The purpose of this posting is to solicit some discussion and hopefully
obtain some of your experience (either positive or negative) on the issue
of predicting settlements.  I would be pleased to assemble and summarize
the responses that I receive, then make them available to those individuals
who may be interested.  I would like to limit this discussion to total
settlement (predicted versus observed) and not the time-rate of settlement
(consolidation) issue.

A comment or observation that geotechnical engineers seem to get fairly
often (at least here in the U.S.) is that we generally over estimate the
settlements that actually occur in the field.  If we accept that the basic
approach or model that we are using is fundamentally correct (effective
stress approach, where we determine the stress-strain relationship for the
soil in the laboratory and apply this to compute volumetric strains in the
field for the anticipated effective stress change), and if we are confident
of the geologic profile and the fact that our test data is
"representative," then it seems that there can be only a few reasons for
this discrepancy.

1.  The size of the sample and method of testing does not provide results
(data) truly indicative of the material insitu.

2.  Our method(s) for predicting the distribution of stress within the
foundation are too inaccurate.

3.  The actual settlements are not being measured (for example, the
foundation settlement beneath an earth structure during construction may
not be monitored, but observations might be made of post construction
settlement of the earth structure).

There may be those in the profession who believe that there is a problem
with the fundamental model, and those comments are most welcome also.

If you chose to participate in this discussion it would be very helpful if
you would also indicate (1) the soil conditions in which you were working
(clay, sand, stratified deposit or homogenous, thick or thin compressible
layer, etc.), and (2) the type and size of loading (40 story high rise
building, 50 foot high earth embankment, height and diameter of a storage
tank, etc.).

Your participation in this discussion is most genuinely appreciated.

Regards,

Dave Kerkes
----------------------------------------------------
David J. Kerkes, Ph.D., P.E.
Consulting Geotechnical Engineer
Houston, Texas 
U.S.A.