ML19329C249
| ML19329C249 | |
| Person / Time | |
|---|---|
| Site: | Davis Besse  |
| Issue date: | 04/06/1977 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML19329C246 | List: |
| References | |
| NUDOCS 8002120965 | |
| Download: ML19329C249 (15) | |
Text
_ - _
DRAFT ECCS SAFETY EVALUATION REPORT SUPPLEMENT DAVIS BESSE UNIT NO. 1 1.0 Introduction In Section 6.3.3 of the FSAR, the acolicant (Toledo Edison Cencany) incoroorated by reference B&W topical reports BAW-10104 and BAW-10105 (References 1 & 2, respectively) into its application to operate Davis Besse Unit No. 1.
Pursuant to the requirements of 10 CFR 50.46, B&W submitted these reports to demonstrate compliance with the ECCS Acceptance Criteria for its 177 fuel assembly plants with raised loops. The basis for acceptance of the principal cortions of the B&W evaluation rodel were set forth in the staff's Status Report of October 1974 (P.eference 3) and the Supolement to the Status Report of November 1974 (Peference 4). Tocether, the Status Report and its Supolement describe the B&W ECCS evaluation model and the basis for the staff's previous acceptance of the model. BAW-10104 des-cribes the general features of the B&W ECCS evaluation model and re-flects the modifications previously required by the staff (References 5 and 6). The original ECCS calculations acolicable to Davis-Besse 1 were submitted in BAW-10105 (Reference 2) using the B&W evaluation model described in BAW-10104 (Reference 1). Later developments on the i
)
I 8002120 966
~
validity of these calculations determined the following:
1.
The B&W method for calculating fuel cladding temperatures during the blowdown phase of a LOCA did not conform to Appendix K because it allowed for a return to nucleate boiling after critical heat flux conditions have been reached.
2.
A steam cooling model was used in the Davis Besse 1 ECCS calcula-tions whicn had not been reviewed by the staff.
3.
Irrproper pin pressure assumptions were employed.
4.
Incorrect value: of certain loop resistances were used.
With regard to item 1 above, Reference 20 provides tne staff evaluation of a revised nucleate boiling lockout logic proposed by B&W. The staff concludes that the revised logic is an appropriate change to be incorporated in the B&W Evaluation Model and that the overall effect 0
on the change on peak clad temperature would be small (s6 F).
With regard to item 2 above, the staff has concluded that the steam cooling model used by S&W is acceptable.* Items 3 and 4 relate to input errors and are discussed in more detail in Section 2.0 of this Safety Evaluation Report Supplement.
Other model changes have taken place subsequent to the ECCS calcu-lations in BAW-10105 (References 23 and 24). These changes have been accepted by the staff and their cumulative effect is not significant to the peak clad temperature.
- See coment in cover letter. The staff steam cooling model safety evaluation should be referenced here if published in time.
x -
2.0 ECCS Analyses t
The background of the staff's review of the revised B&W ECCS evaluation model and its application to Davis Besse Unit No.1 is described in Reference 5.
The applicant's FSAR contains documentation by reference to BAW-10105 of a generic break spectrum appropriate to Davis Besse Unit No. 1.
It is the staff's understanding that the responses to questions submitted on BAW-10105 (References 18, 19) will be made a part of the topical report by B&W. A spectrum of break sizes, configurations, and locations were performed. These 2
analyses identified the worst break as the 8.55 ft double-ended break at the pump discharge. B&W responses to staff inquiries during its review of BAW-10105 determined that incorrect internal fuel pin oressures had been assuned in the ECCS calculations.
B&W subsequently resubmitted analyses in Peference 18 with the corrected pin oressures. These revised analyses also included consideration of an additional flow resistance in the cold legs to accounc for HPI punos injecting ECC water durino reflood. The table below summarizes the results of the revised LOCA limit analyses which determine the allowable linear heat generation rate limits as a function of elevation in the core:
4 3
Peak Claddin Maximum Local Temperature (g )
0xidation (%)
Elevation LHGR Limit F
(ft)
(Kw/ft) 2 16.5 2133 4.01 4
17.2 2n73 3.15 6
18.4 2166 5.25 8
17.5 2164 6.56 10 17.0 2194 7.17 Subsequent to this review, Toledo Edison Coreany informed the staff that an erroneous resistance value to the reactor vessel inlet nozzle had been used in the loss-of-coolant accident (LOCA) analysis. As a result, the applicant submitted a re-evaluation of the Davis Besse In'+ No.1 LOCA analysis based on the corrected inlet nozzle model ar.3 a revised system pressure distribution.(References 21 and 22)
These results show that lower peak cladding temperatures would be obtained for the worst break analysis. The peak cladding temperature obtainea for the retyaluation of the 6 afoot LOCA limit analysis is 2133 F, a value 33 F lower than obtained in BAW-10105 (see tabulation on preceding page).
a reduction in peak cladding temperatures compared to The reason for those reported in BAW-10105 was due to improved reflooding rates in the core following a LOCA. The increased core reflooding rates are based on an improved system pressure distribution (i.e. the new reactor ceolant system total pressure drop is less than the original
~
5-
^
assumed pressure drop). We have reviewed the proposed pressure drops, the derivation of the revised system oressure' distribution and its impact on the LOCA limit analysis, and agree with the apolicant that the proposed linear heat generation limits as a function of core elevation are in compliance with the criteria of 10 CFR 50.46. Also, we conclude that the reevaluation of the 6-foot LOCA limit is sufficient to determine the effect of the revised system pressure distribution on peak cladding temperature for the range of axial power distributions previously analyzed. As reported earlier, the peak cladding temperature following a LOCA was reduced when analyzed at the 6-foot elevation of Similar effects would be expected at other elevations of the core.
Although the reevaluated results are less severe than those the core.
reported in BAW-10105, the applicant will maintain the allowable linear heat generation rate limits for Davis Besse Unit No. I at the same values as previously identified in this report. Additionally, because of the changes described in references 20 through 24, the staff requires that Toledo Edison Company submit within 6 months additional analyses to further quantify existing maigins. The additional analyses should, as a minimum, confirm previous evaluations with regard to worst break size, configuration, and allowable linear heat generation limits as a function of elevation in the core.
Also, we will require the applicant to provide operating reactor coolant system flow data for the Davis Besse Unit No. I which can be used to further verify the assumed total system pressure drom The new pressure drops were based on standard calculation methods supported by operating plant pressure data and the results from scaled reactor vessel model flow tests. B&W has shown that although there are some
6-m design differences between Davis Besse Unit No. 1 and other B&W plants from which measured data were obtained, these differences have a negligible effect on total system pressure drop. We have reviewed the flow path resistances input to the REFLOOD ECCS evaluation code for the revised system pressure distribution, and have checked several flow paths resistance values. We find the methods to be acoropriate for the derivation of loop resistances and accept the reported values as being appropriate for Davis Besse Unit No. 1.
Therefore, the staff concludes that the previous values quoted in the tabulation remain aoolicable to Davis-Besse 1.
The maximum core-wide metal-water reaction was c;1culated to be 0.66 percent, a value which is below the allowable limit of I cercent. As shown in the tabulation, the calculated values for ceak clad tercerature and local metal-water reaction were below the allowable limits specified in 10 CFR 50.46 of 2200'F and 17 percent, respectively. BAW-1010E has also shown that the core peemetry remains amenable to coolino and that long-tem core cooling can be established.
Our review of other plant-specific assumptions discussed in the following paragraohs regarding the Davis-Besse 1 analyses addressed the areas of single failure criterion, lono-term boron concentration, ootential submerged equipment, partial loop operation, and the containment pressure calculation.
_ s 2.1 Sincie Failure Criterion Appendix K to 10 CFR 50 of the Commission's regulations requires that the combination of ECCS subsystems to be assumed operative shall be those available after the most damaging single failure of ECCS equioment has occurred. Babcock and Wilcox has conserva-tively assuned all containment cooling systems coerating to minimize containment pressure and has independently assumed the loss of one diesel to minimize ECCS coolina. We stated in Reference 3 that the aoolication of the single failure criterion was to be confirmed during subsecuent plant reviews.
The acolicant has concluded that no sinale active failure would more severely decrade ECCS than the previous assumotions stated above. A review of the Davis-Besse i pioing and instrumentation diagrans and ECCS motor-onerated valve electrical schematics were conducted by the staff.
From these reviews the staff required valve changes in the LPI discharge lines, LPI-HPI crossover lines, and HPI mini-flow bypass lines. On the basis of the revised plant design, the staff concludes that a bounding single failure analysis has been perforned for the Davis-Besse Unit.10. 1 plant.
+
2.2 Containment Pressure The ECCS containment pressure calculations for 177-FA raised 1000 plants were perforr.ed generically by B&W as described in Reference 2.
The ilRC staff reviewed B&W's evaluation model and published the results of this review in References 3 and 4.
We concluded that B&W's containment pressure model was acceptable for ECCS evaluations. We required that justification of the olant-dependent input parameters used in the containment analyses be submitted for our review of each plant.
Justification for the containment inout data was submitted for Davis-Besse 1 on September 5,1975 (Reference 8). This just-ification allows comparison of the actual containment carameters for Davis-Besse 1 with those assumed in Reference 2.
Toledo Edison Company has ev aluated the containment net-free volume, the passive heat sinks, and coeration of the containment heat-removal systems with regard to the conservatism for the ECCS analysis. This evaluation was based on as-built design information.
The containment heat removal systems were assumed to coerate at their maximum capacities, and lowest expected values for the soray water and service water temoeratures were assumed. The containment pressure analysis in BAW-10105 was demonstrated to be conservative for Davis-Besse 1.
-9 m
We have concluded that the olant-decendent information used for the ECCS containment pressure analysis for Davis-Besse 1 is conservative and, therefore, the calculated containment oressures are in accordance with Appendix K to 10 CFR 50 of the Commission's regulations.
2.3 Lono-Term Boron Concentration The NRC staff has reviewed the oronosed crocedures and the systems desianed for oreventina excessive boric acid buildun in the reactor vessel durina the lona-tern coolino ceriod after a LOCA.
Toledo Edison Comoany has aareed to imolement nrocedures for Davis-Besse 1 which would allcw adeouate baron dilution durina the lono tern and which will coroly with the sinale failure criterion. These orocedures will employ a hot lea drain and hot leg injection network similar to the concept described in BAW-10105. The hot leg drain mode will direct reactor coolant from the hot leg, down the decay heat line to the DHD oump suction.
This coolant draining from the hot len would then be mixed with the dilute water being purned from the containment sumo and would then be curred back to the reactor vessel.
Should a single active component failure not allow operation of the hot lea drain mode, the ocerator then has the alternative of selectino the hot lea injection mode to provide boron dilution. The nrocedure would be to use the relatively dilute water being oumped out of the containment
- s sumo during the long-term recirculation mode and route a minimum of 40 gpm of this sump water to the hot leg to orovide dilution of the water in the upper plenum of the reactor vessel. The aonlicant will be required to demonstrate this minimum flow rate in each mode during preoperational testing.
In addition, the applicant must install flow rate measurina devices to assure that a minimum of 40 opm is continually available followina a LOCA, and to facilitate system tests. With the addition of the flow devices and preocerational tests, this crocosal is acceotable to the staff.
2.4 Submerced t!alves The acolicant has conducted a review o' equinment arrancement to determine if any components inside the containment will becore submerced following a LOCA. Based on this review, decay heat suction valves DH-11 & OH-12 were identified as beino located in an area that will be flooded. The apolicant subsequently enclosed these valves in a water-ticht comrartment to ensure their operability during the long term after a LOCA. The staff will require that an acceptable leakage test of this enclosure be performed each refueling outage. Simple visual inspection would not be sufficient.
~II" 2.5 Partial Looo Analyses To supoort an operating configuration with less than four reactor coolant pumos on the line (partial loco), the staff requires an analysis of the predicted consequences of a LOCA occurring during the proposed cartial loop operatina mode (s). Toledo Edison Company submitted an analysis for oartial loon coeration with one idle reactor coolant oump (three oumos operating) in Reference 9.
Using a reduced oower level of 77% of rated power, B&W performed this analysis assumino the worst case break (8.55 t't DE, CD = 1) and maximum LHGR allowed by Technical 2
Baseo on a sensitivity Specifications for this mode of operation.
study referred to by the applicant in Reference 14, the break selected was located in the active leg of the cartially idle loop. Placino the break at the discharge of the oumo in an active cold leg of the partially idle 1000 (instead of at the dis-charce of the cumn in an active cold leg of the fully active 1000) yields the most dearaded positive flow throuch the core durina the first half of the blowdown and results in higher claddina The maximum cladding temoerature for the one-idle-temoeratures.
pump mode of operation was 1675'F, a value which is within the criterion of 10 CFR 50.46. Therefore, this analysis may be used to suoport the apolicant's proposed ooeration with one idle reactor coolant Dumo.
Since an analysis of ECCS cooling performance with one idle reactor coolant pump in each 1000 has not been submitted, power coeration in this configuration must be limited by Technical Specifications to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.
Single 1000 operation (i.e., operation with two idle puros in one loop) is prohibited by current Technical Specifications without notifyino the staff.
Each croposal for a scheduled single loop test will be considered on a case-by-case basis.
e
3.0 Conclusions The staff has completed its review of the Davis-Besse Unit No.1 ECCS oerformance analyses and has concluded:
a.
The ECCS minimum containment pressure calculations were performed in accordance with Apoendix K to 10 CFR 50.
b.
With the modifications described herein, the sinale failure criterion will be satisfied.
The orocosed orocedures for lono-term coolino after a LOCA c.
are acceptable to the staff. The imolementation of these procedures before startuo is required to provide assurance that the ECCS can be operated in a manner which would prevent excessive boric acid concentration from occurrino.
d.
The proposed mode of reactor coeration with one idle reactor coolant purp is suoported by a LOCA analysis. Coeration with one idle ours in each loop is restricted to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.
Requests for sinole loop oceration will be reviewed on a case-by-case basis, e.
Additional analyses are requirec within six months to further quantify existing nargins.
s
References 1.
B. M. Dunn, et al., "B&W's ECCS Evaluation Model," BAW-10104, Babcock and Wilcox, May 1975.
2.
W. L. Bloomfield, et al., "ECCS Evaluation of B&W's 177-FA Raised-Looo NSS," BAW-10105, Babcock and Wilcox, June 1973.
3.
" Status Report by the Directorate of Licensino in the Matter of Babcock & Wil:ox ECCS Evaluation Model Conformance to 10 CFR 50, Accendix K," dated October 1974 4.
" Supplement 1 to the Status Report to the Directnrate of Licensin9 in the Matter of Babcock and Wilcox ECCS Evaluation Model Conformance to 10 CFR 50, Accendix K," dated November 13, 1974 5.
Lt :ter from A. Schwencer to Mr. Kenneth E. Suhrke, dated January 8,1976.
6.
Letter from John F. Stolz to Mr. Kenneth E. Suhrke, dated January 16, 1976.
7.
"Mir '.um Requirements for ECCS Break Spectrum Submittals," dated April 25,1975.
8.
Letter from Lowell E. Roe to Mr. A. Schwencer, dated September 5,1975.
9.
Letter from Lowell E. Roe to Mr. A. Schwencer, dated October 8,1975.
10.
Letter from Kenneth E. Suhrke to Mr. A. Schwencer, dated December 15, 1975.
11.
Letter from Kenneth E. Suhrke to Mr. John F. Stolz, dated April 6,1976.
12.
Letter from Lowell E. Roe to Mr. A. Schwencer, dated July 9,1975.
13.
Letter from Lowell E. Roe to Mr. A. Schwencer, dated July 21, 1975.
15-14.
Letter from Lowell E. Roe to Mr. Benard C. Rusche, dated May 17, 1976.
15.
Letter from Lowell E. Roe to Mr. Benard C. Rusche, dated June 4, 1976.
16.
Letter from Lowell E. Roe to Mr. Benard C. Rusche, dated November 8, 1976.
17.
Letter from Lowell E. Roe to Mr. John F. Stolz, dated January 3,1977.
18.
Letter from Kenneth E. Suhrke to Mr. John F. Stolz, dated April 6,1976.
19.
Letter from Kenneth E. Suhrke to Mr. John F. Stolz, dated June 8, 1976.
20.
Letter from Steven A. Varca to Mr. Kenneth E. Suhrke, dated February 18, 1977.
21.
Letter fron Lowell E. Poe to "r. John F. Stolz, dated February 8,1977.
22.
Letter from Lowell E. Roe to fir. John F. Stolz, dated April 1, 1977.
j from Kenneth E. Suhrke to Mr. D. B. Vassallo, dated August 20, 23.
Lettee
- 1976, 24.
Letter from Kenneth E. Suhrke to Mr. Denwood F, Ross, Jr., dated 1
June 7, 1976.