ML18036B194

Reactor Bldg Vents Radiation Monitoring Sys Drift Rate Calculation for TVA Browns Ferry Nuclear Plant.

ML18036B194
Person / Time
Site:	Browns Ferry
Issue date:	05/26/1992
From:	Akers D, Frederick S, Laforce J GENERAL ELECTRIC CO.
To:
Shared Package
ML18036B193	List: ML18036B194
References
GE-NE-533-020-0, GE-NE-533-020-0492, GE-NE-533-20, GE-NE-533-20-492, NUDOCS 9303240040
Download: ML18036B194 (18)

Text

GE-.NE-533-020-0492 DRF: D11-00017 REACTOR BUILDING VENTS RADIATION MONITORING SYSTEM (RBVRM)

DRIFT RATE CALCULATION FOR TENNESSEE VALLEY AUTHORITY BROWNS FERRY NUCLEAR PLANT Prepared R(y: k.

J.B. La .Force Reviewed by:

S.A. Fr derick Verified by:

D.D. Akers 9303240040 9303i6 PDR ,ADOCK Q5000259

>"', PDR

~,

y),. ~ ~ <<Q,)~ ~" >

CALIBRATION FRE UENCY EXTENS ION

1. INTRODUCTION The Reactor Building Vents Radiation Monitoring Subsystem channels presently utilize a 90 day calibration frequency. The retrofit of the instrumentation with NUMAC based equipment provides an opportunity to extend the calibration interval to once per cycle. A cycle is defined as 18 months + 254.

A previous investigation (Attachment 1) examined the drift associated with the NUMAC RBVRM instrument loop and provided conclusions and recommendations pertaining to a once per cycle calibration.

This calculation, in conjunction with the aforementioned investigation, will demonstrate that the overall drift rate of the NUMAC RBVRM is within specified limits.

2. COMPONENTS AFFECTING DRIFT Potential drift sources for the RBVRM instrument loop include the sensor and converter's. Geiger Mueller (GM) tube, the sensor and converter's discriminator, the RBVRM's High Voltage Power Supply (HVPS), arid the digital circuits located in the RBVRM chassis (Attachment 1). The discriminator and the digital circuits have been considered to add a negligible amount to the channel drift due to the characteristics of the digital design of the instrument (Attachment 1)

The remaining two items have had the following drift values assigned to them: a 2% of point drift per plant cycle for the Geiger Mueller tube, and a maximum 15 volt per month drift for the HVPS (Attachment 1) .

The 24 of point value for the GM tube was derived by assigning a constant linear degradation of sensitivity over the manufacturer's stated lifetime for the tubes. This value was determined to be 24 (Attachment 1). Because the manufacturer was unable to confirm if the degradation was a catastrophic event at the end of the stated life, it was assumed for conservatism that the decline occurred over the life of the tube. For purposes of this calculation, however, a 44 of point value will be assigned to provide additional conservatism.

The 15 volt value for the HVPS was obtained from the vendor's power supply specification. It should be noted that a change the High Voltage results in a corresponding change in the GM in tube's output, depending upon the manufacturer's plateau slope.

The GM tube manufacturer specifies a maximum plateau slope of 0.34 per volt for the 5 decade instrument and a 0.08~ slope for the 4 decade instrument.

An 804 value (i.e. 12 volts) of the maximum drift (i.e. 15 volts) was utilized, per engineering judgement, to allow additional time for the operator to observe the change in voltage and take corrective action (Attachment 1).

By limiting the drift to 12 volts before corrective action must be taken to reinitialize the HVPS back to 575, a maximum cumulative deviation of only 27 volts would occur volts maximum for the first month without resetting(i.e. 11.9 the voltage, followed by the maximum drift of 15 volts prior to any adjustment, for a total of 26.9 volts). Therefore, a value of 548 volts will be used as the endpoint of the cumulative voltage drift.

3. CALCULATION In order to examine what impact that the GM tube and HVPS have on 1oop drift, the following calculation is performed. It is postulated that a reduced r'eading from the actual mR/hr value would provide the most conservative or limiting case. This is true because the actual tripping action would therefore take place at a higher value due to the under reading of the instrumentation.

To establish the impact that the high voltage reduction will have on the mR/hr reading, the following equation will be utilized:

100 R2 - R1 R1 = S V2 - V1 where V2 = nominal GM tube voltage V1 reduced voltage of interest R2 radiation reading associated with V2 Rl radiation reading associated with V1 S percent per volt relative slope

and utilizing the following association:

Plateau R2 Rl ~ ~ 1 ~ ~ ~ ~ ~ ~ ~ ~ ~

Radiation Level V1 V2 Voltage and where:

V1 548 Volts (See discussion above)

V2 575 Volts (From Attachment 1, sheet 3 of 4)

R2 100 mR/hr (Technical Specification value)

S Manufacturer's maximum plateau slope. For the 5 decade instrument S equals 0.3; for the 4 decade instrument S equals 0.08 Solving for Rl, 100 100 R1 R1 = 0.3 575 548 Rl = 92.51 for the 5 decade instrument and 100 100 R1 R1 = 0.08 575 548 Rl = 97.88 for the 4 decade instrument Therefore, a 30 day drift or reduction of 7.5 mR/hr (i.e. 100

-92.51) could be seen for the 5 decade instrument and a 2.1 mR/hr reduction (i.e. 100 97.88) for the 4 decade instrument.

Combined Drift:

The combined drift is the square root sum of the squares summation of the drifts attributed to the GM tube and the HVPS.

(1). GM tube: Drift is established at 44 of point per plant cycle. This value was derived, as previously noted in the Cycle Calibration Evaluation, from an engineering assumption that the tube sensitivity would drift downward at a constant rate during the manufacturer's predicted life for the GM tube. For the predicted life of the tube, (i.e. 105 years), the loss of sensitivity was determined to be less than 24 of point for the calibration cycle. For the purposes of this calculation, however, a value of 44 of point per will be used to provide con'servatism for this parameter.

The value of 4% of point is directly translated into a mR/hr value by multiplying (i.e..04 X 100 mR/hr =

it times the technical specification value 4 mR/hr) .

(2). Hi h Volta e Power Su l  : Since the High Voltage Power Supply deviations will be maintained within +12 volts (per conclusion 1 of the Calibration Cycle Evaluation), the resultant reading from,a reduced high voltage (i.e. 548 volts) is a maximum of 7.5 mR/hr for the '5 decade device and 2.1 mR/hr for the 4 decade device.

The combined values of these drifts are then:

((4) 2 + (7.5) 2 ) 1/2 = 8.5 mR/hr for the 5 decade instrument and

((4) 2 + (2.1) 2 ) 1/2 = 4.5 mR/hr for the 4 decade instrument Notice that although 8.5 mR/hr and 4.5 mR/hr are 30 day values, they are still applicable since the HVPS deviation will be limited per procedure to 575 +12 vDC and reinitialized exceeds this value. Therefore, the 30 day maximum value is also if it the maximum value that can occur over the plant cycle.

4. COMPARISON BETWEEN 90 DAY CYCLE and 18 MONTH CYCLE 1

The calculated values of 8.5 and 4.5 are less than the stipulated values of 10.2 and 8.14 for the 5 decade and 4 decade instruments, respectively.

5. CONCLUSIONS The calculated drift/error for plant cycle interval is within the permissible values of 8.1 and 10.2

REACTOR BUILDING VENTS RADIATION MONITORING SYSTEM (RBVRH)

CALIBRATION CYCLE EVALUATION FOR TENNESSEE VALLEY AUTHORITY BROWNS FERRY NUCLEAR PLANT DRF: 01140017 Prepared by: 3

0. 0. Akers - Principa Engineer Reviewed By: 9~ 3/2o/qZ S. A. Frederick - Senior Engineer Reviewed By  ! c2, 3 A0- V force - Principa Engineer Verified by. 3/~%e.

U. E. Oennis - Principa Engineer ATTACHM lo SHEET OF

RBVRM - CA IBRATION CYCLE EVALUATION SHEET 2 OF 4 The Reactor Building Vents Radiation Monitor (RBVRM) system Technical Specification, allowable value, is 100 mR/Hr. The RBVRM upscale trip set point, 72 mR/Hr, is the technical specification value minus the loop normal measurement accuracy. The loop normal measurement accuracy, per TVA procedures, accommodates 90 day drift. The drift allocation is 8. 1% of point for the 4 decade sensor and 10.2% of point for the 5 decade sensor.

The objective of this evaluation is to establish a basis for extending the calibration interval to one plant cycle of 18 months plus or minus 25 percent.

The identified drift sources, from sensor to upscale trip, include: GM tube, High Voltage Power Supply, and Discriminator, which are evaluated below.

1. GM TUB CHARACTERISTICS

~ &

The GM tube has no characteristic drift mechanism until onset of end of life degradation. A tube vendor (LND) quotes 5 X 10" counts as usabl.e GM life. The same value is published in the Am~erex data sheet for the ZP1300 tube. Tube life is established at 10 'ounts integrated dose to preclude onset of an end of life condition. Two different GM tubes are used in this application although the Technical Specification trip point and trip set point are the same. There are two vendors for each of the GM tubes which have the same purchase part drawing characteristics and are therefore concluded interchangeable relative to this evaluation.

. The background radiation level is assumed lower than the sensor bug source so that integrated dose is due to the bug source.

EQUIPMENT SENSOR RADIATION COUNTS YEARS PART SENSOR RANGE SENSITIVITY LEVEL PER TO NUMBER mR Hr C/~Sec mR Hr ~HR Hl) ~YA 10~ CTS 140 10'O 10 0.3 3 2.8 E7 350+ YEARS 141 142 10 TO 10 10 .3 9.4 E7 105+ YEARS 143 The 10" accumulated count, end of life, condition does not occur within the first plant cycle.

An alternate evaluation: assume that the tube sensitivity drifts at a linear rate during the predicted life with a linear degradation from 100%

to 0%. For the 105 year predicted life the loss of sensitivity is 1% of point per year or less than 2% of point per plant cycle which is well within the 8. 1% allocation.

1 ATTACHMENT SHEET 1. OF

RBVRN - CA IBRATION CYCLE EVALUATION SHEET 3 OF 4 The GH tube must be vacuum sealed to assure against a drift condition.

Tubes are 100% tested at incoming inspection. A similar test can be designed to verify the continued integrity of the GN tube installed in the sensor per SIL 327 Rev-1. Removal of the tube to perform such test, which might fracture the envelope, is specifically not recommended.

CONCLUSION: The GN tube performance characteristics support a plant cycle calibration interval.

2. HIGH VOLTAGE POWER SUPPLY DR FT RATE IMPACT ON PLATEAU SLOPE The High Voltage Power Supply specified drift rate, including the combined affect for rated change in source voltage, load, and temperature, is 2% of full scale, 1500 volts, for 2 months or 15 volts per month.

The GH tube usable plat'eau width is 500 to 600 volts and the quiescent operating voltage is 575 volts. If the high voltage power supply drifts at the maximum rate then it would drift beyond the plateau limit within two months of operation.

The'endor specified plateau slope is 0.08% per volt for the 4 decade tube and 0.3% per volt for the 5 decade tube. The permitted drift occurs in

8. 1% / .08%/volt / 15 volts/month 6.7 months for the 4 decade tube and 10.2% / .37+volt / 15 volts/month - 2.3 months for the 5 decade tube.

The operating voltage drifts off the plateau before the plateau slope variation on signal gain limits performahce. Power supply drift off the plateau is the limiting consideration.

High voltage power supply performance is monitored by the NUHAC instrument such that the voltage at the detector is observable at the instrument chassis in the control room. The high voltage should be'onitored on a periodic, initially 30 day, surveillance interval. If the indicated voltage deviates by more than 12 volts, half the allowable variation from the set value, adjustment is necessary.

CONCLUSION: Periodic readjustment of the detector high voltage is necessary to limit the drift component to an acceptable value.

1, ATTACHMENT SHEE7 OF b

I RBVRM - CA IBRATION CYCLE EVALUATION SHEET 4 OF 4

3. 0 I SCR IMINATOR CHARACTERISTICS The discriminator is set to discriminate pulse heights greater than 35 millivolts. Detector pulses height is approximately I volt when measured at the discriminator input. The discriminator circuit is constructed of precision components and no significant drift of the discriminator threshold is expected. The discriminator design does not include a threshold set point adjustment. The combination of discriminator set point to pulse height margin insignificant discriminator threshold drift assures long term stable discriminator operation.

CONCLUSION: The discriminator does not contribute to the instrument drift rate.

4. DIGITAL CIRCUITS At the output of the discriminator the signal level is counted into a digital register the contents of which are transmitted to the RS-422 input card in the RBVRM chassis then to the CPU where count rate is established and comparison is made to the digital trip reference point. The CPU is crystal controlled such that there is no significant contribution to drift rate in the digital instrument.

CONCLUSION: The digital'ircuits do not .contribute to the instrument drift rate.

CONC USIONS AND R COMM NDATONS

1. RBVRM signal channels should be recalibr ated once per plant cycle, of 18 months plus or minus 25 percent, provided HVPS voltage is maintained within plus or minus 12 volts of the 575 volt operating point on a periodic surveillance interval.

2. An accumulation of 10" counts is the recommended end of life point for each of the GM tubes.

3. Periodic surveillance of GM tube performance as identified by SIL-327 Rev-I is recommended.

ATTACHMENT SHEET OF b

VERIFICATION INSTRUCTION - RBVRM SYSTEM CALIBRATION CYCLE EVALUATION Sheet 1 of 2.

l. Verify Technical Specification allowable value 100 mR/Hr from Technical Specification BFN Unit 2, Table 3.2.A (See Laforce) ~

vy~

2. Verify Upscale trip setpoint, (72mR/Hr) from document GENE-533-02-0191, DRF Dll-00017. r g~

3. Verify drift allpcation 8.1/ and 10.2/ from document GENE-533-02-0191, DRF Dl 1-00017. v'g~

4. Verify sensor life from Amperex data sheet which becomes a part of the design package, sheet 2 of this document. +g(g

5. Verify that E10 counts is reasonable end of life assumption.

6.

7.

Verify Given life calculations.~

~

that the GM tubes have passed, lOOX inspected, incoming inspection 8.

the nominal 1 volt'ulse amplitude is credible. /

and therefore exhibit a pulse of greater than 1 volt verify that when installed in the circuit, per schematic diagram 945597 confirm whether Verify the 35 millivolt discriminator assertion from 23A5071 paragraph 3.6.5. Confir from the Sensor schematic 945E977 that this is a credible value. ~g

9. Verify the evaluation for accuracy and completeness of the evaluations and conclusions.

Q~~

Responsible Engineer:

D. D. Akers x)~o)S~

Verification Statement:

s)A

+m~

Verified By: ado g~

U. E. Dennis sHsgr s on

1

~ \

~

VERIFICATION INSTRUCTION - RBVRM SYSTBl CALIBRATION CYCLE EVALUATION Sheet 2 of 2.

ZP1300 OPKRATINQ CHARACTERISTICS (Ambient temperature ~ 26 oCI Measured In circuit of Fig,2 Starting voltage 400 V Plateau threshold voltage mex. 600 V Plateau length 100 V Recommended supply voltage 550 V Plateau slope maxa 0.3  %/V Background (shielded with 60 mm Pb with an Inner liner of 3 mm Al), at recommended pF'ax.

supply voltage max, 1 count/min Oead time, at recommended supply voltage nlex. 11 ps LIMITINQVALUES (Absolute mex. rating system)

Anode resistor min. 22 MA Anode voltage max. 600 V Ambient temperature continuous operating nlax. +70 oC Illln. -40 oC storage max. 76 oC

~ L(FE EXPECTANCY Life expectancy at s.26 oC 5 x 10'e count MEASURINQ CIRCUIT R1 <<2,2 MA R2<<47kQ CI <<1

+y~y C1 output R2 C2 on< R1CI ER2 C2 nts)ls4 Fig,2 ATTACHMENT SHEET OF b

'See General Information (paragraph 6.5) 40 June 198S

I rr .t H Y