Effect of sampling locations on reliability of pile groups

Geotechnical site investigations represent an imperative prerequisite in the pursuit of reliable foundation designs. However, site investigations are frequently restricted to a limited number of locations owing to the constraints imposed by budget and time considerations, thereby potentially yielding a range of adverse consequences. Hence, the development of an efficient site investigation plan — one that optimally selects the number and location of tests, is important to gain adequate information for a given cost. This paper proposes a framework to identify optimal investigation locations by minimizing the probability of erroneous decisions (i.e., error types I and II). A pile group was adopted for the demonstration. The best locations corresponding to various numbers of tests were identified based on the derived probabilities of type I and II errors.