Evaluating reliability in geotechnical engineering - Risk and Reliability in Geotechnical Engineering - page 129

evaluating reliability in
geotechnical engineering
J. Michael Duncan and Matthew D. Sleep
3.1 PurPoSe oF relIabIlItY analYSIS
The factors of safety used in conventional geotechnical practice are based on experience,
which is logical. However, it is common to use the same factor of safety value for a given
type of application, such as F = 1.3 for short-term slope stability or F = 3.0 for erosion
and piping, without regard to the degree of uncertainty involved in the calculation in
a particular instance. Through regulation or tradition, the same safety factor value is
often applied to conditions that involve widely varying degrees of uncertainty. This is not
logical.
Reliability calculations provide a means of evaluating the combined effects of uncertain-
ties, and a means of distinguishing between conditions where uncertainties are particularly
high or low. In spite of the fact that it has potential value, the reliability theory has not been
used much in routine geotechnical practice. There are two reasons for this. First, the reliabil-
ity theory involves terms and concepts that are not familiar to many geotechnical engineers.
Second, it is commonly perceived that using the reliability theory would require more data,
time, and effort than are available in most circumstances.
“Reliability” as it is used in the reliability theory is the probability of an event occurring
or the probability of a “positive outcome.” Reliability is the complement of probability of
failure. Thus, if there is a 0.5% probability that the factor of safety could be less than 1.0,
the reliability (the probability that the factor of safety is greater than 1.0) is 99.5%.
Evaluating reliability affords a means of assessing the degree of uncertainty involved in
geotechnical engineering calculations. Christian et al. (1994), Tang et al. (1999), and oth-
ers have presented examples of reliability use in geotechnical engineering, and explained
the underlying theories. The purpose of this chapter is to show that reliability concepts
can be applied in simple ways, without more data, time, or effort than are commonly
available in geotechnical engineering practice. Working with the same quantity and types
of data, and the same types of engineering judgments that are used in conventional analy-
ses, it is possible to make useful evaluations of reliability. Although evaluations of reliabil-
ity are inevitably approximate, they provide a useful complement to conventional factors
of safet y.
The results of simple reliability analyses, of the type described in this chapter, will be nei-
ther more nor less accurate than conventional deterministic analyses that use the same types
of data, judgments, and approximations. While neither deterministic nor reliability analyses
are precise, they both have value, and each enhances the value of the other.
It is not advocated here that the factor of safety analyses be abandoned in favor of reli-
ability analyses. Instead, it is suggested that factor of safety and reliability be used together,
as complementary measures of acceptable design. The simple types of reliability analyses
131