There’s been a big thread going on about whether 18 trains per hour (tph) is feasible on a high speed railway, and specifically HS2. Sceptics demand to know whether any other high speed railway can do 18 tph, and refuse to accept that HS2 could until another railway does.

Two problems with that: 1) If no-one ever did anything until someone else had, we wouldn’t even be living in caves or using fire (‘PROVE it keeps the sabre-tooths away’) and 2) Precedent is not proof anyway as all railways are different.

They differ in what their binding constraint on capacity is – is it plain line headway, or intermediate stations, or termini, or do they simply run no more trains than their demand requires? No, I don’t know of any high speed railway that runs 18 tph.

And if ‘someone else does it’ is the only proof you are going to recognize, stop reading now and do something else. Just look out for sabre-tooths while you’re doing it. But if you have that old fashioned thing, an open mind, …

… let’s have a think about how we know what HS2 can do. Well, precedent can help. 18 tph means running at 3-minute intervals (with 6 minutes per hour to spare) and there’s plenty of high speed lines that do that.

HS2 just has to do it 18 times per hour instead of once or twice. And if you can do it once, the technical capability must be there. But can you do it on HS2, even once, at 360 kph? There’s two ways of finding out – calculation and modelling.

With reasonable inputs, calculation shows the technical headway can be as little as 120 seconds at 360 kph. It’s surprising how low technical headways can be – watch a train fly through Milton Keynes at 125 mph, and the signal is green for the next within 90 seconds.

Of course at higher speeds, train separation distance has to be longer as braking distance is ~l to speed^2, but time taken to run through it only ~l to speed, so time separation is greater. But 120 seconds is plausible. See IRSE News, May 2019.

But to be sure, model it. Now, if anything is my specialist subject, it’s operations modelling. For 30 years ago, I’ve commissioned it and cursed the supplier, supplied it and cursed the client, specified it, developed it and interpreted it.

And in the model, you put the data on the actual railway, HS2 – speed limits, gradients, signalling and its trains. And 120 seconds is again the answer, albeit extending a bit for odd features such as neutral sections (sorry @25kV) and on sustained falling gradients.

And on plain line in open air, even at 360 kph, headway just isn’t a problem. 120 seconds, give or take, supports a timetabled interval of 3 minutes just fine. BUT in slow speed areas where section lengths are dictated by features such as tunnel shafts …

… it’s not so easy. Even so, at the worst places, usage of capacity derived from technical headway * tph falls within the Union Internationale des Chemins de Fer (UIC) guidelines – so long as every sort of engineer works together …

… and if not exactly compromising, does things they wouldn’t necessarily do if left alone. Even operators. Because we aren’t alone, we are part of a system, and if an ideal design of one aspect buggers up the system, we all lose.

So we can plan a timetable based on planning headways of 3-minutes. But can we run a railway like that? What if a train handed over from Network Rail, having started from Liverpool, Glasgow or Newcastle runs late? And the answer is …

… not very much. With 30 – 60 seconds buffer within the planning headway, the consequential delay to other trains decays over the next few trains, they are just running a bit out of order. And the big problem with trains running out of order …

… i.e. fast trains getting stuck behind slow trains (how many times did you hear Virgin announce that? Never mind they were already late which is why they were behind the slow train!) just doesn’t happen on HS2, as - there aren’t any slow trains.

Which is just fine until you reach somewhere where the order of trains does matter, like a flat junction planned around parallel moves. But there aren’t any of those on HS2, except – and this really does matter – THE TERMINUS. Meaning Euston.

And even at Euston, it works. Partly because the track layout includes a grade separation splitting the station into two halves, so that moves in and out of each half can’t conflict. Partly as with 11 platforms, you can plan to work on 10 routinely, leaving the 11th for ‘events’.

So long as that grade-separated, 11 platform station is what is built. So apologies @SirPeterHendy @HuwMerriman @Andrew4Pendle (I’ve only @’d you in this one, read back if you want to) for saying it again, but I’m going to, and I’m not going to stop – do not descope Euston!

Euston is what enables the whole HS2 network to operate as planned, and if you descope Euston you will hamstring the capacity of the whole HS2 network. For ever. Forget plain line in open air, it’s fine. It’s the terminus that is critical. Don’t descope it. Just don’t.