ML19289D823
| ML19289D823 | |
| Person / Time | |
|---|---|
| Site: | Limerick, Shoreham File:Long Island Lighting Company icon.png |
| Issue date: | 01/11/1979 |
| From: | Wu T CALIFORNIA INSTITUTE OF TECHNOLOGY, PASADENA, CA |
| To: | Bates A Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-CT-1084, NUDOCS 7903140561 | |
| Download: ML19289D823 (1) | |
Text
g'.
C A LIFO R NI A INSTITUTE OF TECHNOLOGY PASADENA. C ALIF O R N e A 9112 5 T-auary 11, 1979
= =, ~
.c. ~ c. o.. A - r.,
C2-19W
- PDAL % zh?
Dr. Andrew L. Bates Advisory Committee on Reactor Safeguards U. S. Nuclear Regulatory Commission Washington. D. C.
20555
Dear Dr. Bates:
Please find enclosed two (2) copies of my written report on " Mark II Containment Generic Acceptance Criteria ACRS Fluid Dynamics Subcommittee Meeting - Nov. 28-30, 1978".
It was a great pleasure seeing you in San Francisco during the meeting.
My best regards.
Sincerely,
(' 6d,'f' L
Theodore Y. Wu Consultant for ACRS TYW:hb Enclosures 790314 056/
January 11, 1979 To:
Dr. M. S. Plesset, ACRS From:
T. Y. Wu, Consultant
Subject:
Report on Mark II Containment Generic Acceptance Criteria ACRS Fluid Dynamics Subcommittee Meeting - Nov. 28-30, 1978 San Francisco, CA.
As a result of my attending the meeting, where the Mark II Containment Generic Acceptance Criteria were presented and discussed, and my review of the relevant technical reports and documents, the following comments and recommendations are addressed to several aspects of the various dynamic loads and the NRC Acceptance Criteria.
I.
Methods for combining loads.
In view of the complexity and interactive features of the loading mecha-nisms in the Mark II system that may result from an OBE ground acceleration and a postulated LOCA, it is quite appropriate to adopt a statistical approach in preference to a deterministic one for load combination in the evaluation of resultant response of the system during these events. The present state of knowledge clearly shows that in design and evaluation processes there still exist many uncertainties s aich for all practical purposes can best be accounted for by imposing a desired degree of conservatism on the method of evaluation.
The proposed SRSS algorithm is based on three simplifying assumptions:
1.
the response is a result of statistically uncorrelated stochastic processes, 2 the response has zero mean, and 3.
the model is on linear clastic basis.
This method appears to be well based on sound judgment and extensive experience acquired from carthquake engineering applications rather than on purely theoretical predictions. With such a provisional conservatism as is now incorporated in covering the uncertainties, the chosen OBE ground motion has a low NEP, and the same holds for SRV's actuated concurrence with OBE.
In conclusion, the chosen conservatism seems adequate and no ad-ditional margin is necessary for the evaluation of combined response by SRSS.
Further comments:
1.
The acceptance criteria for SRSS can however be more firmly established if they can be verified on sufficient data basis since the Mark II system is substantially different in structure from conventional buildings.
2 The limitation to short duration - 10 sec. or less - of EQ ground motions seems rather restricted to cover both realistic occurrence and verification of the claimed NEP based on typical EQ energy spectrum (generally spanning over the range between 0. I to 10 Hz).
3.
Modification of SRSS to cover response time functions having non-zero mean is desirable to satisfy possible need in the future.
II.
LOCA-related hydrodynamic loads.
1.
Jet load.
The theoretical model adapted from conventional method (assuming the water jet velocity to equal the maximum vent clearing velocity) to evalu-ating the jet load seems to be overly conservative. The vertical head of an advancing jet, in a form somewhat resembling Hill's spherical vortex, contains nearly all the entrained vorticity and advances at a rate considerably lower than the vent clearing velocity. Further, the unsteady aspects of vent c1 caring and the wall effects due to the proximity of basemat can both modify the jet velocity. The high generic value of 33 psi overpressure on the con-taining walls versus the experimental value of 10 psi could be ascribed to these rather crude simplifications. The overly conservative estimate of jet load could be improved with further refinement of the theory, and this should be investigated in LTP studies and in the new Mark III project.
2 Pool-swell load.
In current practice, only the mass and momentum (initially downward at the downcomer exit) conservations are taken into account, Icaving the enr i gy arpect (especially the heat energy brought into the air bubble with the irr. pinging air) out of the consideration. This is a significant over simplifica-tion since it tends to make the estimate of pool-swellload on the diaphragm less than conservative. The under-estimate of the pool elevation observed in some cases most likely can be attributed to this crude model, and an improved method is needed to correct the poor prediction.
3.
Drag load.
The present method for evaluating the unsteady drag forces on sub-merged structures by potential flow theory appears to be conservative since it is known that the ' induced mass' determined on potential-flow theory is greater than that of the same body in a viscous flow with a wake formation.
Based on this argument, the total coefficient drag, C ',
a structure sub-D rnerged in an unsteady flow is expected to lie in a range from 1 to about 2 times the steady drag coefficient, C The proposed estimated of C
'=3C D.
D D
is therefore conservative for computing the peak load.
4.
Impact load.
The method currently used for calculating the impact load on structures during the inpinging of pool surface on them can be improved by using the same type of water impact theory as used by Plesset. The compressibility effects of water on impact load is too rapidly transient to be of any importance in most practical cases, so its omission is justified. This calculation should be further augmented by adopting the more accurate pool elevation data than previously used.
5.
Asymmetric oscillatory load.
The result of asymmetric loads of observed magnitude as high as 6 times the typical uniform load (covering the range of + 4. 5 psi at 20-3 0 Hz) shows that it has its own structure mechanism uncorrelated to the uniform load. This seems to be little understood and should be further investigated in the follow-up program.
- .