Reliability Assessment of Deepwater Suction Caissons

Abstract

Abstract Currently API provides two sets of guidelines for the design of offshore piles in compression. The first is a working stress design approach (WSD), while the second and more recent is a reliability-based approach that uses load and resistance factor design (LRFD). In recent years, as the offshore industry has moved into deeper water, suction caisson foundations have evolved as alternatives to piles. Presently only the WSD design approach is available for design of deepwater suction caissons. This paper presents some preliminary estimates of the reliability (probability of failure) of suction caisson foundations for deepwater applications. It is an initial step in an overall program intended to provide a similar reliability-based design approach as that currently existing for piles. This paper examines several different aspects of foundation reliability for mooring systems applicable for a range of deepwater structures. The potential impact of spatial soil variability on the overall foundation reliability also is examined. The preliminary results from this study suggest that a factor of safety of about 2.0 against the 100-year design load produces comparable levels of reliability as those currently obtained for compression piles using API design procedures. Although the results from this study are preliminary, they help to focus future work on key uncertainties and provide significant insights into potential foundation performance. Introduction As the offshore industry has moved into deeper waters, several aspects of foundation design have changed. One primary difference is that foundations are no longer loaded primarily in compression. For structural concepts such as Tension Leg Platforms (TLPs), Spars, and monohulls, the primary foundation loads have changed from compression to either uplift or combined uplift and lateral loading. In addition, suction caisson technology has been evolving slowly, initially starting with North Sea applications such as Snorre and Heidrun and then being used in the Gulf of Mexico (GoM) for projects such as the Hoover-Diana floating system. The fundamental mechanisms for providing foundation resistance with a suction caisson are different from the mechanisms provided by piles. Also, because of the high costs associated with extending shallow water site investigation techniques to deepwater, new methods for collecting soils data have been proposed. Some of the fundamental issues associated with site investigations have changed for deepwater because the anchorage foundations for Spars and monohulls are spread over much greater distances (kilometers) instead of, as is typically the case for shallow water foundations, a hundred or so meters. Despite these fundamental changes, engineers still are faced with the challenge of selecting and designing foundations for deepwater applications that will have a comparable level of reliability as those designed for shallow water, at the lowest possible cost. To address this problem, a deepwater technology project was initiated at BP Amoco to use a reliability-based approach to:make consistent cost and reliability comparisons between piles and suction caissons, as well as establish a framework for assessing future foundation types such as vertically loaded anchors, andestablish and define the steps required to develop a robust reliability-based design approach (e.g., using Load and Resistance Factor Design) for deepwater foundations.