Geotech & mining engineering
The application of dynamic loads to loose, saturated, granular deposits may trigger the liquefaction of the ground. Principia has often analysed this problem to determine if it can occur, to propose remediation measures when appropriate and, in some cases, to establish the sequence of events in respect of past accidents.
The studies have been carried out in many different contexts, both onshore and offshore, concerning foundations for liquefied natural gas tanks, electricity generating plants, wind turbines, bridges, slopes, and many singular locations such as the artificial island where the Oresund crossing goes from a tunnel to a bridge connection or the nuclear submarine docks in Devonport in the United Kingdom.
The tailings of the Aznalcóllar pond, pictured earlier, underwent liquefaction during the failure, which allowed them to continue pushing behind the dam in spite of its large displacement.
Study of pile foundations
When the natural soil is unable to provide the necessary stiffness or strength under the design loads, a pile foundation may be required. The study of the response of pile groups may be complex, particularly under dynamic loads (earthquakes, ship impact, winds, etc). In dense pile groups, the interaction between neighbouring piles via the intervening soil has important effects on the dynamics of the group.
Principia uses special purpose software for evaluating the dynamic impedances of dense pile groups, a methodology that the company has applied in many projects of high responsibility: nuclear power plants, liquefied natural gas tanks, large bridges, petrochemical plants, etc.
The photographs correspond to the pile foundation of two large LNG tanks in India, designed by Principia in collaboration with Esteyco. Each tank is supported on 578 concrete piles, 1 m in diameter and about 40 m long.
Aznalcóllar tailings pond
Post-failure investigations are generally far more complex than the original design calculations. The latter can be relatively simplistic, as they only require that all errors are on the safe side. By contrast, the investigation of a real accident must incorporate the actual mechanisms and behaviours undergone; in this context, “conservative” assumptions are simply wrong.
The infamous failure of the Aznalcóllar tailings pond in April 1998 was analysed by Principia for Boliden Apirsa, owner of the facility, in order to identify the causes of the accident. The studies showed that failure occurred because the design did not account for the slow consolidation of the underlying blue clays and the brittleness of their structural behaviour.
The detailed simulation of the failure process is particularly difficult in this case. It requires conducting coupled calculations, using effective stresses in the soil, along the construction period, as well as an explicit representation of the softening response that the brittleness of the clays entails. Further, the onset of failure produces tailings liquefaction and a sudden acceleration of events; the dam moved up to 60 m along hundreds of metres of its length.
The first figure shows the calculated failure mechanism, which is consistent with the one actually experienced. The second one is a photograph of the pond taken after the accident.