The Landslide Blog is written by Dave Petley, who is widely recognized as a world leader in the study and management of landslides.
In the aftermath of the horrific 16 December 2022 landslide at Batang Kali in Malaysia, in which 31 people were killed, an official investigation was instigated by the government. This was not a public inquiry, it was a forensic study undertaken to understand the causes of the landslide and to learn lessons. The official report was released this week in Kuala Lumpur, attracting great interest within Malaysia.
The report, which is in English, is available online. It was led by CKC, the Slope Engineering Branch of the Public Works Department, with input from many other agencies. It’s main conclusions are:
- The landslide was a rotational failure that transitioned into a debris flow;
- The cause of the landslide was heavy rainfall that generated high pore water pressures, in part through seepage;
- There is no strong evidence to suggest that anthoropogenic activities were a factor in the landslide.
It is hard to disagree with any of these points, and I think that the analysis has been done well. CKC are a strong organisation, although I would agree with others that it should be strengthened (see below).
But there are some curiosities in the report that are worth noting. The study concludes that an interesting sequence of events occurred, involving two slope failures. This is illustrated by the diagram below, from the report:-
In the first failure, a large landslide occurred in the upper slope with a runout that reached the final toe. This landslide created a temporary debris dam, which failed 20 minutes later to create the final morphology.
Whilst multiple phases of slope failure are common, it is quite unusual that the second failure did not lead to any retrogression or any increase in runout distance. That is not unfeasible, but it seems surprising.
Two of the three witnesses describe two phases of failure, although neither give much detail about what those failures involved (which is unsurprising as it was night time). The project team has reconstructed the original slope (of which more below) in a model that incorporates seepage. The model indicates the development of a rotational failure that extended to the road, driven by high pore water pressures. This seems reasonable.
The team conclude that after failure, the debris formed a debris dam on the slope, which then remobilised to form the second landslide because the mass trapped water from the scarp area. But note that from the diagram above, the first failure also ran out to the ultimate toe of the slope, as I highlighted previously.
As far as I can see, there is no explanation given for how this interpretation has been made, and there is no consideration, or discounting, of alternative models. For example, a smaller failure lower on the slope could have developed with a lower runout, followed by retrogression in the second failure, which then generated the full runout. Or, the initial failure spanned only a part of the ultimate landslide width, with the second failure involving the remaining material.
I have no evidence to support or refute either of these hypotheses, but I cannot discount them based on what is provided. The interpretation seems to be driven by a 2D slope stability model, but great care is needed in the interpretation of such models. The actual configuration of the landslide is likely to be very sensitive to input parameters that can only be estimated.
A further aspect that is intriguing is the presence of an embankment on the upper part of the slope, on the downslope side of the road. This can be seen in the diagram below, also from the report:-
The embankment, constructed of fill, sits at a really critical point on the slope. It is notable that the rear scarp of the final failure is right at the edge of the road, which is at or close to the boundary of the fill, which is captured quite well in the image on the front of the report:-
The report indicates that this fill embankment first appears in satellite imagery in 1992. The geometry of the embankment as shown in the cross-section is rather unusual. I am intrigued as to why an embankment would be built like this. But more importantly, the role of the embankment is not covered in detail in the report. It seems to me that it could be critical because:
- It appears to have added mass to the crown of the ultimate landslide;
- Fill can be both permeable and can allow trapping of water to generate high pore water pressures;
- Failure of the embankment is a conceivable mechanism for the first failure.
This matters because the report concludes as follows:
I don’t disagree with the conclusion, but it seems to me to have taken a particular view of what is meant by human activities.
In the Malaysian media, the report has had a mixed response. Perhaps the most important long term reaction though is in a letter from Kua Kia Soong, a former MP for Petaling Jaya, who writes:
More importantly, we must thoroughly overhaul the entire system governing hillslope development. Stringent EIA conditions, regular maintenance, and transparent engagement with all citizen stakeholders must be established. Professionalism in engineering projects with a steadfast commitment to stability and user safety must be paramount. Only by addressing these deeply ingrained systemic issues can we hope to prevent such heart-wrenching tragedies from recurring.
Indeed! A good start would be further investment in CKC.
