What a map knows that a timetable cannot
Britain's railway is mapped in remarkable detail by volunteers, and that map knows things no timetable does — how a line is electrified, how fast it is, and which quarter of it is not a running line at all.
The timetable knows a train calls at Sheffield. It does not know where Sheffield is, which physical track the train takes to reach it, or whether that track has wires above it. All of that comes from OpenStreetMap, and it is better than you might expect.
Of 92,910 railway ways in Great Britain, 94% carry an electrification tag and 76% carry a line speed. That is unusually complete for volunteer data, and it is what lets a planned route report what kind of railway it runs over rather than just drawing a line.
A quarter of it is not a running line
The single most useful thing the map knows is which track is not for running trains along. About 25% of Britain's railway ways are service track: 8,902 sidings, 6,963 yard lines, 4,177 crossovers and 3,033 spurs.
Ignore that and a route planner will cheerfully send a train through the middle of a depot, because a depot is beautifully connected and often a shortcut. We charge service track six times its real length when planning, which makes it a last resort without making it impossible.
Penalised rather than banned, and deliberately so: a station's own platform road is sometimes tagged this way, and excluding it outright would turn an unusual route into no route. Crossovers are exempt from the penalty entirely, because moving between two adjacent running lines is exactly what they are for.
The check that made us trust this: across 40 test routes, 28 were comparable before and after the penalty, and all 28 came out identical. It changed the odd cases and left the ordinary ones alone.
Overhead or third rail — and at what voltage
Electrification comes in two useful parts. There is the kind — overhead contact line, third rail, occasionally fourth rail — and separately the voltage.
You need both, because the kind alone does not tell you which railway you are on. Britain runs 25 kV AC overhead across most of the electrified network and 750 V DC third rail across the south, and those are different systems with different trains, not variations on a theme.
The Great Western is a good illustration. Its main line reference reports about 515 track-km at 25 kV and another 513 km with no electrification at all — wired as far as Bristol, diesel onward to Plymouth, in a single line reference. A route that quotes one figure for the whole thing would be describing neither half.
We show the voltage exactly as tagged and never convert it to a tidy number, because dual-system track is tagged with both values at once. Parsing that into a single figure would report half a railway as all of it.
Main line, branch line, and the difference from a siding
The map also records what a line is for: main, branch, industrial or tourism. This is a different statement from the service-track one, and conflating them is easy.
A branch line is a through running line. Trains carry passengers along it at normal speeds; it is simply not a main line. A siding is not a running line at all. Our route legs mention a branch or a heritage line when the route uses one and stay quiet about main lines, on the basis that every route is expected to be on one and saying so adds a word to every leg while distinguishing nothing.
Two line speeds, because an average would hide the point
A leg reports the highest line speed along it and the lowest, rather than an average. The pair is more honest: a 125 mph run with a 20 mph junction in the middle averages to something like 110, which describes no part of the journey and hides the only bit that matters.
Line speed is also not train speed. It is the maximum the infrastructure permits, which a particular train may not be capable of and a temporary restriction may override — see temporary speed restrictions for the other half of that.
Gradients, and the statistic we could not publish
Slopes come from a licensed engineering dataset rather than the volunteer map, and they work well for a journey: a route profile shows where a train climbs and descends, and summing the ascent over a run is meaningful.
What we could not do is quote a steepest gradient per line, and the reason is instructive. We built three versions and checked each against gradients the railway is famous for:
- The steepest single stretch made every main line about 1 in 29. On the Great Western that came from one 102-metre section, when only 0.05% of the line reaches even 1 in 40.
- The steepest 500 m of sections wherever they fall gave King's Cross 1 in 33 against a real 1 in 107 — scattered bad samples add up to 500 m quite happily.
- The steepest continuous 500 m was right on two lines we could check and gave 1 in 29 across the flattest part of East Anglia, which is nonsense.
The cause is that elevations are matched to the track within a tolerance, so a bridge, a flyover or a parallel road can lend its height to the line below. Averaged over a whole journey those errors wash out. An extreme statistic seeks them out. So route profiles have gradients and the line reference pages do not, which is a deliberate gap rather than an oversight.
Track kilometres are not route miles
One last figure that reads wrongly unless you know what it counts. When we report the track length of a line reference, that sums every road carrying the code — the up line, the down line, the loops, the sidings. The Great Western main line comes to about 1,028 track-km against a mileposted extent of 244 miles, roughly 2.6 times.
Neither number is wrong and neither is derived from the other. One answers “how much rail is there”, the other “how long is the line”. We label them separately for the same reason a route measures its distance from the drawn geometry rather than from its own routing cost: once a number has been adjusted for any purpose, it stops being a measurement.