7 February 2012

An official report on the Tumbi Quarry landslide

Posted by Dave Petley

The National Disaster Council of Papua New Guinea has posted online a report into the Tumbi Quarry landslide (see previous posts here, here, and here).  Presumably this is the “official investigation” that was promised.  They should be credited with making this available online, and the document has some useful information on, and images of, the landslide.  It is also helpful to get the official view on the landslide event, and there is a management plan for the victims and survivors of the accident.

However, there are aspects of the report that are perhaps interesting, as follows:

1. The report only very briefly mentions the quarry.  The report states that “naturally high geological weaknesses assisted by abnormally high rain fall combined to cause subsidence to the immediate north west of the quarry triggering a landslide of composite debris.”  The image on the front cover clearly shows a part of the quarry, with some of the benches truncated by the lateral margin of the landslide:

The report is ambiguous as to whether this means that the landslide was associated with the quarry.  It maybe that the quarry access / haul road also crossed the area now destroyed by the landslide (as per the Esso Highlands plan), and yet this is also not discussed.  Clearly at least a part of the quarry was lost in the landslide, so a proper discussion of whether the quarry played a role in activating these weaknesses seems essential to me.

Note also that these naturally-occurring geological weaknesses are not described or discussed (I assume that this means a set of joints?), and the mechanism of “subsidence” is also quite strange.

2. The landslide trigger is definitively described as being rainfall, on the basis that there was no recorded seismicity.  However, although “continuous heavy rainfall” is described, there is no data to support this, and no eye-witness reports.  The next sentence talks about “abnormally high rainfall”, but I can find no other reference to this event.  I wonder what evidence there is for this abnormal rainfall event, given the high rainfall that this area is likely to receive anyway.

3. There are odd aspects of the landslide mechanism as described in the report.  First, the report notes that pools and seepage on the shear face indicate that the “ground water rose significantly above its historical levels”.  I do not understand this logic.  Seepage and pools are likely to occur in the aftermath of almost any deep landslide of this type, and I do not see why they indicate that the groundwater levels were abnormally high, or indeed that groundwater even played a substantial role.  Second, the report notes that the initial assessment team “saw clear evidence of liquefaction of the rock formation”.  This is most surprising.  Limestone is not a material that undergoes liquefaction – I have never heard of such a mechanism in any hard (as defined from an engineering behaviour perspective) rock – and so I just cannot understand this purported process.  Unfortunately, it is not discussed further.

So overall, whilst the description of the site is really helpful, though it would be good to see a more detailed consideration of the ways in which the quarry might have been a factor in activating these geological weaknesses.  In designing a quarry it is normal practice to ensure that such natural weaknesses are not exposed to the point that they can permit slope failure to occur.

To my mind this report should not be considered to be the definitive analysis of this landslide.  Much more detailed analysis is needed, and lessons need to be learnt in terms of other slopes in Papua New Guinea.