Geotechnical Design in the face of uncertaintyE. H. Davis Memorial Lecture

It has always seemed to me that uncertainty is the very essence of geotechnical engineering. Our materials are natural’ in origin, often irregular in form and highly variable in their properties. They are usually obscured from sight and can be investigated only to a small extent at great expense. The resulting uncertainty is often large and can have enormous economic consequences, in many cases being measured in hundreds of millions of dollars. The effects on project safety, and the legal consequences are matters of continual concern. It is important that we ourselves understand the limitations of our technology and are able to communicate them to our clients and the community.

In this talk I will discuss six types of uncertainty that we face and make the point that only one of these is dealt with in what we would normally consider to be an engineering analysis. For this type of uncertainty the way in which safety factors and probability of failure are used to indicate uncertainty will be discussed and the point made that these can be presented in a seriously misleading way. The effects of bias, statistical errors, homogeneity, scale effects, and spatial dependency on estimates of risk will also be discussed. Some data is presented as a first approach to estimating the frequency of occurrence of geotechnical problems in major projects, both mining a~d civil, over the past 30 years.

The point is then made that geotechnical problems only lead to post-construction failures or mining disasters if they are not recognised in the construction or mining period. Case histories indicate that human errors are usually a contributing cause of such failures.

Finally, the investigation and design of the proposed High Wall of the Mae Moh Open Cut Colliery in Thailand will be reviewed as an example of a complex geotechnical problem analysed in the face of uncertainty.