ML20247B319

Forwards Response to NRC 870209 Request for Addl Info Re Application for Amend to License DPR-54,consisting of Proposed Amend 127 Concerning Main Steam Safety Valve Analysis Consistent W/Baw 10043 & Benchmark Analysis

ML20247B319
Person / Time
Site:	Rancho Seco
Issue date:	03/24/1989
From:	Keuter D SACRAMENTO MUNICIPAL UTILITY DISTRICT
To:	Knighton G NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
AGM-NPP-89-040, AGM-NPP-89-40, IEIN-86-005, IEIN-86-5, NUDOCS 8903290353
Download: ML20247B319 (11)

Text

_ -

~ ..

L i es _ . _

( )SMU-SACRAMENTO MUNICIPAL UTILITY DISTRICT O 6201 S Street, P.O. Box 15830, Sacramento CA 95852-1830.(916) 452-3211 l AN ELECTRIC SYSTEM SERVING THE HEART OF CALIFORNIA 1

l AGM/NPP 89-040 March 24, 1989 x

U. S. Nuclear Regulatory Commission Attn: Document Contrid Sask Hashington, DC 20555 Docket No. 50-312 Rancho Seco Nuclear Generating Station License No. DPR-54 PROPOSED AMENDMENT 127. REQUEST FOR ADDITIONAL INFORMATION Attention: George Knighton Pursuant to your letter dated February 9,1987, attached is the District's response to the request for additional information regarding Proposed Amendment 127.

Members of your staff with questions requiring additional information or clarification may contact Mr. Robert Roehler at (209) 333-2935, extension 4918.

Sincerely, h(b d b.tuv+

4 Dan R. Keuter Assistant General Manager Nuclear Plant Manager Attachment cc w/atch: J. B. Martin, NRC, Halnut Creek A. D'Angelo NRC, Rancho Seco gg32;ggggggggggp y

P RANCHO SECO NUCLEAR GENERATING STATION O 1444o Twin Cities Road, Herald, CA 95638 9799;(209) 333-2935

Attachment

• AGM/NPP 89-040 Page 1 of 10

1. NRC REQUEST Sacramento Municipal Utility District (SMUD) has indicated in its letter dated November 14.-1985 that the assumptions used in B&W Document 86-1153322-00, " Rancho Seco: Main Steam Safety Valve (MSSV) Analysis," is consistent with the assumptions used in the B&W Topical Report BAH 10043, " Overpressure Protection for B&W Pressurized Hater Reactors." The former report concluded that Rancho Seco could meet the requirement of overpressure protection of the secondary system when the plant is operated at 112 percent of rated power with three out of nine MSSVs per steam generator inoperable. However, the latter report concluded that the capacity margin for the main steam safety valves is only 6 percent. Provide discussion on the above discrepancy in capacity margin for the MSSVs and justify that the proposed change to the technical specification will provide reasonable assurance of adequate overpressure protection of the secondary system.

DISTRICT

RESPONSE

B&W Topical Report BAH-10043, " Overpressure Protection for B&H Pressurized Hater Reactors" was issued in 1972. This topical report is based on analytical results from the POWERTRAIN computer code and shows that the B&W plant design is adequately protected against overpressure events on both the primary and secondary sides. The POWERTRAIN code simulated the NSSS, i 9

including several key secondary plant features such as the main feedwater system from the feedwater pump suction to the Once Through Steam Generators (OTSGs), and the steam lines from the OTSGs to the main turbine. Major control and relief valving was also included in the model. It was used extensively to .

simulate many upset (anticipated) events and power maneuvers because it included in its simulation a model of the Integrated Control System. Since the version of POWERTRAIN that generated the results cited in BAH-10043 preceded the operation of the first B&H commercial power plant, there was no opportunity to benchmark the code to actual operating plant data prior to its initial application. However, pruder.t selection of conservative inputs was considered adequate to ensure conservative results from the code.

More recent computer codes and models (and later versions of POWERTRAIN) have been refined to provide more realistic estimates of actual plant transient performance. RELAP5/ MODI is one such code that provides best estimate NSSS responses for a wide spectrum of events. Through the application of the code to actual plant analyses and its benchmarking to the Multiple-Loop Integral System Test Program and to operating plant data, the code has been validated quite extensively. The analyses performed to suoport the Rancho Seco Technical l

' Specification revision reflect a more accurate assessment of  !

plant response than did the original P0HERTRAIN work reported in BAH-10043. By continued application of the original acceptance criteria and conservative assumptions, the design basis for overpressu e protection remains unchanged.

1

Attachment AGM/NPP 89-040

[

Page 2 of 10 B&W Document 85-1153322-00 uses the same conservative assumptions as BAH-10043 and shows that the maximum pressure reached on the secondary side is about 1165 psia, which is 5 psi below the acceptance criteria with the assumption of the three lowest setpoint MSSVs out of service. As noted in the previous submittal, however, the MSSV model assumed that all functional MSSVs on each OTSG header lift at 1.M above their nominal setpoints and do not reach full release capability until 3% accumulation. No credit is taken for increases in steam flow rate that would occur as steam pressure increases above the pressure associated with the full. lift point.

Consistent with all other safety analyses, the MSSV flow rates are taken to be vendor-stamped rated, not those expected at full lift.

Even though the analysis assumed the three lowest setpoint MSSVs were out of service for the limiting case that is reported, this valve configuration would not be allowed by the ASME Code. The Code requires at least one safety valve be set at the design pressure for the protected system. Here both of the MSSVs set at this pressure on any one OTSG declared inoperable, another valve would have to be reset to this value or the inoperable valves _ restored to operable condition.

Hence, there is additional margin in the reported results since lower set valves in service have a larger impact on peak pressures than those set at high pressures.

Similar analyses (4), performed using the TRAP 2 code (5) for another B&W 2772 MHt plant with similar installed MSSV capacity, also demonstrate large margins in the installed MSSV capacity. This analysis used the same assumptions as BAH-10043. With the three lowest MSSVs assumed out of service and all remaining MSSVs set at 1100 psig, the turbine trip event from 112% power yielded a peak secondary system pressure of 1181 psia. This result is for a situation more conservative than was analyzed for Rancho Seco because of the assumption of ,

high setpoints for the operable MSSVs; however, the peak secondary pressure is within 16 psi of the RELAP result. The TRAP 2 model was a complete NSSS simulation and showed that for this case the peak RCS pressure was 2681 psia. Plots showing these results are included in Figures 1 and 2. The same case wasrerunwitg81pnly at 1100 psig. two The MSSVs results out ofaservice showed and all others peak pressure in the set OTSG of 1170 psia and peak pressure in the RCS of 2678 psia.

These results, shown in Figures 3 and 4, demonstrate the relative insensitivity of the RCS peak pressure to the MSSV capacity in service.

The operating history of B&W plants since the first unit's startup has confirmed that the original overpressurization analytical results were conservative. A survey of all D&H plant reactor trips from January 1980 through December 1987 showed that only four of 253 plant events resulted in RCS pressure above 2400 psig. W The allowable design overpressure is 2750 psig. This survey is based on B&W Owners

_ _ _ __a

Attachment

~_

,. .AGM/NPP 89-040 Page 3~of-10 Group (B&WOG) Transient. Assessment Reports'. During this period the plants had a high. pressure trip setpoint of 2300 psig and m anticipatory reactor trips on loss of both main feedwater pumps and on trip of the main turbine. However, the PORV setpoint was 2450 psig. The four events that did exceed 2400 psig involved post-trip events in which RCS pressure was increasing-relatively slowly due to slow coolant overheating or because of HPI operation that increased the pressurizer level.. In such cases, the pressure relief capacity requirement is a small, fraction of that required to protect against=the design basis.

event (a continuous rod withdrawal event from zero power).

Operating experience is not well_ documented prior to 1980.

During this period the high pressure trip setpoint was-2355 psig, the PORV was 2255 psig.. No anticipatory trips were-in use. One lift of an RCS safety valve is thought to have occurred at one of the Oconee plants when _the pressurizer spray and PORV were both isolated for a turbine trip test from 100% power. Peak RCS pressure was approximately 2500.psig.

It is not known whether one or both RCS safety valves lifted.

This event and other operating experience demonstrates that.

the B&H operating plants are well protected against RCS over-pressurization. Although none of-these events involved as limiting assumptions 'as were used in the topical report, the wide margin to the overpressure limits exhibited by actual plant experience suggests that the analyses were significantly conservative.

The same observations are true with regard to secondary side peak pressures. It has been a rare occurrence at a B&W plant for steam pressure to exceed 1100 to 1110 psig as measured on the steam line pressure instruments. Most plant trips involve peak pressures in the range of 1060 to 1100 psig. This compares very favorably to maximum design overpressure of 1155 psig.

In summary, the computer codes in current use, i.e., RELAP5 and TRAP 2, have been demonstrated to provide reasonably accurate predictions for plant upsets. As a result, excess margin that was provided for in the original equipment design can be '

quantified and used to provide additional operational flexibility in the plant without jeopardizing equipment safety or the health and safety of the public. The cumulative operating record for B&H designed plants supports the conclusion that approximately 30 to 35% excess MSSV capacity exists in these plants. Forced removal of up to three MSSVs per OTSG, in itself a highly unlikely occurrence, would not jeopardize either primary or secondary side overpressure protection. This conclusion is based en the simulation results which show that three valves can be taken out of service l' without exceeding the acceptance criteria for overpressure protection even when extremely conservative assumptions for several systems or equipment operability are applied.

I Attachment J

.. . AGM/NPP 89-040

2. NRC Page'4 of 10 REQUEST SMUD presented, in B&W Document 86-1153322-00, the results of a benchmark analysis which compared the results of an analysis J using RELAP5 with the Davis Besse turbine trip transient data  !

obtained in November, 1982. However, the RELAPS model does not ,

account for any turbine bypass relief capability while the i Davis Besse transient includes the actuation of the turbine bypass and the atmospheric dump valves. The staff does not consider the above stated benchmark analysis valid.

Provide justification, including supporting code verification and appropriate sensitivity analysis demonstrating the .

conservatism of the methodology used, for the use of the P.ELAP5 code for calculating secondary system overpressure transient.

DISTRICT

RESPONSE

The staff's observation that the benchmark case used in reference 3 did not simulate the turbine bypass valves (TBVs) is correct. This.was known prior to making the comparison case and was done so with the following rationale. At the Davis-Besse plant (from which the benchmark data was taken),

which in this regard is similar to all other B&W units, the TBVs have a lift setpoint of approximately 1015 psig for post-reactor trip steam pressure control. With.an anticipatory trip installed, reactor trips occur promptly after a turbine trip and the TBVs immediately revert to a post-trip control mode. The effect of this is that the TBVs do not begin to open until steam pressure has exceeded their setpoint. The TBVs are air-operated valves with a stroke time of approximately 5 seconds; hence, following a turbine trip the TBVs do not reach their full open position for at least 5 seconds after the setpoint is reached. By comparison, the safety valves have opening times measured in hundreds of milliseconds. The effect on the peak steam pressure is relatively minor since it occurs within a few seconds after the turbine trips.

In the case used for the benchmark, the plant data confirms the model's ability to properly account for the rapid change in the primary-to-secondary heat transfer and to calculate the rate of increase in the steam pressure. The comparison clearly shows that RELAPS predicts a conservatively rapid rate of increase in steam pressure which begins promptly after the steam flow to the turbine stops and before any possible TBV's impact on the secondary steam pressure.

In addition, the RELAP5 model shows good agreement with the TRAP 2 predictions for both the rate of steam pressure increase and the peak secondary side pressure that is reached during the event.

In summary, the RELAP5 model bounds the rate of increase of steam pressure following a turbine trip as evidenced by comparison of the model's prediction to plant data. The RELAPS results should be considered valid and conservative estimates of expected plant response for the turbine trip event.

.'s

~

Attachment

..c .' 'AGM/NPP 89-040.

Page 5 of.._10.

=3. NRC-REQUEST SMUD stated in a' letter dated November 14, 1985 that.the

_ proposed changes of the technical _ specification regarding the -

MSSV availability do not involve' aisignificant' reduction in a margin of ' safety. Justify this _ determination:in light of the

' reduction in pressure relieving capacity margin of the secondary system due to the proposed technical specification .

change to Section 3.4.2.1, which will permit a 33% reduction in z secondary system pressure: relieving capacity during.ful1~ power operation at Rancho Seco.  ;

. DISTRICT RESPONSE-The margin of safety designed into the plant would not be reduced by. permitting plant operation.with up'to three MSSVs out of service because.the safety margin provided by the design remains unchanged. The proposed amendment to the Technical Specifications does not alter the design basis for the i overpressure; protection of.either the RCS or the secondary side.: The design basis _ transient, i.e. the-turbine trip from i overpower conditions, and its attendant conservative-  !

assumptions, are the same as used in the original overpressure l protection _ analysis'. The proposed' amendment _ seeks to take ,

advantage of the additional margin in the plant between the l plant's expected response and_ the original acceptance' criteria for the event.

The Rancho Seco plant design was defined prior to the time that -1 any 177 FA plants were operational, thus preventing direct j feedback of plant data into the design process. Sinc 6 that -

-time, several. features of the plant's_ conservative design have become known'either through direct observation of the plant's behavior or enhanced analytical techniques. The main steam j relief system is one in which.large design margins have been  ;

identif.ied and equipment reliability has been confirmed. j Since the B&W plant design results in lifting of the MSSVs following all medium to high power reactor trips, there is l

- substantial transient experience to verify the reliability of the MSSVs to lift on demand. Studies performed by the B& HOG l have concluded that the HSSVs are v9ry reliable in their action to lift and relieve steam pressure.t8> To enhance the MSSV reliability, especially with regard to the blowdown characteristics, the B&H0G has developed generic maintenance guidelines for these valves. Thus, MSSV reliability to properly reseat should become commensurate with their reliability to lift. Operating experience in 1988 suggests that this improvement is already being achieved.

o

. _ _ _ . .-_____--_____-__-_.--__--__J

Attachnent AGM/NPP 89-040 Page 6 of 10 Excess MSSV capacity has been quantified through the use of_ improved. computer codes and field data, as discussed in District Response 1,. and has confirmed that there is significant margin to the overpressure protection acceptance criteria of the plant. The TRAP 2 analysis is an independent demonstration of this margin in the overpressure protection equipment.td) .B&H expects that a TRAP 2 analysis using MSSV setpoints as assumed in the Rancho Seco analysis would have confirmed the reported Rancho Seco results.

In summary, six of nine MSSVs on each OTSG are adequate to provide for overpressure protection for the secondary side of the OTSGs and steam lines. With this Technical Specification amendment there will be no loss of safety margin of the plant; hence, plant equipment and the health and safety of the public will be adequately protected against overpressurization events, l

i

. ._.._____________________._._w

Attachment

." AGM/NPP 89-040 Page 7 of 10

4. NRC

. REQUEST l- In the IE Information Notice No. 86-05 regarding MSSV test failures and ring setting adjustments, NRC has expressed its concern on the MSSV problems. Describe the testing and maintenance program used at Rancho Seco to preclude similar 1 problems from occurring and thereby provide assurance of I.- proper MSSV operation.

DISTRICT

RESPONSE

The District has a number of programs in place to assure proper MSSV operation.

• Increased monitoring of MSSVs during plant transients using acoustic monitors. The acoustic monitors are calibrated using values established during flow tests performed during our 1988 restart and test program.

a. PMs have been generated for setpoint verification and periodic monitoring (using ultrasonic equipment) to identify leakage.

• The requirements of ANSI /ASME OM-1 have been incorporated into our mainttuance and testing procedures.

• Active participation in the B&WOG Valve Task Force.

Recommendations of the Valve Task Force have been incorporated into our procedures. Valve performance has been evaluated following plant transients using the recommendations of the Valve Task Force.

1

Attachment AGM/NPP 89-040 Page 8 of 10

5. NRC-REQUEST In lir' ' of the industrial experiences on the use of MSSV ident sd in the IE Information Notice No. 86-05, provide justii .ation on how the proposed technical specification will provide secondary system overpressure protection with reliable facilities and high confidence level.

DISTRICT

RESPONSE

The District considers that the procedures, programs, and testing for these valves provide high reliability and resolve the valve disc travel issue. The following summarizes this position:

• Procedures and programs currently in place incorporate vendor and industry recommendations with Quality verification points to ensure that the ring settings after maintenance are in accordance with Dresser Industries recommendations.

• The data from recent trips and testing show better responses for setpoint and blowdown. '

e The District's programs (monitoring and testing) have identified one valve that does not respond well. This valve has been gagged and taken out of service.

• Toledo Edison has recently performed extensive testing of Dresser valves of the same size and model as those at Rancho Seco. The ring setting combinations tested did not result in inadequate disc travel. Based on this testing it appears that the disc travel issue is not applicable to Dresser valves.

p L - - - - - - - - - - - - -

, Attachment

. AGM/NPP 89-040  !

Page 9 of 10 L 6. NRC ]

! REQUEST j Section 1.1 of the B&W Document 86-1153322-00, " Rancho Seco: l l Main Steam Safety Valve (HSSV) Analysis," states that. primary I system pressure is not of major concern in this analysis as L events which require MSSV actuation result in primary side L pressurization to the Reactor Protection System (RPS) high i- pressure setpoint and consequent reactor trip within the first ten seconds of the. event. For design basis event, a turbine trip from overpressure, an anticipatory reactor trip prevents ]

any significant RCS pressure increase. This statement is not  !

consistent with the assumptions listed in Section 3.1 of the I report which states that no direct trip is assumed to result J from turbine trip. Also, the report does not provide I sufficient discussions to address the effect of the primary 1 system pressure transient due to a reduction of the secondary system pressure relieving capacity. Provide the results of such analysis.

DISTRICT

RESPONSE

Section 1.1 of the referenced report (8) intended to make clear the following points. First, the bounding case for secondary side overpressurization does not cause an over-pressure condition on the primary system due to the RPS shutdown of the reactor on high RCS pressure and the action of the pressurizer safety valves to limit the RCS pressure rise.

Second, Rancho Seco has in place an anticipatory reactor trip on turbine trip that would virtually eliminate any RCS pressure increase following a turbine trip, even if from an initial  !

condition of overpower. The RPS high pressure trip function is  !

assumed operable in all safety analyses scenarios except ATHS, whereas the anticipatory trip on turbine trip is not. The acceptability of the proposed Technical Specification amendment is not contingent ors continued use of the anticipatory reactor trip on turbine trip.

In the analysis performed to support the District's submittal, the reactor coolant portion of the OTSGs was modeled by including the appropriate reactor coolant flow rate, inlet temperature and pressure for 1121 power as boundary conditions. These inlet conditions were held constant through the transient. This assumption is j% tified by the fact that  ;

the peak pressure in the OTSGs, whi.rh accurs in about 4 to 6 seconds, is reached before any tempetture feedback due to the upset can affect these inlet conditions. (The RCS loop transport time is about 12 seconds.) RCS pressure results are not available from that work; however, results from later work done on a similar plant show that the RCS overpressure design limit will not be exceeded during the turbine trip event from overpower. These later analyses were performed with assumptions consistent with BAH-10043 assumptions. u) The .

plots of RCS pressure from this analysis are provided in l conjunction with District Response 1. As previously discussed, the maximum RCS pressure from that evaluation

.-______A

e <

• _ , . Atlachment

. AGM/NPP 89-040 Page 10 of 10 varied between 2681 and 2678 psia for the two cases' involving three and then two MSSVs out of service.. All-functioning MSSVs were set:to begin opening 1% above their lift setpoint of -

1100 psig and did not reach their full open relief capacity until 3% above this setpoint. These results show that the peak'.

RCS pressure reached following a turbine trip is somewhat insensitive to small changes in the MSSV relief capacity and that the peak RCS pressure limit of 2765 psia will not be exceeded. These results should bound the Rancho Seco plant since Rancho Seco will continue to use a staggered set of MSSV setpoints.

_ REFERENCES

1. NRC Letter J. F. Stolz to J. E. Hard, " Request for Additional Information Related to the Technical Specification Change Request Regarding Main Steam Safety Valves," Docket No. 50-312, dated February 7, 1987.

2. B&W Document No. 51-1171572-00, " Maximum RC Pressure Data: Peak Pressure Events," dated April 13, 1988.

3. B&W Document No. 86-1153322-00, "SMUD MSSV Analyses," dated November 13, 1984.

4. B&W Document No. 32-1170334-00, "MSSV Analysis," dated 10/27/87.

l l S. B&W Document No. 47-1173004-00, "B&H Owners Group Valve Task Force Final Report," dated August 1988.

l

)

6. _ BAH-10128,. Topical Report: TRAP 2 - FORTRAN Program for Digital Simulation ]

of the Transient Behavior of the Once-Through Steam Generator and l Associated Reactor Coolant System, dated Ap;ast 1976. {

p i L - -