ML20128B142

Discusses Util Reporting Plant Condition Where All Six of APRM Channel Signals Decreased from Measured Thermal Power Level of 68%

ML20128B142
Person / Time
Site:	Monticello
Issue date:	06/15/1972
From:	James Shea US ATOMIC ENERGY COMMISSION (AEC)
To:	US ATOMIC ENERGY COMMISSION (AEC)
References
NUDOCS 9212030573
Download: ML20128B142 (2)

Text

~T t

.eL 4 s

• $ 4** ' UNITED STATES -

f% ATOMIC ENERGY COMMISSION l4 C\- b k)D,1' L W A5HINGT ON, D.C. 20545 JUN 151972 Files (Docket No. 50-263)

THRU: L. emann, Chief, ORB #2, L

' 7(, Ame MONTICELLO - POWER LEVEL SIGNALS FROM ALL (SIX) APRM CHANNELS By letter dated May 25, 1972, Northern States Power Company (NSPCo) reported a plant condition where all six of the Average Power Reactor Monitor (APRN) channel signals decreased from the measured thermal power icvel of 68%, an average 4 4% power, to an average of 64%.of rated power during a 12-hour u; Each of the six AP.RM systems averages input .eignals f rom 24 L are detectors at various radial and axial locations to represent core thermal power. The APRM subsystem (6 APRM cnannels - three channels provide trip inputs to one reactor protcetion system logic chanael and the other three APRMs feed to the other. logic-channel) must continuously indicate the average core power 1cvel and initiate trip signals for -control rod block and automatic reactor

~

scram to prevent excessive ne6Eion~ flux"leveli. ~ Changes in the core power anape uaring start ups caused by xenon transients and variations i

in control roc patterns tend to decalibrate the APRM channels according to the USPCo letter, but according to the FSAR (page 7-4.10), the-APRM provides valid average power measurements during typical rod or flow induced power level change. Expected accuracy over a wide range of power levels was expected to be 5%.

The decalibration of all channels in one directica (downward) with an average resultant APRM accuracy of 5.7% may. represent an extreme statis-tical condition but it is noteworthy because it occurred at essentially stable power conditions (i.e. negligible coolant variation or control rod movemen t) .

In response to questioning during a 5/31/72 telecon, IGPCo representatives stated that the recirculation flow rate prior to APRM _ channel gain -

adjustments was 58%, that all 96 Local Power Range Monitors (LPRMs -

4 x 24) were in service at the time, t hat individual LPRMs had been calibrated with TIP about two weeks ca. 1er, that most of the APRM decalibration occurred during the first 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />, and that several axial scans were made at the time to determine exial power shape.

The axial power level scans showed that peak power occurred initially near the LPRMs at the 7.5 ft elevation of'the core because of xenon peaking at. the bottom and moved downward toward the center causing the '

C LPRMs initially at the- peak power location to decrease noticeably without appreciably changing the lower A and B monitors or the top D monitors. NSPCo' has noted and we concur that such behavior could account for the observed nuclear sensor behavior.

'Q 9212030573 720615

.PDR ADOCK 05000263 S_ PDR -

d

_ Files -

2-JUN 151972 As reported by NSPCo, the nuclear power level signals, although noticeably lower than true power, were capable of providing overpower protection well within the Technical Specification limits and, therefore, reactor-safety was not compromised during the recorded 12-hour period that the nuclear APRK qignals decreased and remained below true power. Ultimately, the APRM channel gain was adjusied to restore calibration.

Conclusion i

In accordance with FSAR Figures 2.1.1 and 2.3.1, the APRM signals could have been low by 7% power level while continuing to provide adequate reactor protection (i.e. prevent core damage due -to overpower transients).

As repor ted by NSPCo, two of the three APRM channels on each of the two reactor protection channels were 2% and 3% below true power, well within the allowable error. The FSAR (page 7-4.10) also notes that periodic calibration of the APRMs -is required. However, there is no documented evaluation of the' maximuin possib'1'e downward' decalibration of the APRM signals or the required frequency of APRM calibration (gain adjustment) at s table power -levels. This omission may be common to all of the Monticello type or larger boiling water reactors. During a telecon on 6/1/72, we requested that NSPCo consider by 7/1/72, a change to the Technical Specification encompassing the requirement that the APRM channels be calibrated following power level changes of specified mag-aJ tude and every four hours thereaf ter until the downward calibration errors are less than 2% and advise us of your decision. A sui. table alternate for such a proposal might be to present an evaluation of APRM errors during reactor power reshaping showing that the maximum downward error for APRMs under all operating conditions is within the '

allowances of Technical Specifications Figures 2.1.1 and 2.3.1. .

1 n~ lq w i

_ C.J.Shea s

perating Reactors Branch #2 Directorate of Licensing cc: D. J. Skovholt T. J. Carter D. L. Ziemann J. J. Shea R. M. Diggs M. Jinks (2)

CO (3)

-,-g,--,.--. y,- .* y. , ,-wa y c _

--y.