Geology Reference
In-Depth Information
be anticipated and the design and construction methodology should
have coped with this.
Highway collapse, Singapore
A major failure of temporary works occurred during the construction
of part of the Circle Line of the MRT, Singapore, on 20 April 2004.
Four persons were killed. The post-failure investigations were pre-
sented to a meeting of the British Geotechnical Association (Hight,
2009).
At the time of collapse, excavation was taking place of a 34m deep
excavation for a cut and cover tunnel between two diaphragm walls.
As the excavation was lowered, the vertical walls were supported by a
system of steel struts, waler beams and kingposts. By the time of the
failure, there were nine levels of struts and some of these were instru-
mented with load cells and strain gauges. About six hours before the
final failure, some of the struts began to lose load rapidly, whilst others
took on more load. This has been interpreted as brittle failure (rapid
and unrecoverable loss of strength) of some of the strutting. A detail of
the strut beams/waling collection was considered the major factor in
the collapse (Magnus
et al
., 2005).
The excavation was in an area of reclaimed land, but most of the
excavation was through very soft and soft to
firm predominantly clays
of the Kallang Formation, which is an extensive and well-investigated
stratum in Singapore (Bird
et al
., 2003). It was concluded that there
had been relative vertical movement between the diaphragm walls and
strutting system as part of the failure mechanism. At the meeting
reported, it was argued by Hight (2009) that the failure also related
to the fact that the temporary works were designed to a FoS of 1.2 yet
contained brittle elements, including the steel strutting/waling connec-
tion and a concrete strut formed at depth using jet piling. It was
suggested that the trigger for the failure was due to ground conditions
deviating from the design assumptions, in that the undrained strength
pro
le was lower than assumed and a complex geology, which involved
signi
cant variations in stratigraphy between the diaphragm walls.
The comment regarding undrained strengths actually does not relate
to incorrectly measured or anticipated strengths but an error in the
way that strength was dealt with in numerical modelling (Magnus
et al
., 2005). The comment on complex geology apparently relates
to a local deepening of the Kallang Clay but such occurrences are
commonplace in Singapore (Bird
et al
., 2003).
In this case, the fundamental cause of the failure was the structural
connection between struts and waling beams that was under-designed.
As reported from the BGAmeeting,
The waling detail that yielded and
underwent brittle failure had a direct load capacity that was only
'
