11 February 2012

Continued questions about the role of the quarry in the Tumbi Quarry landslide

Posted by Dave Petley

Interestingly, over the last few days there appears to have been increased interest in the role of the quarry in the Tumbi Quarry landslide.  Regular readers will remember that the National Disaster Center report indicated that the failure developed as subsidence close to the quarry.  Triggering of the landslide was associated with heavy rainfall, with no evidence to support this trigger, and no other mention was made of the quarry, even though it is clear from the images that a section of it was removed by the landslide.  In an earlier post I questioned both the evidence for rainfall / high groundwater as the trigger and the lack of consideration of the role that the trigger might have played.

Two recent statements have been reported by those associated with the disaster.  First, in a Radio Australia interview, Martin Path, who is the Principal Disaster Coordinator for the  Southern Highlands province in  Papua New Guinea, has given a rather different perspective to that from the official report(LAM is Sen Lam, the interviewer):

LAM: And Martin, at the time of the disaster, the landslide was blamed on blasting at nearby quarries, near the Exxon-Mobil LNG project in the Southern Highlands. Is a clearer picture emerging, of what happened, that early Tuesday morning? Are you getting a clearer picture now?

PATH: Not as yet. We have yet to conduct, er, we are conducting some portion of the independent investigation that the National … Council has endorsed. We believe that an independent investigation unit has been assigned, and we believe that this group would be up here over the weekend. But we also have the geo-hazard technical team on the ground now, since Day One. So they’re still providing the information, so with regards to the quarry that was established some years back, we have yet to establish the actual cause, what actually caused this slide.

There is a strange confusion here – “We have yet to conduct, er, we are conducting some portion of the independent investigation…We believe that an independent investigation unit has been assigned, and we believe that this group would be up here over the weekend.”  It is hard to interpret what is going on from that.

Second, my earlier posts have been picked up by the PNG Industry News website, who then have some comments from an ExxonMobil spokesperson:

An ExxonMobil spokesperson previously told PNGIndustryNews.net that PNG LNG contractors did not conduct any blasting at the Tumbi quarry.

She said a project contractor completed work at the Tumbi quarry in August, 2011.

“There was no need for [PNG LNG-related] blasting at this quarry,” she said.

“The Tumbi quarry has been operating for over two decades.”

There are two issues here.  First, the spokesperson notes that there had been no blasting at the quarry, which is in contradiction to the reports from local people.  I have no way of verifying this either way.  Second, she suggests that work associated with the LNG project ceased six or so months ago.  The unstated implication is that this means that the Exxon-Mobile work was not responsible for the collapse.

However, unfortunately slope behaviour is not as simple as that.  Whilst slope collapses can occur spontaneously during quarrying, there is often a time gap between the processes that lead to failure, and the failure itself occurring.  This is a mechanism known as progressive failure, which was first described in the 1960s.  There are two elements to this:

1. It may well be that the quarrying operations occurred when groundwater levels were low, or indeed that the quarrying caused an initial drawdown (release) of groundwater.  In such circumstances, the groundwater recovers the stability of the slope can reduce, allowing a failure to develop.  Seco

2. Failure of the slope occurs when a shear surface is fully formed that allows the landslide to detach from the surrounding rock.  This is not a spontaneous process, but requires the formation and growth of planes of weakness.  Such a process can take weeks or even years.  I have written several papers on this very process, such as this (click on the title to download the paper):

Petley, D.N., Higuchi, T., Petley, D.J., Bulmer, M.H., and Carey, J.  2005. The development of progressive landslide failure in cohesive materials.  Geology, 33, 3, 201-204

It can be the case that cutting a slope in a process such as quarrying initiates the progressive failure mechanism, and allows the slope to collapse some time later.

None of this inevitably means that the quarry was the cause of the landslide.  It may have been the direct cause, it may have been one of several causes, or it may be irrelevant.  However, understanding its role is critical.  Once again I would emphasise that this can only be uncovered with a proper forensic investigation by a team who understand the complexity of this type of landslide.  This is not a trivial task, and will involve detailed field mapping; structural measurements; examining aerial and ground images before and after the landslide; examining the quarry design and operation; looking at rainfall records; etc. The methodology for such an investigation is well-established, but undertaking it requires very specialised skills.  Usually this will require an international team.

The loss of a significant number of lives really demands that this landslide is investigated properly.  I do hope that this process is now underway.