Incorporating Uncertainty and Variability in Engineering Analysis

Abstract

Engineering education has always been strong in providing stu- dents with the tools needed to solve problems. As civil engineers, we learned about methods for calculating runoff and solving for flows and pressures in pipe networks and applying various opti- mization techniques. The basic concept is to understand what data are needed, where or how that information can be acquired, how to choose and apply the appropriate methodology, and then how to use the results from that analysis to design or analyze a system. The underlying assumption is that if you use the correct data and select the proper methodology then the resulting answers will be the correct solutions. However, to a lesser degree engineering students are also exposed to the concept that engineering is not quite as black and white, or more appropriately, deterministic, as may be suggested by strict application of formulas. There are shades of gray that must be considered in the form of variability and uncertainty that are factors in all engineering and other decisions. It is this topic that I will discuss here both in general and as it applies to the subject of this special issue of the Journal—water distribution system analysis.