Actions Brad Adams Enclosure 2

t ( 11 0 Future Plans/Actions Exel n.

Nuclear

" Project plan in progress to install common FW header UFMs

- Testing has determined common header UFMs are free of noise

• Scheduled completion for Byron/Braidwood units is September/October 2003

• Independent technical review

• Broad review of decision making initiated

- Results will be shared with NRC 28 Enclosure 2

I

-- 4 Exelkn,.

Nuclear Closing Remarks Jim Meister Enclosure 2

. 3 t ,

Closing Remarks Exelnt Nuclear

" Root cause is UFM correction factor error caused by pressure pulses in FW piping

• Low safety significance of overpower condition

• UFM implementation at Byron Units 1 and 2, and Braidwood Unit 2 pending formal evaluation of root cause and corrective actions

• Actions were taken historically to investigate power level anomalies

• Previous evaluations were rigorous and resource intensive

" Overall review of decision making has been initiated 30 Enclosure 2