ML20214P503
| ML20214P503 | |
| Person / Time | |
|---|---|
| Site: | Sequoyah  |
| Issue date: | 11/25/1986 |
| From: | NRC |
| To: | |
| Shared Package | |
| ML20214P315 | List: |
| References | |
| NUDOCS 8612040239 | |
| Download: ML20214P503 (7) | |
Text
.
ENCLOSURE 1 j
SAFETY EVALUATION REPORT ENVIRONMENTAL EFFECTS OF MAIN STEAM LINE BREAK OUTSIDE CONTAINMENT SEQUOYAH NUCLEAR PLANT, UNITS 1 AND 2 DOCKET NOS: 50-327/328 I.
Introduction By letter dated August 13, 1986, the Tennessee Valley Authority (TVA), the licensee, submitted a report (Reference 1) entitled, " Main Steamline Break Environmental Qualification Study for TVA Sequoyah Units 1 and 2 Main Steam Valve Vaults," in response to IE Information Notice 84-90 which addresses the concern of Main Steam Line Breaks (MSLB) outside containment with superheated steam blowdown. TVA identified this issue in Volume II, section III.6.0, of the TVA Nuclear Performance Plan (NPP), which was submitted by letter dated July 17, 1986.
For certain MSLB accidents, steam generator tube bundle uncovery occurs which results in the production of superheated steam. The effect of the superheated steam is to raise the temperature in the main steam valve vaults above that previously calculated. Consequently, the environmental qualification of equipment located in the valve vaults needs to be reevaluated.
The valve vaults are located adjacent to the Sequoyah Units 1 and 2 containment buildings. Each unit has two valve vaults, i.e.,
East and West valve vaults.
TVA reanalyzed the qualification of equipment in both valve vaults.
II. Review and Evaluation The mass and energy release data from WCAP-10961 (Reference 2) were used as input to the Westinghouse computer code COMPACT for calculating the temperature profiles in the valve vaults. A thermal lag analysis was then performed to obtain the component temperature response.
h20 39 861125 P
CK 05000327 PDR
y A.
Mass and Energy Release Data The mass and energy release data used are presented in Westinghouse Topical Report WCAP-10961. The data for Sequoyah Units 1 and 2 are tabulated in
" Category 2" of the topical report, which was prepared under the auspices of the Westinghouse Owners Group - High Energy Line Break /Superheated Bl.ow-down Outside Containment subgroup program. The Westinghouse computer code LOFTRAN was used for this calculation. The LOFTRAN code was modified to account for heat transfer to the steam during steam generator tube bundle uncovery. This modification is described in Westinghouse Topical Report WCAP-8860, Supplement 1, which the staff found acceptable in its safety evaluation that was transmitted to Westinghouse by letter dated May 27, 1986 (Reference 3).
The licensee postulated a spectrum of breaks, including a double ended rupture of the steam line (1.4 fte), a 0.9 ft2 break upstream of the main steam line check valve, and a 0.9 ft2 break downstream of the main steam line check valve. The 1.4 ft2 break results in automatic isolation of the main steam isolation valves (MSIV) and the most rapid tube bundle uncovery, and, therefore, the earliest onset of superheat.
The 0.9 fta break up-stream of the check valve is similar to the 1.4 fte break except that the blowdown rate is lower and the duration of blowdown is icnger. Even though automatic isolation of the MSIVs does not occur, the check valve prevents the other three steam generators from blowing down. The 0.9 fta break down-stream of the check valve does not initiate MSIV closure, and, therefore, all four steam generators blow down. As a result, tube bundle uncovery occurs late in the transient. The total blowdown energy from the four steam generators is significantly higher than that from one steam generator.
The results of the analyses indicate that the 0.9 fte break downstream of the check valve is the limiting case.
B.
Compartment Temperature and Component Thermal Lag Analyses The Westinghouse code COMPACT was used for calculating compartment temper-ature profiles. The bouyancy force, due to temperature stratification and the density of the steam, is represented by the gravity term in the momentum
~1 equation. The licensee found that buoyancy initiates a natural circulation pattern that pulls cold outside air into the vault and pushes hot air out through the blowoff roof panel. Natural circulation significantly reduces the temperature in the vault. The natural circulation phenomenon and its effects, originally identified in the COMPACT code calculations, were later confirmed by a TVA calculation using the RELAP5 computer code, and were also confirmed by the staff's consultant, Battelle Pacific Northwest Labor-atory (PNL), using the COBREE computer code.
In calculating the valve vault temperature response, the concrete walls and steel structures were accounted for as heat sinks. Condensation heat transfer based on the Uchida correlation was modeled until the surface temperature reached the saturation temperature corresponding to the pressure in the vault. Afterwards, natural convective heat transfer was modeled.
For the components, different heat transfer coefficients were used to maximize the component surface temperature responses. Four times the Uchida correlation and forced convection heat transfer coefficients were used in modeling the condensing mode and saturation mode, respectively.
This approach is conservative and in accordance with the staff guidance in NUREG-0588, " Interim Staff Position on Environmental Qualification of Safety-Related Electrical Equipment," and therefore, is found acceptable by the staff.
C.
Results of the Analysis Six cases were analyzed by Westinghouse for the two valve vaults using the COMPACT computer code. The rapid blowdown of the steam generator for the 1.4 fte and 0.9 ft2 breaks upstream of the check valve cause natural cir-culation to occur early in the transient. Therefore, the coolino effect of natural circulation mitigates the temperature rise in the valve vaults. However, the 0.9 fte break downstream of the check valve results in all four steam generators blowing down and delays the natural circulation effect. This delay results in a higher vault temperature.
From the results presented in the submittal, it is found that the 0.9 ft2 break downstream
1 of the check valve in the West valve vault is the worst case.
For this case, the vault air temperature rises to 302 F from 140'F in the first 10 seconds after the break. Thereafter, the vault air temperature slowly rises to 323*F by 250 seconds. At this time the tube bundles start un-covering and the vault temperature increases to 430*F at about 510 seconds.
The vault temperature stays at about 430*F for 70 seconds. At 543 seconds, the mass ralease rates have dropped enough for natural circulation to begin.
Natural circulation and the termination of the blowdown at 600 seconds cause a rapid cooldown of the vault to temperatures below 200 F.
j A sensitivity study showed that the results are not sensitive to the nodalization model chosen for the valve vault. A blowoff roof flow area sensitivity study also showed that the compartment air temperature rise is only slightly sensitive to the flow area.
The resulting surface temperature profile for a MSIV is shown in Figure 6.3-5 of Reference 1; the peak temperature is 365*F. Similarly, the resulting surface temperature profiles of an ASCO solenoid valve and conduit are shown in Figure 6.3-11 and Figure 6.3-19 of Refer $nce 1, respectively, with peak temperature of about 380*F in both cases.
It is noted that these peak component surface temperatures are higher than the qualification temperature limit of 325*F.
D.
Confirmatory Analyses performed by TVA and PNL The analyses discussed above were performed by Westinghouse for TVA, using the COMPACT computer code. TVA performed an independent, confimatory analysis using the RELAPS computer code. The analysis results based on RELAPS are similar to those obtained using the COMPACT code, with respect to the shape of the temperature profiles and the phenomenon of natural circulation. The predicted timing of the temperature spike and the onset of natural circulation cooldown were in close agreement between the two calculations. The predicted peak temperature and steady state temperature values were also close, with RELAPS results being somewhat higher.
Using RELAP5, TVA ran additional cases assuming a smaller break size (0.3 fte) and different initial power levels (102% and 70%).
The effect of initial power on the vault temperature response was insignificant, and the
temperature response for the smaller break size was less severe. There-fore, TVA concluded that the spectrum of break sizes chosen in the Westing-house COM MT analysis was acceptable. The staff concurs with the licensee on the h quacy of the break spectrum analyzed.
The Battelle Pacific Northwest Laboratory (PNL), at the staff's request, performed an independent confirmatory analysis using the COBREE computer code. This code has previously been used for the calculation of compart-mental pressure / temperature response following a postulated high energy line break. The results of the PNL analysis show good agreement with the shape and timing of the temperature profiles obtained for the three cases analyzed in the Westinghouse COMPACT analysis, i.e., the 1.4 ft2 break, the 0.9 ft2 break upstream of the check valve and the 0.9 fte break downstream of the check valve, in the West valve vault. The PNL results confirm the effect of the natural circulation phenomenon identified in the licensee's analysis. Quantitatively, the COBREE calculations predicted higher room temperatures but lower component surface temperatures. One of the main reasons for this is the way in which the COBREE code models heat transfer.
The current version of the COBREE code uses the same heat transfer coefficient for structural heat sinks and safety related components. The COMPACT code, however, minimizes heat transfer to the structural heat sinks and maximizes the heat transfer to the safety related components.
This approach is more conservative for component surface temperature calculations, and is consistent with the guidance in NUREG-0588.
Therefore, the staff finds the component surface temperature profiles calculated with the COMPACT code to be acceptable for equipment qualification.
III. Conclusion The calculated component surface temperature profiles presented in Figures 6.3-5 (MSIV), 6.3-11 (ASCO valve) and 6.3-19 (cable in conduf t) of the licensee's submittal (Reference 1) are acceptable for equipment qualiff-cation. However, the peak temperatures exceed the qualification temper-ature limit of 325 F.
The licensee contends, however, that component in-ternal temperatures are actually lower than what internal parts have been tested to. The licensee's justification for this is still under staff review.
e IV.
References:
1.
Letter from R. Gridley (TVA) to B. Youngblood (NRC), dated August 13, 1986, transmitting the report, " Main Steamline Break Environmental Qualification Study for TVA Sequoyah Units 1 & 2 Main Steam Valve Vaults."
2.
WCAP-10961 Rev.1 (Proprietary) and WCAP-11184 (Non-Proprietary),
"Steamline Break Mass / Energy Releases for Equipment Qualification Outside Containment".
3.
Letter from C. E. Rossi (NRC) to E. P. Rahe Jr. (Westinghouse), dated May 27, 1986, subject: Acceptance for Reference of Licensing Topical Reports WCAP-8822-P-SI/WCAP-8860-S1 and WCAP-8822-P-S2/WCAP-8860-52,
" Mass and Energy Release following a Steam Line Rupture."
O DISTRIBUTION:
a Docket File NRC PDR Local PDR PRC System NSIC PWR44 Reading FDuncan BJYoungblood Reading TAlexion TYAOP (3) S. Richardson AR 5029 HCentan Jiaylor BHayes GZech, RII fl Grace LSoessard KBarr SAConnelly
~
Driuller TNevak BJYoungblood JHolonich CStable TKenyon WLong BKSingh KHooks ACRS (10)