Preventing disruption from autumn leaves

Leaves, particularly in wet conditions, can be compressed onto the railhead by the weight of train wheels. In turn, this can reduce adhesion between the wheel and the rail and may also cause poor electrical contact between them. This leaf ‘mulch’ is to rail what ice is to roads.

These innocuous-seeming leaves can cause huge disruption, not only because they make rails slippery—negatively affecting train braking—but also because their presence may mean that track circuits fail to detect the presence of a train. If a wrong-side track circuit failure (WSTCF) occurs, there is a risk of collisions or failure of trains to initiate the operation of level crossings. As a minimum, it can cause delays and frustration for passengers and train operators.

Working with the Adhesion Research Group and Seasonal Challenge Communications Group, we have identified inconsistencies between routes in how track circuits are managed and how well they perform. Our research project ‘Understanding how railhead and wheel contamination affects track circuit performance’ (T1222), which ran from 2022 to 2023, has provided greater insight into the effects of railhead and wheel contamination on the reliability of track circuits.

Findings and what they mean

This project conducted track trials and laboratory tests—as well as evaluating national datasets and reviewing case studies on a number of GB lines—and has revealed a number of meaningful findings.

Specifically, track trials at Wensleydale Railway revealed that:

• Leaf layers applied to the railhead produced a consistent and predictable increase in voltage and resistance. 
• When strongly bonded to the railhead, the leaf layer persistently maintained high voltage levels between each pass of a train. This could lead to a failure of train detection in high-voltage cases.
• Sand applied via sanders created erratic, high spikes in voltage but was removed the first time a train passed. This would be expected to generate a short ‘flickering’ train detection failure.
• Sand applied on top of a leaf-contaminated railhead caused erratic, high-voltage measurements that persisted with the next pass of a train, which would be expected to cause the most issues with train detection.
• The inductance of the relay coil of the track circuit is a key component in causing large voltage spikes. Changing this component, perhaps to a solid-state relay, could reduce the impact of some train detection losses.

In laboratory tests, leaf contamination resulted in elevated contact resistances until/unless it disintegrated under the application of sheer force (i.e., the weight of the train). On the other hand, increased surface roughness decreased electrical resistance in the presence of leaf contamination.

Furthermore, resistance created by sand depended on the amount of normal load being applied. At lower loads, the sand produced high electrical resistance, but at higher loads, the sand particles broke down, resulting in lower resistance and improved electrical conductivity at the contact.

Meanwhile, rust contamination was removed very quickly upon the application of load, and it only created isolation upon the first contact.

The track trials and laboratory tests successfully created contaminant layers that created high electrical resistance in the contact. According to the final report, trials with leaves, sand, and rust characterised the materials as ‘predominantly ohmic conductors’.

It was also observed that wheel load/axle load had an important effect—specifically, the lighter the load, the more detection issues there were. This knowledge could be used in the future to assess the risk of WSTCFs, especially when routing light or empty vehicles through areas of known contamination.

The factors outlined above could inform possible WSTCF risk areas. They could also help industry assess lines on which track circuit issues may occur if, for example, traffic loads change, vegetation is not adequately managed, or railhead treatment schedules change.

Action points for industry to consider

The project also produced two good practice guides containing a number of key action points, which include:

• removing vulnerable DC track circuits (where practicable)
• undertaking regular remote condition monitoring (where available) to allow for the prediction of future WSTCFs
• identifying rarely used lines, where rust may be more likely to form, and taking steps to mitigate related WSTCFs
• considering the installation of axle counters (where appropriate) where contamination-based WSTCFs are a persistent issue
• investigating and utilising new technology for visual inspection (such as drones), which provides the potential to undertake inspections while keeping personnel off the track
• undertaking further engagement across industry in order to standardise reporting of WSTCFs, which will facilitate future data analysis (on a national scale) and learning from incidents.

What’s next?

To build upon these findings in the future, the research team proposes:

• simulating audio frequency and impulse track circuits in the laboratory
• looking at the effect of multiple (i.e., more than two) train passes on different railhead contaminants
• determining what’s driving the electrical breakdown of the leaf layer
• running the same sorts of tests on different track circuit types and varied voltages.

Ultimately, the hope is to be able to use these findings to more fully understand what causes track circuit issues and in what situations they are most likely to occur. Knowing this may help industry predict WSTCFs and take steps to mitigate them before they cause disruption.

Watch our video summarising the findings of the project below.