Lawrence Livermore National Laboratory Probabilistic Seismic Hazard Codes Validation

Savy, J B

Lawrence Livermore National Laboratory

关键词: Validation;    Lawrence Livermore National Laboratory;    Testing;    54 Environmental Sciences;    Ground Motion;   

DOI  :  10.2172/15004907
RP-ID  :  UCRL-ID-155282
RP-ID  :  W-7405-ENG-48
RP-ID  :  15004907

美国|英语

来源: UNT Digital Library

【 摘 要 】

Probabilistic Seismic Hazard Analysis (PSHA) is a methodology that estimates the likelihood that various levels of earthquake-caused ground motion will be exceeded at a given location in a given future time-period. LLNL has been developing the methodology and codes in support of the Nuclear Regulatory Commission (NRC) needs for reviews of site licensing of nuclear power plants, since 1978. A number of existing computer codes have been validated and still can lead to ranges of hazard estimates in some cases. Until now, the seismic hazard community had not agreed on any specific method for evaluation of these codes. The Earthquake Engineering Research Institute (EERI) and the Pacific Engineering Earthquake Research (PEER) center organized an exercise in testing of existing codes with the aim of developing a series of standard tests that future developers could use to evaluate and calibrate their own codes. Seven code developers participated in the exercise, on a voluntary basis. Lawrence Livermore National laboratory participated with some support from the NRC. The final product of the study will include a series of criteria for judging of the validity of the results provided by a computer code. This EERI/PEER project was first planned to be completed by June of 2003. As the group neared completion of the tests, the managing team decided that new tests were necessary. As a result, the present report documents only the work performed to this point. It demonstrates that the computer codes developed by LLNL perform all calculations correctly and as intended. Differences exist between the results of the codes tested, that are attributed to a series of assumptions, on the parameters and models, that the developers had to make. The managing team is planning a new series of tests to help in reaching a consensus on these assumptions.