Research over the last 10 years has proven conventional driven pile designs to be highly unreliable, with problems particularly severe at sites underlain by sand. The most widely used design methods can result in very high scatter and bias.
Several database studies have shown that standard deviations of about 60% between predictions and measurements apply in even the best controlled cases.
One example is the competition run by Ciria and Imperial College that was reported at the 15th International Conference on Soil Mechanics and Geotechnical Engineering, held in Istanbul in 2001.
International experts were invited to predict the test response of piles installed at the Dunkirk dense sand research site in France. The highest prediction made for a 10m long 457mm diameter driven steel pipe pile was six times the lowest; the standard deviation was about 60%.
The major Euripides study conducted in The Netherlands by Fugro, Geodia and their partners provides another recent example. Tests on 762mm pipe piles driven to 43m in very dense sand gave short-term capacities 65% higher than expected by the standard American Petroleum Institute offshore pro edures; longterm capacities were higher still, far exceeding the test rig capacity.
In other cases, such as the Jamuna Bridge project in Bangladesh (GE November 1998), tests on driven piles gave capacities far smaller than those expected by conventional design approaches.
Recognising these and other shortcomings, researchers at Imperial College London decided to investigate the basic mechanics of driven pile behaviour. A series of PhD studies were carried out by myself, Andrew Bond, Barry Lehane and Fiona Chow from the early 1980s. These involved laboratory, theoretical, field testing with instrumented piles, full scale tests and in-service monitoring.
Funding came from the Marine Technology Directorate (MTD) and a group of industrial participants, most notably Amoco UK, the Health and Safety Executive and Shell UK.
Radically different design procedures were developed for clays and sands, summarised in a book et (Jardine and Chow) published by MTD in 1996. These pro edures at least halved the scatter associated with conventional design methods.
Designers could make considerable savings in many cases or improve safety and reliability in others. Shell UK has applied the MTD approach in its offshore projects since 1996 and others have used the procedures for onshore and offshore projects from Azerbaijan to Venezuela.
Recognition of the work includes prizes from the Royal Academy of Engineers and Shell UK.
Research has progressed steadily since 1996, aided by further masters and PhD projects and field experience.
As the MTD has been disbanded, the authors turned to Thomas Telford to publish a new edition of the book in March 2005, suggesting the methods should now be known as Imperial College Pile (ICP).
The new volume broadens and updates the original recommendations, discussing the ICP methods in relation to recent research and practical experience. Other key points include:
The methods can now be applied to a much wider range of projects and pile types, with guidance on square and H piles as well as open and closed ended pipe piles.
It considers conditions other than just static tension and compression loading. It also covers cyclic, seismic and group action effects.
A wider range of soil types is explored, ranging from calcareous sands to diatomaceous clays.
Attention is drawn to the potentially dramatic effects of ageing on shaft capacity in sand. This was explored in recent fi eld research at Dunkirk, proving previously untested piles develop far more dramatic increases with time than indicated by the trend (found from retesting the same piles) in the first MTD booklet.
It gives detailed descriptions of how the procedures should be applied by geotechnical engineers, supported by worked examples involving well known case histories.
It presents further evidence of the methods' reliability. A substantially updated database of high quality pile tests shows how the ICP approaches lead to good predictions.
It includes sections on how reliability theory may be used to assess the appropriate factors of safety, or load and resistance factors.
Richard Jardine, professor of geomechanics at Imperial College London, led the research and developed the new methods with Dr Fiona Chow, lead geotechnical engineer at WorleyParsons.
Dr Jamie Standing, senior lecturer in soil mechanics at Imperial College, played a key role in the recent field research.
Dr Robert Overy, principal geotechnical engineer with Shell UK, was on the research steering committee and has been prominent since in many offshore applications of the procedures.