ML20062M666

To Safety Relief Valve In-Plant Test Plan

ML20062M666
Person / Time
Site:	Fermi
Issue date:	07/31/1982
From:	Lehnert R, Leonard J, Roberts C NUTECH ENGINEERS, INC.
To:
Shared Package
ML20062M638	List: ML20062M635 ML20062M666
References
DET-22-014, DET-22-14, NUDOCS 8208200168
Download: ML20062M666 (15)

Text

'

DET-22-014 Revision 0 July, 1982 Safety Relief Valve In-Plant Test Plan For the Enrico Fermi-2 Atomic Power Plant Unit 2 Prepared for:

Detroit Edison Company Prepared by:

NUTECH Engineers, Inc.

San Jose, California Issued by:

Issued by:

n f' J.

R.

Leonard, P.E.

C. W.

Roberts Project Engineer Project Director Approved by:

R.

A.

Lehnert, P.E.

)

Engineering Manager 8208200168 820818 PDR ADOCK 05000341 A

PDR

Fermi-2 SRV In-Plant Test Plan Table of Contents Page 1.0 Introduction 1

2.0 Test Objectives and Scope 1

3.0 Test Program Rationale and Instrumentation 4

3.1 Plant Unique Analysis Report (PUA) Model 4

3.2 Test Bay Selection 6

3.3 Test Instrumentation 7

10 4.0 Test Program 4.1 Procedures 10 4.2 Test Matrix 10 4.3 Data Reduction and Reports 12 13' 5.0 Program Schedule 6.0 References 13

.DET-22-014 i

-Revision 0

_.~

Fermi 2 SRV In-Plant Test Plan 1.0 Introduction Fermi 2 is equipped with safety relief valves (SRV's) to control primary system pressure transients.

When a SRV is actuated, steam is released from the primary system 5bdcompressestheairwithinthesafetyreliefvalve discharge line (SRVDL).

This compressed air enters the pool in the form of high pressure bubbles which oscillate, resulting in pressure loads on the torus shell and internal structures.

Section 1-1.3.2 of the Fermi 2 Plant Unique Analysis Report (PUAR) (Reference 1) provides a more detailed discussion of the SRV discharge phenomena.

The evaluation of the Fermi 2 SRV discharge loads presented in the PUAR utilized the alternate methodology contained in Section 2.13.9 of Appendix A of NUREG-0661 (Reference 2).

As such, SRV in-plant tests will be performed for Fermi 2 after fuel load to confirm that the methcdology used in the PUAR for evaluating SRV discharges is conservative.

This document describes the planned SRV in-plant tests for Fermi 2.

2.0 Test Objectives and Scope The objective of the Fermi 2 in-plant SRV discharge test is to confirm that the loads and structural responses documented in the Fermi 2 PUAR for SRV discharge related loads are conservative compared to_the loadings and structural responses which occur during actual SRV discharges.

DET-22-014 Revision 0 1

The test program planncd for Formi 2 is being developed in accordance with NRC guidelines for in-plant tests contained in NUREG-0661 and NUREG-0763 (Reference 2 and 3).

A discussion of the applicability of these guidelines to Fermi 2 is contained in the paragraphs which follow.

The characteristics of the T-quencher device being used in Fermi 2 are the same as those used in other Mark I plants and tested at Monticello.

The SRV discharge methodology contained in NUREG-0661 and that which is based on results obtained from Monticello in-plant tests are applicable for use at Fermi 2, as discussed in the PUAR.

The scope of the Fermi 2 in-plant test is limited to confirmation of the SRV discharge methodology used in the Fermi PUAR.

This method-ology was developed using the alternate criteria.for SRV discharge loads contained in NUREG-0661.

The criteria and guidelines contained.in NUREG-0763 for performing limited confirmatory in-plant tests are therefore, applicable.

Conservative analysis techniques are utilized to demonstrate acceptable Fermi 2 torus attached piping.and submerged structures response to SRV loads.

Section'2.13.9 of NUREG-0661 states that test measurements are not needed if such conservative analysis techniques are used.

ThNrefore, the Fermi 2 SRV in-plant test does not include measurement of torus attached piping or submerged structures response to SRV loads.

DET-22-014 Revision.0 2

Tha SRV in-plcnt test plann d for Formi 2 will focus on the following areas:

1)

Measurement of the line clearing reaction loads on the SRV discharge line and T-quencher supports.

2)

Measurement of peak pool boundary pressures during air clearing and steam discharge due to a single valve discharge (normal water leg, cold pipe) 3)

Measurement of the frequency content of the T-quencher air-bubble-transient pressure signatures.

The SRV discharge methodology used in the Fermi 2 PUAR utilizes generically developed conservative methods as defined by NUREG-0661 for determining spatial variations in air-bubble loads, load superposition methods for evaluating multiple valve actuations, load changes that accompany consecutive valve actuations, and shifts in bubble frequencies that result from variations in back pressure during air clearing.

Therefore these methods do not require confirmation by in-plant test.

Supp-ression pool thermal mixing tests will not be conducted since the evaluation of pool temperature response to SRV transients described in the PUAR demonstrates compliance with the required pool temperature limits.

Additional specific items included in the Fermi 2 in-plant test are discussed in the sections which follow.

DET-22-014 3

Revision 0 i

3.0 Test Program Rationale and Instrumentation The test procedures and instrumentation planned for the Fermi 2 SRV in-plant test are being developed in accordance with the NRC guidelines for in-plant tests contained in NUREG-0661 and NUREG-0763 (References 2 and 3).

Since a large data base has already been developed for evaluating the effects of SRV discharge in a Mark I containment, the scope of the instrumentation and testing for Fermi 2 will be limited to that necessary to meet the confirmation objectives discussed in Section 2.0.

The key elements of the Fermi in-plant test program are discussed in the sections which follow.

3.1 Plant Unique Analysis (PUA) Model The analytical approach used to evaluate the response of the suppression chamber to SRV discharge torus shell loads is discussed in the Fermi 2 PUAR.

The approach utilizes a coupled load-structure analytical model developed.in accordance with NUREG-0661 criteria.

Predicted SRV discharge torus shell pressure load magnitudes and time characteristics are developed using an analytical model based on Monticello in-plant test data.

It was demonstrated in the PUAR that the analytical model results in load magnitudes which envelop those measured in the Monticello tests.

DET-22-014 4

Revision 0

The etructural evaluation of the suppression chamber for SRV discharge torus shell loads is performed using a finite element model of a representative segment of the suppression chamber.

The analytical model includes a finite element representation of the suppresison pool to account for fluid-structure interaction effects.

A forced vibration analysis of the suppression chamber is performed for th'e Monticello-based SRV torus shell loads using the finite element

~

model discussed above.

Calibration factors, developed using Monticello in-plant test results are applied to the results to convert the forced i

vibration response to a free vibration response.

The approach is based on the observation that the phenomena associated with an SRV discharge into the suppression pool are characteristic of an. initial i

value or free vibration rather than a forced vibration condition.

The calibration factors used in the Fermi 2 suppression chamber analysis for SRV discharge torus shell loads are documented in the PUAR and will be confirmed using_ Fermi 2' in-plant test results.

l

\\

DET-22-014 5

Revision 0 l

l l

l l

3.2 Test Bay Selection The SRV discharge line (SRVDL) selected for the in-plant test is one of the shortest in length, measured from the SRV to the T-quencher ramshead along the pipe center-line.

Compared with other SRVDL's the line contains the smallest volume of air between the SRV and the submerged portion of the SRV piping in the wetwell.

This characteristic results in air clearing loads which are closer in frequency to the dominant frequency of the suppression chamber than those produced by longer SRVDL's.

As the frequency of the SRV discharge torus shell loads approach the dominant frequency of the suppression chamber, the dynamic response of the suppression chamber increases.

Although the frequencies of the air clearing loads are expected to be less than the dominant frequency of the suppression chamber, use of the selected SRVDL is expected to maximize dynamic amplification effects l

and provide a basis for confirming the model calibration factors discussed in Section 3.1.

The magnitude of the SRV discharge air clearing loads used in the Fermi 2 PUAR are developed using Monticello-1 based analytical models, as discussed in Section 3.1.

The analytical models account for different line geometries and other parameters which affect load DET-22-014 6

Revision 0

6-characteristics.

The SRVDL selected for the Fermi 2 in-plant test will provide sufficient basis to confirm that the load magnitudes used in the PUAR are conservative and, in general, confirm the adequacy of the analytical approach used in the PUAR to develop SRV discharge air clearing loads.

The suppression chamber segment selected to be instrumented is located away from major suppression chamber attachments which may have an affect on the suppression chamber frequency.

The structures contained in the selected suppression chamber segment are representative of those contained in al1 suppression chamber segments.

3.3 Test Instrumentation In order to meet the test objectives, measurements must be made of SRV discharge bubble pressures, T-quencher internal pressure, torus shell pressure and response of the torus shell and support system.

To accomplish this, pressure transducers will be placed on the T-quencher arms to measure internal and external source pressure, on the torus shell to measure the resulting torus shell pressure, and inside the SRVDL to ensure that the SRVDL water level for subsequent actuations has returned to a steady state value following SRV closure.

Transient reflood DET-22-014 7

Revision 0

charac,toristics will not be measured since a normal 1

SRVDL water lovel is assumed in the PUAR for all subsequent actuation cases.

Strain gages will be placed on the internal and external surfaces of the torus shell in a rosette configuration to provide measurements of extreme fiber and membrane stress intensities on the shell.

The pressure tranducers and strain rosettes located on the torus shell are fewer in number but arranged in a manner similar to those of the Monticello in-plant test.

Uniaxial strain gages will be located on the columns and saddle supports in the test bay to record the total integrated reaction load of the suppression chamber support system.

Uniaxial strain gages will also be located on the support columns in the adjacent bays to measure attenuation effects.

A summary of sensor characteristics is provided in Table 3.3-1.

An air bleed system will be installed on the SRVDL in the drywell to equalize the pressure between the discharge line and the drywell air space prior to SRV actuation. The existing plant SRVDL temperature and pressure sensors will be used to detect a leaking SRV.

The test instrumentation will require approximately 125 recording channels with a maximum frequency response of 200 hz.

The signals from the instru-DET-22-014 8

Revision 0

t uM

<a Ps aw PN o:

uo H

ou 1

Table 3.3-1 Summary of Sensor Characteristics Sensor Type Location Range Environment A.

Pressure transducer 1) low pressure Torus shell, external 0-100 psi Water, air, steam @ 50 psi &

quencher, air bleed 270 F (max) 2)

High pressure Internal quencher, 0-1000 psi Water, air, steam @ 700 psi &

so SRVDL 400 F B.

Strain Gages 1)

Weldable Internal torus shell, 0-0.02 in/in Water, air, steam @ 50 psi &

quencher supports 270 F (max) 2)

Foil External torus shell 0-0.02 in/in Air @ 14.7 psi & 100 F (max)

J q

w '

1 mentation will'be processed by appropriate signal s,

conditioning equipment and stored on; magnetic tape in digital format.

Each sensor will be scanned at approximately 1000 samples per second.

Approximately 25 percent of the channels'will be processed on site for comparison with the test acceptance criteria.

4.0 Test Program 4.1 Procedures Necessary procedures will be provided to outline the requirements for sensor placement and installation, qualification of test personnel, calibration of ins'trumentation, establishing pre-test conditions and conduct of the matrix tests.

4.2 Test Matrix The test matrix is presented in Table 4.2-1.

- 'a Shakedown test (s) will'be conducted to verify

^

operational procedures, to optimize test cperat' ions

• ! i.

and to establish recorder settings for realtimeii

' instrumentation.

Matrix testing consists.,of' D

t y at least four test pairs to evalu~ ate =the effects of a single valve actuation (SVA). and a subsequent consecutive actuation (CVA) of the same valve.

The number of SVA/CVA test pairs to be conducted will depend upon the data scatter encountered as determined q(

z 'y.

by a statistical review of the real-time data

' k ', b DET-22-014 10

)

T Revision 0 i

s jt}

~

~

Table 4.2-1 TEST MATRIX gg (3 M

- --. INITIAL CONDITIONS VALVE PtPE 4 F3 4

F l VALVE CLOSilRE COOLING f.[j TEST TEST TO DE SRV POOL POWER DISCHARGE TIME PHEOR PRIOR TO O L NUMBER TYPE ACTUATED PIPE TEMP (*P)

LEVEL (1)

TING (SEC)

TO CVA TEST D C)

. c3 $[

SD1 SD SRV-2066 CP,NWL See Note 1 See Note 2 10 N/A See Note 3 MT1 SVA CP.NWL 10 N/A See Note 3 wr2 CVA HP, AWL 10 1 Min 1 Min MT3 SVA CP NWL 10 N/A See Note 3 MT4 CVA HP. AWL 10 1 Min 1 Min MTS SVA CP,NWL 10 N/A See Note 3 MT6 CVA IIP, AWL 10 1 Min 1 Min MT7(4)

SVA CP.NWL 10 N/A See Note 3 MTS(5)

CVA HP, AWL 10 1 Min 1 Min MT9 SVA CP.NWL 10 N/A See Note 3 MT10 CVA HP, AWL 10 1 Min 1 Min Notes:

(1) The starting pool temperature is unknown prior to the test but all tests shall be run with pool temperature i 10* P of the starting temperature (2) Power levet sufficient to support steady steam flow through an SRV discharge line at 1000 poi at the SRV.

(3) Pipe temperature to be within i 10* P of the temperature before test MT1 or at least 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> of cooling should be provided following a previous actuation.

I (4) After performing test MT7, additional SVA tests shall not be performed unless required.

(5) After performing test M8, additional CVA tests shall not be performed unless required.

Shake Down.

Abbreviations: SD SVA - Single Valve Actuation CVA - Consecutive Valve Actuation Cold Pipe (SRVDL)

CP Hot Pipe (SRVIE)

HP NWL - Normal Water Level (In the SRVDL)

AWL - Actual Water Level (In the SRVDL)

• l7

y

/1/ -

.following the fourth and cubsequent tests but

,/

will not exceed six te's'ts.

Real time data will

/

also be evaluated against acceptance criteria

\\:

after each shakedown t,est and during the cooldown t

period following each$ pair of SVA/CVA tests.

i l

The duration of discharge selected for each test will be 10 seconds with 60 seconds between SVA and CVA tests based upon an analysis of the trends-reported in Reference 3.

4.3 Data Reduction and Reports 3.

In addition to insitu data processing, filtered i

time history and power spectral density plots in engineering units will be generated for each data channel and for all tests.

Further post processing to produce frequency averaging over u-

/ i '

several tests hcomponent bending and axial i

r strains fop!(aJbarticular' gage location, and stress intensities for strain rosette configurations e

i

~

will be conducted as required.

v;o.-

v S

e

4-

,A' final ' report;will be prepared which will contain s,

J J

l)*

$(1) a discu'ssion of the instrumentation locations, calibrations, signal conditioning system, instrument uncertainty, data collection and reduction, (2) i

~

[, /

a tabulation of maximum and minimum values of j

I all data channels for each test condition, r

(3) A discussion of test results and comparison to expected results, (4) representative plots of.all ej Q*'

DET-22-014

.' V -

12 Revision 0-(

data'in engincaring units for all conditions tested, and (5) a summary of conclusions regarding the confirmation of the Fermi PUA for SRV discharge loads.

5.0 Program Schedule Testing will be conducted with the plant operating on the bypass system at 10% power or greater and will span a period of from one to two weeks (actual test time will not exceed three days including shakedown tests).

Data reduction and analysis will follow and culminate in the issuance of the final report.

4 6.0 References 1.

NUTECH report DET-04-028-1, "Enrico Fermi Atomic Power Plant Unit 2 Plant Unique Analysis Report, Volume 1",

Revision 0, April 1982.

2.

NUREG-0661, " Safety Evaluation Report Mark I Containment Long Term Program", July 1980.

3.

NUREG-0763, " Guidelines for Confirmatory Inplant Tests of Safety-Relief Valve Discharges for BWR Plants", May 1981.

DET-22-014 13 Revision 0 4