Role of Soil and Structural Heterogeneity in Geotechnical System Redundancy

It is well known that redundancy generally improves system reliability. For example, a geotechnical support system comprised of a single monopile will have the same failure probability as the monopile itself. Alternatively, if the geotechnical support system is comprised of two piles, each of which can support the load with probability 1−pf, then the system failure probability will lie somewhere between pf and p2f. In this case, if the failure probability of an individual pile is pf=1/100, then the system failure probability will lie between 1/100 and 1/10,000, depending on the degree of statistical dependence between the piles. Clearly, redundancy in the geotechnical system has the potential to significantly reduce the system failure probability. From a design point of view, since redundancy generally increases system reliability, the individual system elements (e.g., piles) need not necessarily be designed to the same level of reliability. In other words, if the supported load is distributed amongst a number of footings or piles, redundancy should be taken into account to achieve construction savings while maintaining overall safety. This paper looks specifically at the effects of redundancy in pile support systems on the overall system reliability. It is assumed that the support only fails when all piles have failed and piles fail randomly according to the local ground strength (pile structural capacity is not considered). Two load transfer models between failed (excessively displaced) and surviving piles are considered. Pile system reliability is then estimated as a function of the distribution of pile resistance, the load transfer model, the number of piles, and the target design reliability of individual piles. Charts are produced to allow the selection of individual target design reliability for a given number of piles and the target system reliability.