Accounting for uncertainty and risk

The hydraulic engineer must always keep in mind the level of certainty inherent in data measurements, computational methods, and information provided by others. For example, the frequency of flows developed by a hydrologist is, at best, a statistical derivation with confidence limits. The hydraulic engineer can inspect the steepness of the frequency curve, as well as the confidence limits to determine the range of flows that should be associated with a given recurrence interval. If the hydrologist had no gage data from which to develop frequency information, the hydrology would probably be considered even less reliable.

As described in this course, numerous methods in hydraulic engineering were developed from empirical studies. The designer should know what situations are and are not applicable to a given methodology. Whenever simplified methods are employed, the designer should be aware of the sacrifice in confidence of results.

One typical response in the attempt to minimize the risk due to uncertainty is to use factors of safety and be conservative. However, it is critical that the designer apply factors of safety to the correct calculations and be conservative in the correct aspect of the analysis. To be conservative from a flood control perspective is to design a larger than necessary channel. However, if the goal of the project is to reconnect the flood plain, that designer’s conservatism may lead to design failure.

The designer should keep track of each computational attempt to account for uncertainty. In each case, an adjustment should be justified by a description of the source of the uncertainty and reasoning regarding the magnitude of the adjustment. In many cases, a conven-tional factor of safety will have been established by the field of hydraulic engineering. Standard freeboard heights for channel design are also conventional.

Finally, the hydraulic engineer may wish to more fully document the impact of uncertainties by modeling what-if scenarios, considering extreme values of one or more parameters.

Congratulations, you have completed the knowledge section of the course.

You may now complete the course by successfully passing the course quiz.