Loss adjusters would be wise to consider measuring soil suction for subsidence claims, says Andrew Ridley.
Five and a half years ago (GE April 99) Nick Ramsey wrote an excellent Talking Point extolling the virtues of properly conducted ground investigations in connection with house subsidence claims.
At the time, the insurance industry was moving rapidly towards the use of disturbed samples in an attempt to minimise costs.Water contents and Atterberg Limits were being favoured over soil suction and tests were frequently carried out on samples that had been allowed to dry because of delays.
So where has the insurance industry gone in the new millennium?
Unfortunately the answer is possibly towards, if not into, a black hole! In search of the ultimate cost reduction loss adjusters are now turning to knowledge-based systems to remove the need for investigation all together.
This in itself might be commendable were it not for the fact that key information being used to generate the database of knowledge is fundamentally flawed.
Investigations by some now consist of a hand augered borehole, followed by the measurement of volume change in an oedometer when the recompacted soil is flooded. Volume increases are interpreted as indicating desiccation and therefore the potential for subsidence.
However, as most geotechnical engineers will appreciate, the volume of recompacted clays, particularly those that are desiccated, can increase, decrease or most ambiguously remain constant when wetted depending on (a) the density at which they have been compacted and (b) the magnitude of the applied vertical stress during flooding.
Moreover, even fully saturated recompacted clays can swell when flooded in an oedometer. Therefore volume increases can really only provide an indication that the material being tested is predominantly clay (something that should have been identified by a tactile examination of the soil).
Therein lies the problem and possibly the solution. Having identified that the soil is clay, the loss adjusting community has realised it can identify those properties that are vulnerable to subsidence and monitor them directly using surveying methods. This provides simple pictures that are easy to interpret and for this they do not require engineers.
However, some loss adjusters still measure their success through customer satisfaction as well as reduced costs. The former can only really be successfully delivered if all cases of subsidence are correctly identified. This may be achieved using the approach being adopted but possibly at the expense of misdiagnosing a larger proportion of the claims.
Better, I suggest, that the database of knowledge is built on the strength of a more reliable measurement. Soil suction in itself defines desiccation and will correctly identify the problem even on disturbed samples, if the effect of disturbance is properly accounted for.
In contrast to the insurance industry, the geotechnical fraternity is embracing the measurement of soil suction.
This is because recent research has led to a better understanding of how soil suction can be used and the development of new, easier to apply, and economical techniques of measuring it both in the laboratory and the field.
Some of the ways in which measurements of soil suction are being used are:
to identify disturbance in samples removed from the ground;
to estimate the average insitu effective stress once the effect of disturbance has been quantified;
to apply more rigorous controls to the compaction of fill materials where swelling or collapse is likely to cause problems; and l to assess the stability of natural and engineered slopes.
Projects such as Heathrow Terminal 5, The new Wembley Stadium development and London's Crossrail have made extensive use of suction measurements, which have contributed to cost reductions through more appropriately designed laboratory testing programmes and the efficient use of temporary works.
Andrew Ridley is a founding director of Geotechnical Observations and part-time lecturer at Imperial College, London.