Wet rail monitoring, moisture sensing and adhesion hot spots
Building on research conducted in 2014, which collated existing knowledge of the effects of moisture on rail adhesion and developed a better understanding of what has come to be known as 'wet rail’ phenomenon, we are exploring a range of options for improving wet rail monitoring, moisture sensing and our understanding of adhesion hot spots.
We have shown that an “Internet of Things” approach using low-cost self-contained moisture sensors can be used to measure the presence of railhead moisture. These sensors, which have been installed at vulnerable Transport for London and London Underground sites, take readings every 90 seconds and relay the data directly to London Underground’s adhesion management software. Networks of the sensors are now being trialled and are expected to demonstrate the operational advantages of an integrated approach to moisture sensing.
Anecdotal evidence suggests that conditions at sites that are considered high-risk – and which are consequently the focus of rail cleaning activities, lineside vegetation management and driver briefings – are not always as severe as expected and conversely there may be high-risk sites which have yet to be identified.
This autumn RSSB, in partnership with West Midland Trains, the Network Rail London North Western Route and the University of Birmingham, will be implementing moisture sensors across the Birmingham cross-city route. The data we collect will enable us to determine whether varying levels of moisture on the railhead can be correlated against sites that are typically considered low adhesion hotspots. This data, in conjunction with other data collected on the same route, including adhesion forecasts and adhesion levels experienced in braking by in service trains, will also allow further validation of the WILAC and LABRADOR models and support our work on quantifying the effects of railhead treatments on adhesion.
- Internet of Things
- Moisture sensors
- Wet rail monitoring
- Adhesion forecasting models
- Low adhesion hotspots