30 March 2021
The resilience of rail infrastructure: reports from the Network Rail earthworks and weather advisory task forces
Posted by Dave Petley
The resilience of rail infrastructure
On 12 August 2020 a very serious rail accident occurred at Carmont near to Stonehaven in Scotland when a fast-moving train struck a pile of debris on the track. The train derailed, with some of the set crashing down an embankment. Fortunately the train had only nine passengers and crew onboard, but sadly three of them were killed.
The accident occurred on a section of track that had been subject to a number of landslides and other geotechnical failures in recent years, and it happened during a well-forecast period of heavy rainfall. Inevitably this focussed attention on the state of the geotechnical assets managed by the UK rail infrastructure operator, Network Rail. It also raised questions about the viability of an early warning system for heavy rainfall events on the rail system.
In the aftermath of the accident, the Secretary of State for Transport commissioned two reports in addition of course to the accident investigation. These reports were:
- An Earthworks Taskforce Management report led by Lord Robert Mair
- A Weather Advisory Taskforce Report led by Dame Julia Slingo
In addition, Network Rail have produced a report providing an update on the Resilience of Rail Infrastructure.
The reports have now been published online. I’m going to state clearly an interest here as I was a member of the Earthworks Taskforce, and indeed within that report there is a separate report on The Role of Technology in Slope Management that I authored.

The front cover of the Network Rail report on Earthworks Management
As I was a member of the Earthworks Management task force I’m not going to comment on the reports – that is for others. It it is, I believe, an exceptionally detailed and comprehensive review of the topic (including the appendices it runs to over 540 pages). It provides an excellent history of the development of earthworks on the UK rail system, the mechanisms of failure, management approaches and potential areas of research.
I think the two reports will be considerable interest to a wide variety of people, not just those in the UK and those working on rail networks.
The Executive Summary of the Resilience of Rail Infrastructure report makes the following key observation:
Most earthworks and drainage beside our railway were built more than 150 years ago. They were constructed without detailed engineering design and not to current standards at a time when the risks associated with earthworks were not scientifically understood. Consequently, cuttings and embankments were constructed with steep and unreinforced slopes. Commonly, drainage systems were not built with the original earthworks and the drainage that was installed was not designed to cope with floods. While they have served us well, and despite many improvements over the years, they are not as robust as a modern-day equivalent. Earthworks structures are complex to manage due to underlying geology, adjacent environments and assets, local weather patterns and a propensity for very localised failure. Rebuilding thousands of miles of earthworks and drainage systems to modern-day standards is not practicable, either from a funding or deliverability perspective.
The magnitude of the challenge is very clear, but these two reports are intended to provide a clear roadmap to keep the UK railways safe.
In the section on InSAR I was very happy to see links to Terra Motion’s online land motion maps of the UK and Germany. Readers of the report who follow these links will see complete coverage over all land cover classes. This is a recent development in satellite InSAR pioneered by Terra Motion which contrasts with the traditional PS and DS approaches that are mentioned, offering a solution over all land cover classes, rural and urban alike, without the need for corner reflectors. It recognises the non-persistent nature of natural surfaces and seeks a solution through a more in-depth analysis of the temporal coherence of each pixel. Over the last decade it has been the subject of many peer-reviewed papers and a recent Nature paper showed that it was able to detect precursors of the Brumadinho tailings dam failure when other approaches had failed. It has been particularly successful at monitoring the seasonal rise and fall of peatland surfaces and recent work by the University of Nottingham (unpublished as yet) on a spate of incidents at the end of last year in Ireland is showing that it has a remarkable potential to predict bog bursts in open and forested terrain.