The Irish railway network is supported by earthworks mostly constructed in the late 1800s. These earthwork assets are generally not constructed to modern design standards. When combined with extreme rainfall events, which are becoming increasingly common, they can therefore pose a serious challenge to the safe, reliable running of the railway services. In 2014, Iarnród Éireann/Irish Rail (IÉ) appointed Gavin and Doherty Geosolutions (GDG), a specialist geotechnical engineering consultancy, to develop an earthworks decision support tool (DST). Following over two years in development, this DST provides comprehensive risk assessment, prioritisation and cost/benefit investment analysis of the railway’s earthwork assets. This is a significant achievement for IÉ in that it delivers an objective model to support the decision-making across a nationally diverse asset base of earthworks in the key areas of railway safety, operations, cost and investment.

Challenge­­s facing the network

The Irish railway network of over 2,800km is supported by 3,700 earthwork assets. These earthworks, comprising both cuttings and embankments (C&E), total 1,300km in length.  Cuttings are where the railway corridor has been cut into the natural topography.  Embankments support the track where it is elevated higher than the natural ground levels. Even minor failures in cuttings when close enough to the track can cause derailments, while derailments from embankments are potentially more serious but are generally only caused by major failures. The vast number of assets, coupled with the effects of cuttings versus embankments, as well as their own many individual characteristics, creates a complex network of slopes that need to be managed safely for the on-going operation of railway services and effectively to ensure optimisation of investment. [caption id="attachment_33710" align="alignright" width="300"]manulla Figure 1: Manulla landslide 2007[/caption] In addition, side slopes of soil and rock, due to their natural variability, perform with different levels of reliability throughout the network. Extreme rainfall events are becoming more common. These rainfall storms can trigger slope failures and further increase the variability in performance of assets on the network. The safe operation of the railway relies on the stability of the earthworks and therefore it is critical to manage the risk of slope failure. Slope instability and landslides are not uncommon occurrences. Every year across Europe, significant damage to infrastructure occurs, with slope instability responsible for consequences such as derailments and even fatalities. [caption id="attachment_33711" align="alignright" width="300"]rushbrook-rockfall-2014 Figure 2: Rushbrook rockfall 2014[/caption] Since the Irish railway was mostly constructed in the late 1800s, IÉ earthworks are generally more than 100 years old and do not meet modern design standards. As a result, the network is subject to multiple instances of landslides and rockfalls every year (Fig. 1-3). Sometimes resulting in train damages, derailments, temporary line closures and expensive remediation measures. The critical challenge facing the railway network, however, is how all these difficult factors are assessed. Standard railway network practice is to routinely inspect all assets visually and record the findings in an asset database. IÉ inspectors carry out hundreds of inspections annually and record observations and asset-condition scores in the Irish Rail Infrastructure Asset Management System, an overarching asset database. The reports are mostly qualitative rather than quantitative, as they are primarily reliant on visual inspections and so the results of these inspections tend to be subjective. One of the primary objectives of the C&E DST was to provide an additional objective assessment of the assets.

Developing the slope management framework

GDG, with its substantial experience in landslide risk assessment R&D, developed a robust scientific framework to quantify the relative risk of different slope assets across the network. This framework and Decision Support Tool utilises best engineering practice for slope stability calculations and supplements this by probabilistic simulations. [caption id="attachment_33712" align="alignright" width="300"]waterford-rockfall-2013 Figure 3: Waterford rockfall 2013[/caption] GDG and IÉ C&E inspectors collaborated through workshops to identify observable features that are frequently neglected in geotechnical calculations. These ‘degradation factors’ range from vegetation types on slopes and drainage issues to history of slope failures and are weighted by importance in causing or preventing failures. IÉ records of slope failures back to the early 1900s supported the inclusion of these degradation factors. The calculations for probability of failure were then updated with these degradation factors to give more refined results. Since slopes further back from the track may fail but not impact the track, the consequence of failures was calculated using height/slope angle ratios and clearance to the track values and further prioritised by train frequency and train speeds. The project brings innovative solutions for obtaining the three universal features of hazard, consequence and risk in a quantitative, objective, scientific and repeatable manner. To create a DST that would stay up-to-date with the asset register, it was essential to pair the decision support tool with the overall Irish Rail asset-management system. This required restructuring of the asset register and the creation of synergies between the base-asset data information with that required to run the model. An accurate DST fundamentally requires a database of high quality, objective and consistent information. This was achieved using information already available and by establishing strict criteria for all 36 inputs. GDG analysed cross-sections from the 2011 Lidar railway network survey at 10m intervals for the 1,300km of earthworks and established material properties for each of the 3,700 assets from geological maps and on-site testing. In summer 2016, IÉ engineers collected information on degradation factors from recent inspection reports and inputted over 50,000 values such that the model is now fully populated. The resulting probabilistic calculations use bespoke software code to interact automatically with IÉ’s asset database and enables thousands of simultaneous, automatic and resource-free calculations.

Innovations of risk-management software

Commercial slope stability software typically analyses the stability of individual earthworks. The challenge at IÉ was the requirement to analyse all 3,700 assets concurrently. GDG coded a standalone program to carry out simultaneous probabilistic and risk-assessment calculations of all assets. Such complex calculations need to be housed in easy-to-use software. To create an intuitive platform for the DST, GDG produced a graphical user interface (GUI) with several interactive modules (see Fig. 4). Users can chose to focus on an individual asset, localised areas or look at the entire network and select remediation techniques which manipulate the asset data to model the benefit to the asset or assets. [caption id="attachment_33713" align="alignright" width="300"]figure3 Figure 4: Graphical user interface of decision support tool (CLICK TO ENLARGE)[/caption] From the outset, this DST needed to be innovative to handle such a large number of assets. Analysing over 3,000 assets meant typical site investigations to determine material properties for each individual asset was unfeasible. To account for uncertainties in the network, probabilistic calculations rather than ‘factor of safety’ calculations were undertaken. A key feature of the model is that the sophisticated software can also carry out weather simulations – something of particular benefit in a time where increasingly severe weather events are more frequently impacting on railway service operations. For any model, verifying the results is a crucial step. GDG carried out a suite of independent calculations to ensure the coding in the model was correct. Additionally, since many slope failures in recent years were not predicted by condition-based inspections, IÉ engineers also tested the model by inputting information from the inspection prior to an unexpected failure. When the model was run for the entire network with asset information included from these failed assets, the results showed these assets as critical. This gives further confidence to the objective and sophisticated analysis capabilities of the DST.

National and international impact

The implications of this project for Irish Rail are far-reaching. Improved assessment of the earthworks that support the track creates a safer, more reliable railway network. As the DST is implemented in planning maintenance works for critical assets, the reduced likelihood of landslides causing derailments or emergency line closures will improve customer confidence in the railway. As the greenest mode of transport, a more trusted railway is also good for the environment. A rule of thumb for emergency remediation works is that it costs ten times that of a preventive intervention. At a time where the requirement for funding dramatically outstrips the available financial resources, it is important that the effectiveness of these scarce resources is maximised. With the DST identifying critical assets for investment before failures occur, an optimisation of limited funding is assured. Finally, to the author’s knowledge, this is the first time worldwide that a network-wide geotechnical slope analysis has been carried out. So it is fair to say, this project really has been groundbreaking. Author: Catherine Joyce is a project engineer with Irish Rail. She is a chartered engineer with over ten years’ experience in the transportation industry. Following design and construction supervision of structures on several motorway projects, she joined Irish Rail in 2011. Joyce is now specialising in asset management of earthworks and bridges at Irish Rail.