Sensitive and Uncertainty Analysis Shell (SUNS)

Introduction:

SUNS is an acronym for Sensitivity and Uncertainty Analysis Shell.  SUNS for Windows is a 32-bit application that runs under Windows 95/98 and Windows NT.  It is designed to aid in statistical analyses for a broad range of applications.  The class of problems for which SUNS is suitable is generally defined by two requirements:

1.A computer code is developed or acquired that models some process for which input is uncertain and the user is interested in statistical analysis of the output of that code.

2.The statistical analysis of interest can be accomplished using Monte Carlo analysis.

SUNS can generate a statistical sample using either Latin Hypercube or random sampling and, if desired, induce correlations among sampled variables.  Once the user's application analyzes the sample and provides output results in the required file format, the SUNS Results Viewer provides statistical analysis and graphical outputs of the results.

Benefits:

SUNS provides a powerful sensitivity and uncertainty analysis capability that can be coupled to almost any user's process model with relatively little effort.  Coupling between SUNS and the user's application is done via a comma-separated text file.  As a result, the only modifications to the user's application are those necessary to enable it to read this file and to provide its output in similar form.

Capabilities:

Uncertainty and sensitivity analysis capabilities provided by SUNS include:

• Latin Hypercube (LHS) or random sampling,
• The option to induce correlations among input variables,
• Identification of input variables that most affect output results, and
• Several graphical outputs including histograms, cumulative distribution functions, complementary cumulative distribution functions, paretos, and scatter plots.

Figure 1. Simple Rank Correlations between Input Variables and a Result (MTNF).

Figure 2. Distribution of Calculated Results (MTBF) and Scatter Plot
(Hard Drive Failure Rate vs MTBF).

Applications:

SUNS has been applied to a wide range of problems such as:

• Reliability analysis of repairable systems,
• Hydrocarbon economic analysis,
• Groundwater contamination and contaminant transport, and
• Radioactive waste disposal in geologic formations.

References:

SUNS User's Reference Manual, Sandia National Laboratories, December 1998. SUNS is currently being beta tested.