Rail temperature decides whether trains slow down in summer, so it is measured by probes clamped to the rail and reported to the control room. The probe at one site in a north facing cutting reads eight degrees above the two either side of it, at night, in December, with no sun on it since October.
I look after the remote condition monitoring on a route, and the rail temperature probes are the simplest thing I have.
How they work.
A probe is clamped to the web of the rail, in the shade of the head, wired to a logger in a lineside box that reports every ten minutes. There are about forty on my route, placed where the track is most at risk from heat: long straights in the sun, and anywhere the ballast is thin.
The number they report is used. Above a threshold in summer, speeds come down on that section, and the threshold is not a guess, it is calculated from the rail's stress free temperature.
They are checked annually against a calibrated contact thermometer and they read within a degree.
The probe at one site, in a deep north facing cutting, reads about eight degrees above the two either side of it.
Not in summer, when a cutting with no sun would be cooler, which is what you would expect and what the other two show.
All year. At night. In December, when the rail either side is at minus one and this one reads seven.
A cutting with no sun on the rail from October to March has no source of heat for a rail. The air is the same air and it is colder in the bottom of a cutting than on the bank.
What has been checked.
The probe, replaced twice, and once with a probe from a different manufacturer with a different sensing element.
The logger and its wiring, replaced when the site was upgraded in 2021.
The rail. A thermal camera from the cess in the dark shows the rail as a uniform cold line along the cutting, and then, for about forty metres either side of the probe, a lighter one. Not at the probe. The rail. Eight degrees, along forty metres, on a rail that has nothing near it but ballast and the cutting sides.
The obvious causes: a traction return current heating the rail, a cable fault, a buried heat source. The traction return is measured and it is balanced. There are no cables under that rail. The cutting is cut through clay and there is nothing in it.
And by hand, with a contact thermometer, which I have done on eleven nights.
It reads eight over. The instrument is right.
What that means for the track, because it is the only reason anybody funds looking at it.
A rail eight degrees warmer than its neighbours has eight degrees more expansion in it, and that is a real force in a continuously welded rail. It is inside the design margin. The stress free temperature at that site was re-measured in 2022 and the rail is not in distress.
Nobody's safety turns on it. A rail is warm for no reason, in a cutting, at night, and that is the whole of the finding.
What is done about it.
The probe's output is excluded from the automatic speed restriction logic for that section, which is a change I asked for and which was approved, because in a heatwave the eight degree offset would restrict trains a section early. The two neighbours cover it.
Excluded from the automatic logic, not switched off. I still see it.
What I do.
I read the forty probes every morning and the one in the cutting is the one I look at first.
The one thing I changed is how the annual calibration is done at that site. The procedure is a contact thermometer on the rail beside the probe, and the rail is at head height in a cutting, and you do it on your knees in the cess.
I do it standing on the bank above the cutting now, with a thermometer on a pole, which is not in the procedure, and I wrote the procedure.
The reason is a night in January 2023 when I was kneeling by the probe with my hand on the rail, which is a thing you do to steady yourself, and the rail was warm under my palm the way the log said it was.
It was warmer towards the probe and cooler away from it, along forty metres, and the warmest point sat about two metres down the rail from the probe rather than at it, and when I moved my hand there the steel was the temperature of a hand.