ML20212F250

Forwards Verification of RELAP5-FORCE Hydraulic Force Calculation Code & Other Addl Info Re NUREG-0737,Item II.D.1,per NRC 861216 Request.Results of as-built Reanalysis Scheduled for Apr 1987

ML20212F250
Person / Time
Site:	Rancho Seco
Issue date:	03/03/1987
From:	Julie Ward SACRAMENTO MUNICIPAL UTILITY DISTRICT
To:	Miraglia F Office of Nuclear Reactor Regulation
Shared Package
ML20212F255	List: ML20212F250 ML20212F265
References
RTR-NUREG-0737, RTR-NUREG-737, TASK-2.D.1, TASK-TM JEW-87-172, NUDOCS 8703050051
Download: ML20212F250 (13)

Text

r

$'SMUD SACRAMENTO MUNICtPAL UTIUTY DISTRICT O P. O. Box 15830, Sacramento CA 95852-1830,(916) 452-3211 AN ELECTRIC SYSTEM SERVING THE HEART OF CALIFORNIA JEW 87-172 March 3, 1987 Director of Nuclear Reactor Regulation Attention: Frank J. Miraglia, Jr.

Division of PWR Licensing-B U S Nuclear Regulatory Commission Washington D C 20555 Docket 50-312 Rancho Seco Nuclear Generating Station Unit #1 NUREG 0737, ITEM II.D.1 REQUEST FOR INFORMATION

Dear Mr. Miraglia:

Your letter dated December 16, 1986, requested additional information to complete your review of NUREG 0737, Item II.D.1. Attached is the District's response to this request. As noted in the response to question number 7 the results of the as-built reanalysis is scheduled for April 1987.

If you have any questions pertaining to this submittal contact John Atwell of my staff at extension 3906.

Sincerely,

/

Deputy General Manager, Nuclear 1

Attachments cc: Syd Miner, NRC - Bethesda A. D'Angelo, NRC - Rancho Seco A oMo g3oRDoEkfsNh e P

DISTRICT HEADQUARTERS O 6201 S Street, Sacramento CA 95817-1899

ATTACIDENT 1 NUREG 0737, ITEM II.D.1 QUESTION 1 Provide the torque setting for the PORY block valve operator at Rancho Seco 1 and the torque produced at this setting. If the torque is less than 82 ft-lbs (the mini === torque tested by EPRI), it is the staff's position that it is not adequate to conclude proper operation solely on manufacturer's calculations. The problems encountered with Westinghouse gate valves on closing, which were traced to the calculations used to size the valve operator torque requirements, indicate the need to experimentally verify the adequacy of the block valve / operator combination. SMUD should provide test data to demonstrate the SMB-00-10 operator at Rancho Seco 1 is capable of providing adequate torque to close the block valve.

RESPONSE: The torque setting for the PORV valve has been determined to be 87.6 ft-lbs. This value has been determined as a result of the District's MOVATS testing program (IEB 85-03). The important parameter for gate valve operability is the thrust applied to the stem. MOVATS testing of the PORV block valve has confirmed that the required thrust value (7186 lbs.) will be generated by the valve notor at a torque setting of 87.6 ft-lbs. As a result of this testing, it is the District's position that the PORV block valve motor will provide the required torque to close the valve under the expected operating conditions.

r l

l

NUREG 0737, ITEM II.D.1 (CONT.)

QUESTION 2: Discuss how REIAP-FORCE calculates piping forces from REIAPS output and describe how the code was verified. Provide comparisons of RRLAP-FORCE calculations and EPRI/CE data.

RESPONSE: The thermal-hydraulic analyses were performed with the RELAPS-FORCE code, version 14, developed by Gilbert Associates.

This code was developed by revising REIAPS/ MOD 1 to include the hydraulic force equation. Verification of this modification was performed by running a sample problem without the force option and comparing it to RELAPS/ MOD 1 results to ensure that the modifications did not alter the basic RELAP5/ MOD 1 calculations.

The adequacy of the hydrodynamic force option was subsequently verified by comparing analytical results to test results from several tests including the EPRI/CE SRV test program. The report indicated reasonable agreement with test data and that RELAP5-FORCE is an acceptable program for thermal-hydraulic analyses of the S/RV discharge piping. A copy of the verification report is enclosed as Attachment 2 to this letter.

l l

l

NUREG 0737, ITEM II.D.1 (CONT.)

QUESTION 3: The submittal stated intensification factors were used to compensate for a lack of sufficient detail in the REIAP5 thermal-hydraulic analysis. How were the intensification factors developed? This was not clear from the submittal. How was the approach, i.e., applying intensification factors to results from a model with less than sufficient detail, verified? Supply the results of verification analyses of appropriate EPRI/CE tests using this approach for our review to assure this method conservatively bounds the discharge piping loads.

RESPONSE: In the nodalization study performed in Appendix C of reference 1, the minimum control volume size was 0.5 ft. in length. Using this length control volume for the Rancho Seco model would have greatly exceeded the RELAP5 program capacity. In order to overcome this problem, intensification factors were developed to use as multipliers for the forcing functions in the piping structural analyses. The forcing function for each straight pipe segment was multiplied by an intensification factor yielding values which are representative of the values expected from a finer model.

The intensification factor was determined by performing a nodalization sensitivity study similar to that performed in App.

C of reference 1.

The discharge line for valve PSV 21507 was chosen for this study as being representative of the three Rancho Seco lines. It was modeled with an average control volume length of 1.96 ft. Force time histories for both the steam discharge and subcooled water cases were developed and applied to the model. The line was

! then reanalyzed using a model with double the number of control volumes.

l The study showed that when more than 5 control volumes where used for a straight pipe segment, doubling the number of volumes increased the peak force by less than 10%. In two volumes the I force actually decreased. If fewer than 5 volumes were used in a segment, then doubling the volumes increased the forces from 10% to 30% in the steam blow-down case and 10% to 68% in the subcooled water case.

With this information, the intensification factors were applied as multipliers to the forcing functions as described below:

For pipe segments in which both the fine and coarse models had fewer than 5 control volumes an intensification factor of 2 was used, i

l b

c NUREG 0737, ITEM II.D.1 (CONT.)

RESPONSE 3: (Cont.)

For the remaining segments, which contained more than 5 control volumes in the coarse model, an intensification factor equal-to the ratio of peak forces in the fine model volume to those in the coarse model volume (with round-up margin) was used.

A more detailed discussion on intensification factors was provided in our April 12, 1985, submittal in response to question 8b. The overall thermal-hydraulic model as well as the coarse and fine nodalization models are shown as Attachments 6, 7, and 8, respectively, to that submittal.

REFERENCE:

1. EPRI Report, " Application of RELAP5/ MOD 1 for Calculation of Safety and Relief Valve Discharge Piping Hydrodynamic Loads" EPRI-2479, December,1982.

i

~

l l

i

NUREG 0737, ITEM II.D.1 (CONT.)

QUESTION 4:- The maximum time step used in the thermal-hydraulic analysis for subcooled water was 2 X 10-4 sec, based on Reference 1.

However, this reference did not consider subecoled water discharge. Analyses by EG&G Idaho, Inc. have shown that with subcooled water discharge a A x/ At ratio of about 20,000 f t/see is needed to optimize calculation of acoustic wave propagation.

With the coarse nodalization used at Rancho Seco the mini = =

node size was 1.25 ft. This would require a maximum At of 6.25 X 10-5 see rather than the 2 X 10-4 seco used. For the fine nodalization study the maximum A t should be one-half that calculated above. Provide assurance the maximum time step used in the RELAP5 analysis was small enough so that the forces calculated from the RRLAP5 output are not underestimated due to numerical smearing.

RESPONSE: The time step used in a finite difference solution is chosen according to the Courant criteria, as follows:

dt(dx/C Where:

C= acoustic wave velocity dx= node length dt= time step The saturated steam C is approximately 1500 ft/sec. The corresponding time step for a minimum node length of 1.25 ft. is therefore 0.0008 sec.

For subcooled water C is approximately 4000 ft/sec. The corresponding maximum time step for a node length of 1.25 ft. in this case is therefore 0.0003 sec. It should be noted that the acoustic velocity of 20,000 ft/see suggested in the question is on the order of the speed of sound in steel. This velocity is Considered to be overly Conservative for use in subcooled water analyses.

l l The maximum time step used in the thermal-hydraulic analysis was j 0.0002 sec. We feel that this was well within the criteria j limits to guarantee the RELAP output was not underestimated by l numerical smearing. This is consistent with the recommendations of section C4 of reference 1 (Question f3).

l l

l l

NUREG 0737, ITEM II.D.1 (CONT.) i QUESTION 5: The RELAP5 thermal-hydraulic analysfa used a PORY opening time of 0.50 see for the Dresser PORY installed at the plant. A review of the test report showed opening times for the Dresser PORY were less than 0.23 sec. Confirm the PORY opening time

-used in the analysis is appropriate and results in the calculation of conservative piping loads.

RESPONSE: The original thermal-hydraulic analyses employed a PORV opening time of 0.050 sec. , as correctly stated in our April 12, 1985, responses to questions 8a and 8b. However, the opening time was mistyped as 0.50 sec. in our response to question 8c.

Therefore, the PORV opening time of 0.05 seconds was used in the analysis and is less than the 0.23 seconds cited.

NUREG 0737, ITEM II.D.1 (CONT.)

QUESTION 6: The submittal stated a cutoff frequency of 50 Hz was used in the piping analysis. The cutoff frequency of 50 Hz appears to be too low for piping r.nalysis based on EG&G Idaho, Inc.

experience. Most piping analysis uses a cutoff frequency of 100 Hz. The submittal mentions that 50 Hz is more conservative than the NRC guideline of 33 Hz. This guideline, however, is only for earthquake analysis. Provide assurance the use of a cutoff frequency of 50 Hz does not invalidate the analysis performed.

This also affects the maximum time step used in the analysis since the maximum time step is the reciprocal of 5'X (highest natural frequency).

RESPONSE: In the structural analyses of the pressurizer S/RV discharge piping, a cutoff frequency of 50 Hz was used in the seismic portion of the analyses. In the thermal-hydraulic dynamic analyses, however, a cutoff frequency of 300 Hz was used. The results of these two analyses were then combined by the SRSS method to obtain the total response of the system to dynamic loading.

The time step used in the analyses of thermal-hydraulic loads was 0.0005 see for the saturated steam case which compares favorably with the maximum recommended time step of the reciprocal of 5 times the highest forcing function frequency or 0.00067.

L i

4 1

4 NUREG'0737, ITEM II.D.1 (CONT.)

QUESTION 7: The submittal' indicated the limiting transient for piping stresses was the steam discharge transient and not subcooled water discharge. This result is different from the results of a similar B&W plant (TMI-1) which found, with similar initial conditions, that subcooled water discharge was the limiting case. .In addition, EPRI/CE testing found water seal discharge resulted in greater piping loads than steam discharge without water seals. Therefore, provide more details on the differences in the stress analysis between the steam discharge case and the subcooled water discharge case. Include reasons why the steam discharge case was the limiting transient. Also include a table comparing the calculated stress with the allouable stress for the most highly loaded pipes and supports for both the steam and water discharge cases.

RESPONSES: As discussed in our April 12, 1985 submittal (Question 8) the Rancho Seco thermal-hydraulic analyses considered transients involving both steam and subcooled water discharge through the SRV's and.PORV. Under these transients, the saturated steam discharge' cases generally produced higher thermal-hydraulic piping reactions. It was judged that these would be the more credible transients especially when combined uith seismic forces. The stress analyses for the subcoolel cases were therefore not completed.

The concern of greater piping loads from water seal discharge is not applicable to Rancho Seco since there are no loop seals at the safety and relief valves. Thus the extremely high dynamic forces due to a slug of cold water being forced down the discharge piping by high pressure steam were not analyzed at Rancho Seco.

Based on these analyses, the piping system was upgraded in 1986 to meet the anticipated thermal-hydraulic and other loads.

During the performance of confirmatory analyses to verify piping stresses under the as-built support locations it was determined that the system modeling required additional refinements to accurately predict the system response to dynamic loads. A reanalysis effort is currently underway to improve the model and fully verify system adequacy under all the design basis loadings including the subcooled cases. Upon completion of this analysis the results will be provided to the staff. It is anticipated that the reanalysis will be completed by April, 1987.

L

NUREG 0737, ITEM II.D.1 (CONT.)

QUESTION 8: SMUD in its. response to our request for information stated that ,

the qualification requirement of NUREG-0737, Item II.D.1, '

pertains to qualification of the block valve and does not address qualification of the PORY control circuitry.. However, environmental qualification of the PORY control circuitry is exactly what was required by Item II.D.1. It is the staff position that in order to demonstrate the Rancho Seco 1 control circuitry is qualified to the requirements of NUREG-0737, the design qualifications must be compared to the environment the control circuits will be exposed to. Provide documentation to show the PORY control circuitry has been qualified under 10 CFR 50.49, or to allow a complete review of the qualification of the control circuitry for the PORY under NUREG-0737, provide the following:

A. Provide a list of all PORY control circuitry needed to mitigate NUREG-0737 transients.

_ _ . . _ B. For each item of equipment identified in A, provide the following:

1. Type (functional designation)

2. Manufacturer

3. Manufacturer's type number and model number

4. Plant ID/ tag number and location C. For each item of equipment listed above, provide the environmental envelope, as a function of time, that includes all extreme parameters, both maximum and minimum values, expected to occur during NUREG-0737 transients, including post accident conditions.

D. For each item of equipment identified above, state the actual qualification envelope simulated during testing (defining the duration of'the environment and the margin in excess of the design requirements). If any method other than type testing was used for qualification, identify the method and define the equivalent " qualification envelope" so derived.

E. Provide a summary of test results that demonstrated the adequacy of the qualification program. If any analysis is used for qualification, justification of all analysis assumptions must be provided.

7 NUREG 0737, ITEN II.D.1 (CONT.)

~ QUESTION 8 -(Cont.)

F. Identify the qualification. documents that contain detailed

. supporting information, including test data, for items ~D and E.

RESPONSEt' NUREG'0737, Item II.D 1 requested licensees to review relief and safety valve operating conditions expected.due to accidents and anticipated operational occurrences referenced in' Regulatory Guide 1.70, Revision 2. Resulting qualification of the safety and relief valves was to. include qualification of associated .

. control circuitry, piping and supports in addition to the valves themselves.

The District has_ developed a. list of equipment required for accident mitigation based on the accident and transient analyses contained in Chapter 14 of the Rancho Seco Updated Safety Analyses Report (USAR). This list was developed in response to

~

10CFR50.49 and-has been documented by a District calculation.

The results of this review reveal that the PORY is not required to provide primary system overpressure protection as a result of any USAR design- basis transient. In all Chapter 14 events where overpressure occurs in the primary system the safety relief valves are assumed to lift, thus providing overpressure protection. This is consistent with the description / intent of, the EMOV (PORV) and Code Safety Valves found in the USAR (Section 4.2.4). The EMOV is required to remain closed to maintain the pressure integrity of the RCS while exposed to post accident environmental conditions. As a result, the solenoid has been passively qualified to ensure that the valve remains in the closed position. Thus failure of the solenoid will not result' in a spurious actuation of the PORV as a result of short circuit, open circuit or loss of power in accident conditions.

Additional information on the qualification status of the PORY control circuitry is provided in Table 1. As noted in the Table all devices associated with the PORV circuitry have been qualified to 10CFR50.49 for the environment in which each is expected to operate.

The only time Rancho Seco requires use of the EMOV for overpressure protection is during LTOP conditions. Under the conditions that would exist in this event the current EMOV and associated circuitry has been found to be adequate to fulfill the overpressure protection function.

Based on this, it is the District's position that the PORV control circuitry does not require further upgrade of its qualification beyond the current qualification as presented in Table 1 and described above. All qualification documentation is available at Rancho Seco for your inspection.

s 7

. i 1

- TABLE 1 QUALIFICATION STATUS OF PORY CONTROL CIRCUITRY ENVIROIRENT&LLY QUALIFIED DEVICE TAC MANUFACTURER'S TYPE NO. PLANT LOCATION IBISER ,

NO. TYPE (PtBICTION DESIGNATION) MANUFACTURER AND MODEL NO. (BQ ENVIRONMDff) 10CPR50.49 S2A1 Vital 480V AC MCC Ceneral Electric Co. 325x915 West 480V Swar. Rm. Not Required Aux. Bids. (Mild)

PSV-21511 Solenoid General Electric co. S/N BM09686 Containment (Harsh) Yes*

Connector BNC to Mirco Dot Adaptor TEC Mone Containment (Harsh) Yee FT-21038 RCS Pressure Transaitter Rosemount 1153CD9 Containment (Harsh) Yes PT-21040 RCS Pressure Transmitter Rosemount 1153CD9 Containment (Harsh) Yes PT-21092 RCS Hot Leg "B" Control DHS Could PC3200-0$M Containment (Harsh) Yes Pres. Permissive Interlock PSH-21511 Pressuriser Relief Bailey Control 6623819-1 [H4 PIA] Control Room .Not Required (Mild)

PSH-21092 EM07 Low Setpoint Bailey control 6623819C1 [H4 PIA] Control Room Not Required (Mild) ,

IE-21524 Position Indicator on PORY TEC 2273A Containment (Harsh) Yes.

PSV-21511 XE-21525 Position Indicator on PORY TEC. 2273A Containment (Harsh) Yes PSV-21511 XY-21524 Charger Amplifier for TEC 504A Containment (Hersh) Yes XE-21524 XY-21525 Charger Amplifier for TEC 504A Containment (Harsh) Yes XE-21525 H3CRR Aux. Control Relay Panel General Atomic None West 480V Swgr. Rm. Not Required Aux. Bldg. (Ra. 217)

(Mild)

H1RC Reactor Coolant Console Babcock & Wilcox None Control Room (Mild) Not Required H4PIO1 NNI Cabinet Not Found None Control Room (Mild)'- Not Required H4EFA Nuclear Interface Automation Industries None NSEB Elect. Equip. Not Required Instrumentation Rack Room (Mild)

H4EFB Nuclear Interface Automation Industries None NSEB Elect. Equip. Not Required Instrumentation Rack Room (Mild)

H7RP50 Containment Penetration Conas 2325 Containment (Hersh) Yes H7RP63 Containment Penetration Conax 7073 Containment (Hersh) Yes H7RP21 Containment Penetration Conax 7073 containment (Harsh) Yes Splices Electrical Connection Raychem- Various Containment (Harsh) Yes Terminal Termination (H7RP50, Kulka 7TB8/7T512 Containment (Hersh) Yes Blocks H7RP63 H7RP21)

N/A 600 Volt Cable Brand Rex XLPE Insulation / Containment (Harsh) Yes Hypelon Jacket N/A 600 Volt Cable Cerro XLPE Insulation / Containment (Harsh) Yes Neoprene Jacket N/A 600 Volt cable Eatom XLPE Insulation / Containment (Harsh) Yes Hypelon Jacket-N/A 600 Volt cable Rockbestos XLPE Insulation / Containment (Harsh) Yes Neoprene Jacket

• Passive qualification meets 10CPR50.49, not qualified as an active component.

Ok ,

(

ATTACHMENT 2 VERIFICATION OF THE RELAP5-FORCE HYDRAULIC FORCE CALCUIATION CODE

>