Centre for Offshore Foundation Systems

Risk and reliability analysis

Further information

Search the UWA Research Repository

COFS has positioned itself at the forefront of developing the analytical models and methodologies required to accurately predict the risk of failure of offshore foundations, pipelines and platform systems.

While traditional assessment methods separate soil systems, currents, wave loading and other risks, these factors need to be treated as a whole system to accurately predict stability and safety. Our approach bridges the disciplines of geotechnical engineering, structural analysis and wave mechanics by developing numerical models in a framework that allows an integrated assessment of offshore structures and pipelines. Our ultimate aim is to establish a practical framework for assessing the stability, serviceability and fatigue life of the structural and riser systems that will underpin tomorrow’s marine energy developments.

risk and reliability renewables deep water engineering

Research Highlight: Jack-up failure

Jack-up platforms are installed by jacking the hull up off the water, while large ~20 m diameter spudcans are pushed into the seabed. This installation is challenging because of the prevalence of buried sand layers and highly stratified seabeds.

Potential exists for unexpected punch through failure, where a spudcan uncontrollably pushes a locally strong zone of soil into an underlying softer material. Such failures can lead to leg-buckling and in extreme cases can even topple the platform.

As many as five jack-up failures are recoded annually, mainly due to rapid leg penetration through seabeds of inconsistent strength.

At COFS, we have developed new analytical solutions for spudcan punch-through in two-layered and multi-layered soils. These allow a prediction of the largest vertical load before punch-through occurs.

COFS' Director, Prof Mark Cassidy together with other COFS' researchers, and in collaboration with Dr Marco Uzielli of GeoRisk, Italy, are defining a method to provide estimates of the probability of punch-through using a Bayesian framework. These predictions can be made prior to the jack-up installation and account for uncertainties within soil profiles and allows the models that we develop to be consistently adapting to offshore design and practice.


This new model provides engineers with a quantitative framework for evaluating possible failures before the jack-up arrives at site. More importantly, the probabilistic contour predictions can be updated moment by moment during installation as more data is gathered.

This research improves safety by developing more reliable models for the installation, operation and retrieval of jackups and provides guidance to the inclusion of monitored data in real-time offshore decision making.