85
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
1
McClelland lecture
Analytical contributions to offshore geotechnical engineering
Conférence McClelland
Contributions des méthodes analytiques à la géotechnique offshore
Randolph M. F.
Centre for Offshore Foundation Systems, University of Western Australia
ABSTRACT: The theme of this paper, the written version of the 2
nd
McClelland Lecture, is the contribution of analysis to offshore
geotechnical engineering. The application areas considered range from the axial and lateral response of piles, to seabed infrastructure
associated with deep water applications, including shallow skirted foundations, anchors, pipelines and risers. The emphasis
throughout is on analytical solutions, including appropriately framed outcomes of numerical studies. Most of the material is
retrospective, summarising key contributions in an effort to facilitate access, and thus help close the gap between theory and practice.
RÉSUMÉ : L’objet de cet article, la 2
e
conférence McClelland, est de présenter les contributions des méthodes analytiques à la
géotechnique offshore. Il couvre plusieurs champs d’application, de la capacité axiale et horizontale des pieux au comportement des
structures géotechniques associées aux développements en eaux profondes, incluant notamment les fondations superficielles avec
jupe, les systèmes d’ancrages et les pipelines. L’accent est notamment porté sur les solutions analytiques, dont certaines sont basées
sur des résultats de solutions numériques. L’essentiel du contenu de cet article résume les contributions antérieures les plus
significatives, de façon à en faciliter l’accès et ainsi réduire l’écart entre théorie et pratique.
KEYWORDS: Analysis, consolidation, offshore engineering, penetrometers, pile foundations, pipelines, shallow foundations.
1 INTRODUCTION
I was privileged to meet Bram McClelland on a few occasions
and have always held him in the highest regard. Much of my
early exposure to the offshore world was through interactions
with the London and Houston branches of the consulting
company, McClelland Engineers, that he founded. It was
therefore a great honour to be invited to give this, the 2
nd
,
McClelland Lecture, and I am gratified that the written version
of the lecture is to form part of the proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical
Engineering. Following in the footsteps of the first McClelland
Lecturer, Don Murff (Murff, 2012), is no easy feat, although I
must admit to having become somewhat accustomed to this
during my career. More times than I can remember I have found
(often retrospectively) that an analytical contribution I have
offered has been covered elegantly by Don in a prior
publication. It is fitting, therefore, to continue the theme of his
own McClelland lecture, in targeting the gap between theory
and practice, drawing attention to and summarising various
analytical contributions.
In an era where virtually any geotechnical application can be
modelled numerically, with idealisations potentially limited
only to those associated with the constitutive response of the
soil, it is tempting to wonder whether true analytical solutions
still have a role. At the opposite extreme, design guidelines such
as API (2011) and ISO (2003, 2007) are inevitably slow to
evolve and in many places rely on somewhat dated suggestions,
either empirical or quasi-analytical. There is limited incentive to
refine them through analysis without clear evidence of lack of
conservatism, or the reverse, excessive conservatism.
The potential of analysis is its ability to provide a direct,
ideally quantitative, link between a required output and the
various input parameters for a given application. At a basic
level, dimensional analysis should indicate appropriate non-
dimensional forms for input and output quantities. Analytical
solutions will typically contain idealisations, either of the
problem geometry or of the soil response, for example linear
elasticity for stiffness solutions, or perfect plasticity for capacity
solutions. However, they still provide a framework linking the
outcome to the various input parameters, highlighting the
critical sensitivities of the response, facilitating parametric
studies and quantifying the effect of different idealisations.
The paper takes a retrospective look at some of the analytical
contributions relevant to offshore geotechnical engineering,
drawing attention to the potential application of the solutions in
design guidelines and day to day practice. The first part of the
paper revisits solutions for the axial and lateral response of pile
foundations, which are still the main type of foundation for
offshore structures in moderate or shallow water depths and for
tension leg platforms in deeper water. The remainder of the
paper then focuses more on applications relevant for deep water
developments, including subsea foundations, anchors and
pipelines. Of necessity, restrictions on the length of the paper
have required me to focus on a few specific issues within each
topic, in particular where solutions point the way towards
improved design recommendations, and recent work addressing
developing areas of offshore geotechnical engineering.
Before discussing the applications themselves, I should
clarify what I intend by the word ‘analytical’ within the present
context. I include within this term appropriately conceived
parametric studies undertaken through numerical analysis.
These should lead to algebraic expressions or charts that may be
used in design, identifying the relative contribution of non-
dimensional groups of parameters that affect the result. By
contrast, an algebraic fit through experimental data will rarely
provide comparable insight, and should instead be taken as
encouragement to quantify the phenomenon through analytical
or numerical means. That said, I have always been a strong
proponent of the need for high quality experimental data, but
with the primary objectives of stimulating understanding of the
problem for subsequent analysis, and where necessary to
calibrate specific areas of uncertainty in analytical models.
Lecture
