ML19282B804
| ML19282B804 | |
| Person / Time | |
|---|---|
| Issue date: | 02/28/1979 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML19282B803 | List: |
| References | |
| NUDOCS 7903160206 | |
| Download: ML19282B804 (6) | |
Text
.
ENCLOSURE TOPICAL REPORT EVALUATION Report No. &
Title:
WCAP-8904, "Wes tinghouse Emergency Core Cooling System Evaluation Model for Analyzing (N-1) Loop Operation of Plants with Loop Isolation Valves."
Originating Organization:
Westinghouse Electric Corporation Reviewed By:
Analysis Branch Summary of the Topical Report Westinghouse designeo Nuclear Steam Supply Systems (NSSS), containing re-circulation loop isolation valves, can be operated with one recirculation loop out of service. These valves isolate the steam generator and reactor coolant pump from the primary system, thereby pemitting continued reactor operation at reduced power. Topical report WCAP-8904 documents how the computer programs, SATAN and WREFLOOD, will be used in modeling a loss of coolant accident for Westinghouse designed PWRs equipped with loop isola-tion valves, and operating under (N-1) loop configuration.
The topical report presents system sensitivity studies and discussions of modeling as-sumptions for a Westinghouse 4-loop (15x15) and a Westinghouse 3-loop (17x17) power plant.
Regulatory Evaluation LOCA analyses performed on plants operating under (N-1) loop configuration require no modification to the analytical models approved for the SATAN and WREFLOOD computer programs. Only minor noding changes are required to re-flect the modified primary system configuration.
79031602o6
The SATAN blowdown calculations for normal plant analyses (all loops in operation) model the primary system with 46 nodes. When modeling an active or an inactive loop break under (N-1) loop configuration, Westinghouse proposes to describe the primary system utilizing three additional nodes. One node models the valved off hot leg, the second node models the valved off cold leg, and the third node models the accumulator that is connected to the valved off cold leg.
Thus, when modeling a break in the inactive loop, Westinghouse proposed to model the broken leg as one node.
When modeling the cold legs, the one-dimensional SATAN code does not account for the momentum flux pressure gradient between the downcomer and the adjacent cold legs. The momentum flux is not calculated at this location because the local fluid conditions in the downcomer, adjacent to the cold leg pipe, are unknown.
This is due to the three dimensional behavior of the downcomer. Sensitivity studies conducted by Westinghouse showed that when modeling the break as two nodes (neglecting the momentum flux between the vessel and the cold leg, and considering the momentum flux between the two cold leg nodes), a slight increase in calculated peak clad temperature (13 deg. F) was observed. The use of two cold leg nodes to model the break is consistent with previous sensitivity studies (Ref. 3). Thus, Westinghouse agreed to model the broken leg using a minimum of two nodes, and accounting for the momentum flux between the cold leg nodes, but not between the vessel and the broken leg.
- The WREFLOOD code analyzes the reflood phenomena occurring during a postulated LOCA. When modeling an active loop break, Westinghouse uses the same noding structure as previously approved for (N) loop operation, with the addition of one node connected to the downccmer which models the inactive loop accumulator.
The dead-ended hot and cold leg volumes were not simulated.
This is acceptable since the dead-ended nodes have no influence on the reflood transient.
When modeling an inactive loop break, the WREFLOOD noding structure is modified to account for the dead-ended hot leg piping.
The dead-ended (val';ed shut) pipe was modeled as three consecutive nodes with decreasing areas and increasing loss coefficients.
This is acceptable since it adequately models the function of the dead-ended hot leg pipe.
Sensitivity studies have been parformed for postulated breaks in both the active and inactive loops.
In the cases analyzed, the peak clad temperature occurred for the postulated active loop break.
For the postulated inactive loop cold leg breaks, no steam venting from the dead-ended hot leg occurs, thereby increasing potential steam binding effects. However, this is offset by a prolonged blowdown negative core flow, which results from the single-ended cold leg break.
The prolonged negative core flow during the blowdown phase provides suffi-cient cooling to reduce the stored energy in the core, thereby reducing
- the amount of steam generated during the reflood transient.
The net effect results in a higher peak clad temperature for the active loop breaks.
To assure that this trend is consistent for all applications, (N-1) loop plant analyses for both an active and inactive loop break should be performed.
For a postulated cold leg break in the inactive loop, a potential con-cern exists regarding cold leg and downcomer plugging during the reflood transient. The downcomer mixture level model in WREFLOOD assumes that the level does not rise above the bottom of the cold legs. Consequently, the code could not properly assess potential plugging of the steam venting path in the cold legs and around the downcomer while the accum-ulators are injecting into a filled downcomer.
The plugging of this path has a potential for increasing steam binding and retarding the reflooding rate. Westinghouse conducted a series of sensitivity studies which artificially increased the system pressure losses such that the plugging behavior was conservatively bounded. The studies showed this concern to have a small (less than 10 F) influence on the calculated peak clad temperature. The sensitivity of the reflooding process to this effect was confirmed by independent staff calculations.
Regulatory Position The Nf'.C staff has completed its review of WCAP-8904 which describes the models used to evaluate LOCAs for Westinghouse designed PWRs with isalation valves segregating one loop from the primary system. We con-clude, with the stipulation of modeling the inactive loop break with a
. minimum of two nodes and proper accounting of the momentum flux effects, that the methods stipulated in WCAP-8904 are acceptable for ECCS per-formance evaluation for Westinghouse designed 3-loop (17x17) and 4-loop (15x15) PWRs with loop isolation valve.
In addition, specific applicants should verify that the limiting peak clad temperatures occur for postulated active loop failures. This report may be referenced in related licensing applications as an acceptable analytical model.
- References 1.
Letter from C. Eicheldinger, Manager, Nuclear Safety Department, Westinghouse Electric Corporation, to Mr. John F. Stolz, Chief, LWR Project, NRR, dated June 27, 1977.
2.
Letter from C. Eicheldinger, Manager, Nuclear Safety Department, Westinghouse Electric Corporation, to Mr. John F. Stolz, Chief, LWR Project, NRR, dated September 7,1977.
3.
WCAP-8341, " Westinghouse Emergency Core Cooling System Evaluation Model - Sensitivity Studies, July,1974.
-