The following elaborates on the research conducted over the course of the program, broken down by past research, which includes Phase I (2003-2008) and Phase II (2009-2015), and current research, under which Phase III (2015-2020) currently falls. Phase IV (2020-2025) research priorities are currently under development, and will be approved upon with industry and government stakeholders.
Past Research: Phase I
At the outset of Phase I, researchers were provided access to a railway industry database of ground hazard incident reports collected since the beginning of the last century. With this information, researchers developed a ground hazard classification system that identified and defined all types of ground hazards encountered by railways. Researchers also developed methodologies for mapping and characterizing ground hazards and their trigger thresholds, performed innovative Geographic Information System (GIS) based geospatial analysis, and created decision support tools and ground hazard event detection systems. The technologies that were investigated by our researchers include:
- the use of seismic monitoring for detecting rock fall activity;
- the application of Micro-Electro-Mechanical Systems (MEMS) sensors for monitoring the dynamic and static movement and deformation of railway sub-grade under cyclic train axial loading; and
- the use of Interferometric Synthetic Aperture Radar for the detection of rock slope movement prior to failure.
The discoveries and technical developments were shared with members of the RGHRP, regulatory agencies, and invited guests through publications and annual workshops.
Past Research: Phase II
After Phase I, it became apparent that more research was needed to better understand and assess the magnitude of risks that ground hazards pose to the railway industry. This led to the development of new instrumentation for field investigation and a fundamental examination of several ground hazard sites. Phase II’s research included the development of risk assessment methodologies, the application of risk to railway ground hazards, and the development and adaptation of monitoring technology for ground hazards.
Current Research: Phase III
Building on the research of Phase I and Phase II, the specific research objectives of Phase III are to:
- Quantify the economic risk associated with ground hazards in terms of economic impact on national and industry levels
- Investigate the impact of an increase in frequency and severity of extreme weather events
- Develop change detection and hazard monitoring technology for larger scale applications, and
- Evaluate the efforts being made in risk mitigation and remediation efforts being implemented as a result of the RGHRP’s continued research.
The results of this research will inform and contribute to necessary improvements to infrastructure over the next couple of decades. This will increase the reliability and resiliency of Canada’s railway networks and enable industry operations/infrastructure to adapt to new rail traffic requirements and the increasing frequency of extreme weather events. The dissemination of results to the wider technical community and the training of HQP to solve real world engineering problems are essential elements of this program.
(1) Economic Risk Quantification
Key Research Focus
- Evaluate the Transport Canada (TC) database of freight movement within the Canadian rail transportation system,
- Quantify the effects of known major track outages (loss of capacity and time), and
- Adopt an economic metric and map distribution system for track outages caused by ground hazards.
The reduction of the resiliency, or the ability of the railway network to move backlogged freight after the interruption of traffic flow, has increased the sensitivity of railway operations to track outages. This has led to an increase in the economic impacts of ground hazard events. To achieve a comprehensive risk analysis methodology that initiates real investment to mitigate the impact of ground hazards on the operations of the railways, the economic risks of the larger scale ground hazards need to be considered. TC has developed a database of the movements of intermodal containers through the transportation system. CN, CPR and TC have agreed to grant the UofA access to this database. The research focuses on the quantification of the effect of track outages on the wider railway network and adopts an economic metric for the consequences of ground hazards. Sections of CPR and CN railway will be selected for risk assessment based on the ground hazards present along its alignment. This risk analysis method will be tested for short sections of railway and for subdivision-long (hundreds of kilometers) sections of railway. The length of the section will determine the level of detail of the risk assessment. The final output will include the methodology and example applications. The risk levels and assessments will be presented as thematic maps generated with the aid of GIS to increase the accessibility of the results and to aid in decision-making. This method, in turn, can be used as a model for other situations requiring risk analysis such as highways and pipelines.
(2) Extreme Weather Events
Key Research Focus
- Conduct an inventory of ground hazards and earthwork failures triggered by the 2013 southern Alberta floods and their impact on the CPR in order to conduct a failure mode and effect analysis, and
- Evaluate ground hazard conditions that railway structures were designed to resist. Develop thematic maps of the susceptibility of railway infrastructure to failure during severe weather. Assess the resiliency of the railway network to severe weather events.
Recently, the impact of ground hazards has been associated with severe weather events that trigger multiple ground hazards. The 2013 Alberta floods are a prime example of this. CPR’s mainline track runs from Calgary through Banff. The flooding triggered a multitude of ground hazard events that directly impacted the rail line, including debris flows, slope failures and washouts of track. Individually, these ground hazards would have been managed with minimal impact to operations. Collectively, these ground hazards presented a crisis.
Several major infrastructure refurbishments and upgrades had been conducted within the area over the past several decades. At many locations, the capacity of the infrastructure to handle water had declined relative to the increasing observed and predicted maximum flow rates. In addition, the flooding exceeded historical precedents and the predicted 100-year maximums. This has led to the re-evaluation of the water management capacity of the infrastructure and the vulnerability of the railway network.
In the hazard scenarios described above, the initiating event is usually an intense rainstorm. Bunce et al. (2003) reviewed the records compiled over the past 122 years by CPR and concluded that at least 30% of the larger volume hazards occurred synchronous with severe weather events. Severe weather events are defined as climatic conditions including antecedent conditions that develop over months or years that have a return period of at least 10 years. To reduce the impact of these events on the safe and reliable railway operation, the railways retain the service of a weather-information provider. This information comes in three forms: (1) synoptic reporting of current conditions, (2) forecasts of predicted conditions, and (3) Weather warnings (as defined by Environment Canada and the National Weather Service (NWS) in the US).
Class 1 railways in North America use weather information in numerous ways. Temperature information is used to limit train speed during severe cold and hot conditions to reduce the potential for broken rail and rail sun-kinks, respectively. The track, bridge and structures maintenance personnel also use weather information, such as rainfall amounts, to predict increased potential for damage to the track and structures. Heavy rainfall forecasts frequently prompt increased track and structure inspections. In rare cases, trains have been slowed over tens-of-miles of track in response to high rainfall conditions and forecasts. According to Bunce et al. (2003), the use of weather is non-systematic and empirical. While maintenance personnel are empowered to invoke multi-mile slow orders, they often need additional climatic information to support their observations and cause for concern. Furthermore, maintenance personnel are not equipped with any means of quantifying or communicating the hazard level. As a result, these severe climatic events are not always responded to in a proactive manner.
The current research is aimed at developing a forecasting methodology that could evaluate the impact of a severe weather event on the track infrastructure such that a proactive management strategy for rainfall-induced hazards can be applied. The methodology will consider, as a minimum, the weather event, a simple failure mode and effect analysis, and the processes that can impact the railroad safety.
Although the specific effects of climate change on climatic events may be poorly quantified at this time, it is recognized that climate change will likely affect the return-period for severe weather events. Higher frequency and increased severity of extreme weather events, predicted by some as being a product of climate change, will test the robustness of the railway infrastructure.
(3) Change Detection and Hazard Identification
Key Research Focus
- Review the space borne, aerial and terrestrial remote sensing techniques’ ability to detect changes on complex slope in order to develop and apply the best method for identifying, characterizing, rating and monitoring complex slopes,
- Investigate ground hazard conditions for various geological areas in eastern and western Canada, then analyze the factors for hazardous rock slope failures and the influence of predictable weather-climate cycles,
- Verify and validate the natural slope monitoring system based on the methods developed by high quality personnel (HQP), and evaluate the applications, impacts and improvements to the railway network,
- Validate the use of remotely sensed aerial slope data as input for a formal system using quantitative change detection methods. Identify the optimal applications of remote sensing/change detection techniques for the railway network, and
- Investigate optimal applications for remote sensing/change detection techniques for different slope failure models. Evaluate failure modelling approaches to support the rating/monitoring of natural slopes.
There are thousands of kilometers of railway lines in high relief terrain. The steep valleys and geology of these regions result in complex unstable earth and rock slopes and areas prone to rockfalls (Peckover, 1975; Peckover & Kerr, 1977; Piteau, 1977). During Phase II, pilot studies for the use of LiDAR, InSAR and photogrammetric technologies were conducted on several known hazardous areas in northern Ontario and B.C. The use of LiDAR and photogrammetric analysis within the White Canyon (Mile 93-95 on CN’s Ashcroft Subdivision; Fig. 2) and at Coldwell (Mile 71-72 on CPR’s Heron Bay Subdivision) was very successful in demonstrating the utility of these technologies for monitoring changes in the terrain. The analysis of sequential 3D LiDAR and photogrammetric slope models has recorded several distinct failure processes, delineated incipient failures and has identified several locations at which failure appears imminent. These pilot projects have developed the workflow and expertise to conduct this monitoring (Lato et al., 2009, 2012).
The investigations conducted in Phase II and the regular monitoring conducted by CN and CPR has been focused on known ground hazards. There have been several recent ground hazard activations at locations not previously known to be hazardous. A prime example is the sudden failure of a rock slope in November 2012 at Mile 109.4 on CN’s Ashcroft Subdivision. This failure closed the mainline track and required a great effort to clear and rebuild. A warning of where these failures are likely to occur, and monitoring for changes that may precede failure, would greatly assist in planning for the reduction of risk and the impact on railway operations. Formal and systematic rock slope hazard rating systems are in use by Canadian railways: these are limited in application to cut slopes which can be inspected from track-level. Bi-annual helicopter inspections of the whole of the railway network are required; however, these are visual inspections and are limited in their ability to detect and quantify new hazards, being particularly dependent on the inspectors’ memory and attention to detail. The wide application of aerial- or space-based change detection technologies has the potential to detect and monitor ground hazards not visible during routine inspections. These technologies and methods show promise for inclusion in railway rock engineering and slope assessments, particularly for inaccessible locations.
The current research will further develop, test and optimize the methodology for including remotely sensed data into a formal system to identify, characterize and monitor hazardous natural slopes. This involves an expansion of our understanding of how best to apply space borne, aerial and terrestrial remote sensing techniques to complex slopes, and various ground hazard types, for the purposes of identifying, characterizing, rating and monitoring hazardous locations. The development of the technology will continue at the White Canyon as it provides a safe vantage point for the monitoring, the slope reliably provides failures, and it has been well characterized during Phase II. The optimization of the methodology focuses on identifying the characteristics of each failure modes and the effect of the complex nature of the slopes.
(4) Effectiveness of Risk Remediation and Mitigation
Key Research Focus
- Review remediation methods for stabilizing and reinforcing embankments constructed on soft soils. Analyze data sets resulting from the geotechnical instrumentation to determine the effectiveness of piles,
- Correlate rates of rockfalls to the rates of precipitation aggregated on monthly and yearly time scales. Evaluate the effect of large scale weather events on the rate of rockfall occurrence,
- Synthesize findings from Thompson River valley studies. Conduct spatial analysis of hazard posed by the slopes. Estimate the range and spatial distribution of hazard in terms of economic impact,
- Supervise field programs and provide addition supervision of HQP. Evaluate the effectiveness of inspection methods currently implemented on tracks. Develop an understanding of the current level of risks accepted by the railway industry,
- Evaluate triggering factors for large scale rock slope failures. Determine how predictable these events are and on what scale, and
- Develop a model for an early warning system based on the identified precursor conditions. Validate the model using historical and real time weather events and recorded rockfalls.
The understanding and quantification of risks is only useful if it helps determine how best to reduce these risks. The development and application of new hazard monitoring technologies and remediation methods through the RGHRP has been very successful. New techniques have been identified and tested in both the laboratory and in the field. Regulatory partners have provided the context for broader scale implementation.
(4.1) Thompson River Landslides
In southern B.C., both the main CN and CPR lines run along the lower valley slopes of the Thompson and Fraser Rivers. Up to 80 trains per day, some with lengths up to 4 km, run through these valleys. Combined, these railway lines form the major transportation corridor for freight between Canada and its busiest port and gateway to the Pacific, the Port of Vancouver. The Thompson River valley between Ashcroft and Spence’s Bridge has been the site of several large landslides, and at present, there are 14 landslides that have been identified within this corridor, 12 of which are traversed by railway tracks.
Early failures of the slopes were first noted in the 1860s and reported by Stanton (1898) with periodic (re)activations since then. These landslides have the potential to sever the railway lines and greatly reduce the flow of freight from the Port of Vancouver, resulting in economic losses that grow exponentially with the duration of service interruption (Bunce & Chadwick, 2012).
The research conducted by the RGHRP has resulted in a clearer understanding of the causes and risks associated with the landslides along the Thompson River valley. The surfaces of rupture of the landslides through this section of the Thompson River valley appear to occur along a common weak layer of clay located below the bottom of the river and develop as compound slides (Eshraghian et al., 2007). The river level has been shown to control the rate of movement. Eshraghian et al. (2007) showed that decreasing river elevations after unusually long higher river levels was a trigger for the slides.
The most recently activated landslide is the Ripley Slide, which is currently moving slowly and is traversed by the mainlines of both CN and CPR. This slide presented the opportunity to study the movements of an active landslide within this strategic and dynamic area and to test the technologies used to mitigate risk and serve as an early warning system (Bunce & Chadwick, 2012; Hendry et al., in press). As a recent initiative of the RGHRP, this landslide has become a test bed for monitoring methods. These methods include both near and real-time GPS; subsurface geotechnical (SAA and piezometers) and fibre optic monitoring systems; and trials of LiDAR, photogrammetry and InSAR for change detection (Bobrowsky et al., 2014; Huntley et al., 2014).
The research will (1) continue the evaluation and comparison of the landslide monitoring technologies installed at the Ripley slide; (2) use the resulting movements of the slide to evaluate the stability of this section of the Thompson River valley; (3) combined with an economic-based risk assessment, evaluate the economic feasibility of the large-scale remediation of the known active landslides based on the risk and economic impact of ground hazards on railway operations.
(4.2) Rock Fall Detection and Risk Analysis
Rock falls are frequent hazards for transportation corridors through mountainous areas. Railways across the mountainous regions of western Canada and in the Canadian Shield in eastern Canada regularly deal with rock fall events. Many areas with frequent rock fall events are protected by fences, which provide warnings for such events. Other sections of track have very dispersed rock falls such that they need to be patrolled by railway employees to ensure the safety of the track. These hazard events have small volumes relative to other slope failures; however, their frequency and rapidity may lead to derailments if not detected. Thus, rock falls pose a threat to the safety of railway personnel and to the ecosystems of the rivers that the tracks typically follow through these regions.
The research conducted in Phases I & II made advances in the 3D modeling of rock falls (Lan et al., 2007; Ondercin et al., 2014), the kinematics of rock fall slopes and talus movements (Gauthier et al., 2012, 2013; Kromer et al., 2014) and the potential of the use of change detection and monitoring for hazard assessment and risk management (Hutchinson et al., 2013). Further, a close relationship between rock fall occurrences and weather has been developed. Macciotta et al. (2013) showed strong correlations between rock fall occurrences and precipitation, and between rock fall occurrences and freeze-thaw cycles on the first 60 km of CPR’s Cascade subdivision.
The research project builds upon analyses conducted in Phases I and II, and further develops an understanding of the various failure modes associated with the complex slope and rock conditions found along the railway corridors and the triggers for these hazards. The research focuses on precursor displacements and developing correlations between the temporal distributions of rock falls to climatic conditions (such as precipitation, freeze-thaw cycles, snow accumulation for both the monthly average distribution and the hydrological year totals) as well as the effect of large-scale weather events such as El Niño or La Niña. This research is being conducted on CPR’s Cascade subdivision, CN’s Squamish Subdivision between North Vancouver and Lillooet, B.C., and the Ashcroft Subdivision through the White Canyon. The expected results will be threshold values for displacement, precipitation, freeze-thaw cycles, etc. that are usually exceeded before the occurrence of a rock fall event and corresponding measure of confidence that rock falls will occur once thresholds have been exceeded.
This research is anticipated to provide the basis for the development of an early warning system.
(4.3) Remediation of Embankments over Muskeg Foundations
Long sections of Canadian railways are built over very weak organic soils (muskeg bogs). Many of the existing embankments were constructed more than a century ago with the standards and methods of that time. Since 1996, trainloads have doubled and the length of trains has increased up to 4 km. Larger cyclic strains are produced within the peat due to train loading. The large concentrated loads associated with trains on railway embankments are particularly destructive to rails, ballast and tie structures due to the wear produced by large cyclic movements of the embankment and foundation. These cyclic strains can also lead to embankment failure after years of normal operation.
Hendry et al.’s (2011) study of a previously remediated embankment has shown potential benefits of using piles to stiffen the structures and to reduce the stresses on the muskeg foundation. CN is currently moving forward with the use of piles to remediate several problematic sites. Instrumentation was installed at CN’s Lévis subdivision to evaluate the effectiveness of the piles in increasing the strength of muskeg foundations under heavy axle loading. Instrumentation was installed in November 2012 to measure the response of the muskeg subgrade to train loading before the installation of piles. Helical piles were added in November 2013 , and were instrumented with strain gauges to measure the amount of load that they are bearing.
The researchers will analyze the data collected from the Lévis site in order to determine the effectiveness of the piles at reducing the stresses on the muskeg foundation and at reducing the impact of the soft subgrade on the rail-track structure.
(1) Ground Hazard Risk Identification and Analysis
In order to properly assess and reduce the risks posed by ground hazards, the railways must be able to track the hazard and the associated risk. This can only be accomplished if the ground hazards are classified and described using a consistent technical process. The purpose of this work was to establish Ground Hazard Terminology and Scenarios for documenting and reporting ground hazards and ground hazard events. The results to date have led to the development of a detailed classification system for all railway related ground hazards base on mechanism and triggers. Ongoing work includes the development of an industry standard for ground hazard reporting and documentation.
(2) Ground Hazard Event Triggers
In order to properly assess and reduce the risks posed by ground hazards, the railways must also understand the conditions that trigger ground hazard events. Ground hazards can be triggered by precipitation, seismic events, freeze-thaw cycles, river scour and erosion, rapid snowmelt, and dynamic loading conditions as trainloads increase. To date, an evaluation of the environmental factors that influence risk have been conducted, drawing correlations between meteorological and ground hazard events from 122 years of records.
(3) Technology for the Monitoring & Evaluation of Ground Hazards
The purpose of this project was to evaluate new and emerging technologies for the monitoring of ground hazards, such as large slopes and rock faces, which may pose threats to the railway. Several technologies were evaluated, including:
- Static laser scanning (LiDAR) for the analysis of rock faces and rock slopes to assess stability and associated risk. This included the development of progressive scans to measure movement and changes in surfaces due to rock falls.
- Terrestrial-based InSAR (satellite radar interferometry) for the stability of large rock slopes and cuts along railway corridors. This technology can measure millimeter movements from distances of over a kilometer.
- A comparison of several rockslide monitoring technologies at a large-scale instrumentation installation where the CN line crosses the large active Gasçons rockslide on the Gaspé Peninsula.
- Ground Penetrating Radar (GPR) to evaluate soft soil foundations to determine mechanisms resulting in track alignment degradation.
(4) Rock Fall Detection
The current method for monitoring rock fall events was with a trip wire system. While this system was reliable in detecting rockfalls, it was prone to false alarms and required time to reset once triggered, during which trains must run at very low speeds. The purpose of this work was to develop a seismic rock fall detector that could replace existing trip-wire detectors (or rock slide fences) to decrease the time required to reset systems and return to normal train speeds and line capacity. The system was tested for reliability and ability to differentiate quantity and proximity of fallen rocks to rails. This work led to improvements in train reliability, safety, and on-time performance.
(5) Heavy Axle Loading over Soft Subgrades
Both CN and CPR experience continuing problems with large stretches of embankments built over peat. These problems range from excessive settlements requiring increased amounts of maintenance to sudden failures of these embankments.
The purpose of this on-going work is to study the impact of heavy axle loading on aging track infrastructure and the soft peat foundations under these structures. To date, several sites have been investigated and extensively instrumented. This instrumentation has included the development of new Micro Electro Mechanical Systems (MEMS) base technology to measure cyclic displacement of soil with depth during the passage of trains.
(6) Risk Mapping of Sensitive Clays
A large portion of the railway lines in Eastern Canada are built over thick deposits of soft sensitive clays and are vulnerable to sudden landslides. The purpose of this work was to use new understanding of sensitive landslides and new mapping and monitoring technology to pinpoint areas of concern along tracks in relation to sensitive clays. The results of this work included comprehensive maps and guidelines to help prevent activation of disastrous slides. Operators were provided with developed maps and guidelines to reduce hazard levels which can also be used to protect other types of transportation corridors (pipelines, roads, etc).
(7) Ballast Degradation
Degradation of ballast (the layer of crushed rock or gravel upon which railway track is laid) is a process by which ballast is filled with fine-grained material. The result of this is a reduction in the ability to drain and a reduction of shear strength. The purpose of this work was to develop a field sampling methodology and analysis of the mechanism and source for fouling. In addition to sampling, the project involved: evaluating the use of Ground Penetrating Radar (GPR) for determining the degree of ballast degradation on sites; evaluating the effectiveness undercutting as a ballast remediation technique; and evaluating the effectiveness of mitigation techniques such as drainage and reinforcement on degradation mechanisms.