ML17263A003

From kanterella
Jump to navigation Jump to search
Research Information Letter5 0054, the Set Equation Transformation System
ML17263A003
Person / Time
Issue date: 05/15/1979
From: Levine S
Office of Nuclear Regulatory Research
To: Harold Denton, Dircks W
Office of Nuclear Material Safety and Safeguards, Office of Nuclear Reactor Regulation
References
RIL-0054
Download: ML17263A003 (3)


Text

  • UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D. C. 20555

.t511t MEMORANDUM FOR: Harold R. Denton, Director Office of Nuclear Reactor Regulation William J. Dircks, Director Office of Nuclear Material Safety and Safeguards FROM: Saul Levine, Director Office of Nuclear Regul a.tory Research -$, ~i ~

SUBJECT:

RESEARCH INFORMATION LETTER ff~ THE SET EQUATION TRANSFORMATION SYSTEM This memorandum describes work performed to improve the Set Equation Transformation System (SETS computer code) for fault tree evaluations*

of nuclear safety systems. The present version of the SETS code, developed by Sandia Laboratories,, represents a major advance over the fault tree codes used in the Reactor Safety Study (RSS), particularly for large and complex trees. Whereas the codes used in the RSS could only efficiently compute minimal cut sets (the basic combination of components which cause the system to fail) up to double or perhaps triple component failures, the SETS code can usually process the same size trees with no such limitations on the size of minimal cut sets produced. For example, the code has been successfully used to process a tree containing approximately 300 gates, 300 primary events, and over 64 million minimal cut sets. The ability to process large trees with no strict limitations on minimal cut set size is of particular importance when SETS is used for common cause searches of large cut sets having a common cause failure potential. In addition to serving as an efficient analysis tool, SETS has another important benefit in that its input and output are a standardized form of documentation for the analysis. The analysis can be repeated or modified for sensitivity studies, and a given analysis of a fault tree can be combined with other related analyses as future needs arise. SETS has been well tested and is an accepted code which is currently in use at Lawrence Livermore Laboratories, Los Alamos Scientific Laboratories, and Sandia Laboratories.

As a result of the work performed in the SETS computer code project, an improved version of the SETS code has been documented and made available for use by NRC contractors and personnel for projects requiring an efficient tool for the analysis of complex systems.

I The major results of the SETS p~ogram have been:

1. Development of an automatic tree decomposition algorithm (documented in the SETS user's manual, Attachment 1) which will further increase the speed and efficiency of SETS for processing large trees, and which is currently befog incorporated into the stan.dard *version of SETS.
2. Development, in preliminary form, of basic minimal cut set quantification procedures. These procedures, which are currently being incorporated into the standard version of SETS, form the beginnings* of a comprehensive quantification capability. * *
3. Development of a standardized version of the SETS code for the CDC 6600 computer, and installation of this version of the code at the Brookhaven National Laboratory Computer Center for use by NRC personnel. (A11 future updates, and enhancements to the standard version of SETS will also be made available to NRC.}
4. Preparation of a SETS users' manual oriented specifically for the fault tree analyst (s~e Attachment 1).
5. Presentation of two courses for NRC personnel and contractors on the use of SETS for fault tree analysis.

If you require further information or have any questions r_egardi_ng the SETS code, please contact Francine F. Goldberg {492-8388}.

~. b Office of Nuclear Regulatory Research

Attachment:

1. A SETS User's Manual for the Fault Tree Analyst

MEMORARDUM FOR: Harold R. Denton. Director Office of Nuclear R~actor Regulation

~ ' .

William J. Dircks. Director . . .

Office of Nuclear Material Safety and Safeguards FROM: Saul Lev1neD Dfrector

  • Office of Nuclear Regulatory Research.

SUBJECT:

RESEARCH INFORMATION LffiER f 54 TH£ SET EQUATIQN TRANSFORMATION SYSTEM This memorandum describes work performed to *improve the Set* Equation Transformation System (SETS computer codel for fault tree evaluations*

of nuclear safety systems. The present version of the SETS code. developed by Sandia laboratories. represents a major advance over the fault tree codes used in the Reactor Safety Study (RSS). particularly for large and complex trees. Whereas the codes used in the RSS could only efficiently compute minimal cut sets (the basic combination of components which cause the system to fafl) up to double or perhaps triple component fa1lures11 the SETS code can usually process the same size trees with no such limitations on the size of minimal cut sets produced.. For example. the code has been successfully used to process ~ tree contafn1ng approximately 300 gates. *,

300 primary events, and over 64 million minimal cut sets. The ability to process large trees with no strict limitations on minimal cut set size is of particular importance when SETS is used for colliilOO cause searches of large cut sets havfng a CORl!Jon cause failure potential. In addition to serving as an efficient analysis tool, SETS has another important benefit 1n that its fnput and output are a standardfzed form of documentation for the analysts. The analysis can be repeated or modified for sensitivity studies, and a gfven analysts of a fault tree can be combined with other related analyses as future needs arise. SETS has been well*** tested and is an accepted code whfch 1s currently in use at Lawrence Livennore Laboratories, Los Alamos Scientffic laboratories, and Sandta Laboratories.

As a result of the work performed in the SETS computer code project, an improved version of the SETS code has been documented and made available for use by NRC contractors and personnel for projects requiring an efficf ent tool for the analysis of complex systems.

I &..'°' , J

  • U . S . GOVERNMENT PRINTING OFFICE1 1978 - 828-824