Thread by @25kV

THREAD: Every so often during a discussion about electrification at bridges, the subject of COASTING comes up. I denigrate the idea without explaining why, then move on. I've been meaning to do better than that for a while: so here it is, buckle up!

1/

Let's start by defining coasting as follows:

Coasting is when an electric train passes through a section of railway without taking power, using only its own inertia to make it through the gap.

2/

Things that AREN'T coasting:
1) a bi-mode train switching from electric to diesel
2) A train switching to batteries that were charged while the train was under OLE
3) Hydrogen or other vapourware

3/

I'm also excluding low-speed systems such as electric charging buses and trams - this is about mainlines.

A bridge recon typically costs ~£1-£5m depending on complexity, and these route clearance costs have traditionally made up 1/3 of the total cost of electrifying a route.

4/

Track lowering is the other option, but this makes bridge reconstruction look cheap: think £5-10m, even more for a long tunnel. So well-meaning people faced with these costs often reach for simple solutions... like coasting.

5/

The thought process goes like this: why electrify the hard bits? Why not just electrify the easy, low cost bits and coast through the hard bits? This concept is known as discontinuous electrification.

6/

One simple test that you should always apply to an idea is: who is already doing it, and how are they getting on? So lets look at existing discontinuous electrified railways...

7/

The reality is that no-one AFAIK is operating a mainline electrified railway with multiple discontinuities. There are some railways which will operate with a single discontinuity, but these are unusual. They also tend to be lower speed.

8/

So why is that? Why do railway owners, when faced with very high route clearance costs, spend the money and avoid having any gaps in their OLE? Well it is partly to do with operational constraints, and partly engineering constraints.

9/

Let's start with a pretty obvious point: an electric train will come to rest if it spends too long without power. How long depends on simple physics: how much inertia does it have, and how quickly will aerodynamic drag, wheel-rail friction and gravity bring it to rest.

10/

Second obvious point: if a train comes to rest in a coasting section, it cannot move. Your railway is now closed until further notice. Not good. So you can't have coasting through a station or at a signal which can display a red aspect. That rules out a LOT of railway.

11/

At this point you might be thinking "how about using gravity to get you out of trouble? Have your coasting section on a gradient. Take off the brakes and roll out."

12/

Unfortunately railways operate in two directions. If you still need to electrify in the uphill direction, you still need to rebuild the problem bridge. So you might as well electrify both lines.

13/

But lets assume that you've found a location where you think you can coast without significant risk of coming to rest. Hurrah! Take your £2m bridge recon cost and spend it on party hats.

Not so fast roadrunner...

14/

Before you can start coasting you need to lower the pantograph. Easy enough. But what happens if the driver forgets? One smashed pan, one smashed bridge, one disrupted railway.

15/

We don't tolerate single point human error failures like this - we build systems that prevent the mistake from having a consequence. In the UK this means trackside balises that lower the pan automatically. This process is known as Automatic Power Changeover - APCo.

16/

The process is complex. A control balise tells the train to lower the pantograph; but as a failsafe a further "zero" balise is placed beyond the lower zone as the last line of defence; it notifies the driver to lower the pan or stop the train.

17/

If the pantograph fails to lower, the driver must bring the train to a stand to avoid damage. So the zero balise must be at least the braking distance before the OLE end.

The cumulative effect of this whole process is that at 125mph the changeover zone is ~ 4km long.

18/

With all this going on, the changeover zone must be carefully selected so that the driver is not overloaded with other information or duties. So no signals, stations or junctions.

19/

This can push the changeover zone a long way back from the actual end of OLE - for a while the GW one was East of Didcot, several miles from the actual OLE end which was West of Didcot; because the junctions and station would have resulted in driver overload. 20/

Of course on the other side of the gap you need to raise the pan again - you can't do this just anywhere, you need to select a spot where the OLE is boring so dewirement risk is minimised. So again, no junctions or stations.

21/

But lets assume that you've found a location where you think you can fit all of this in. Hurrah! Time to take your £2m bridge recon cost and spend it on party hats.

Actually... we're still not done

22/

Remember OLE does two jobs: power transmission and sliding contact. You no longer need the second part through the bridge, but you still have to transmit power to the other side of the gap. But you have no OLE to do that. What do you do?

23/

Well, you need to run HV insulated cables in trough route along each side of the railway, and connect them to the OLE at each end of the gap. This isn't particularly cheap: I've heard figures of £500 per metre quoted.

24/

All of this equipment - APCo, balises, power feeds - will cumulatively erode the savings that you make by not rebuilding the bridge. But lets pull back from the detail and look at it from a much simpler point of view.

25/

Coasting is appealing - it looks like a great money-saving idea, right up until the first time it doesn't work; then you have a stranded train, a rip down, a pan smashed on a bridge, or all 3.

Your gap will be a permanent operational headache that requires an expensive fix.

26/

That's not to say that coasting is *never* a solution; but I hope everything in this thread has convinced you that it is not a mainstream solution, and it is certainly not practical to use it multiple times on a route.

ENDS/

THREAD EXTENSION: a number of you have raised additional points and questions - so welcome to Episode 2 of:-

COASTING: HONESTLY, DON'T GO THERE

28/

Firstly, a few extra objections to coasting raised by the rail twitterati:

@seb_barrow noted that, every second you coast you are losing time. Rail is supposed to be about shorter journey times, not longer ones

29/

My learned colleague @bowdidgea pointed out that APCo can only be used by ECTS-compatible stock. So you are locking in a specific set of stock, and locking out open-access operators with older cascaded stock and ideas, or even the TOC wanting to beef up service frequency.

30/

The estimable @driverbod125 reinforced this by noting that some electric stock loses auxiliary power when off the wire - so goodbye brake reservoir top-up!

31/

… @clloyd3003 reminded me that GW has a diverging non-electrified route to Oxford. Diverging routes hugely complicate APCo because now the system needs to differentiate between trains to switch off, and trains to leave alone.

32/

How hard can that be, right? Well quite hard when you consider that a bi-mode train could start its journey with Plan A and then be diverted onto the non-wired route mid-journey due to disruption ahead. It's an operational and data nightmare.

33/

... @PeteWilson1967 asked why not have pans at either end of the train so one is always on the wire? Well aside from the fact that this train would have to be veeeeery loooooong, once again you are locking in stock. Railways grow and change over time - OLE lasts 80 years.

34/

... @srpnor asked - I *think* in all seriousness - why not just use 3rd rail. Suffice to say that:
a) it's bloody dangerous
b) it doesn't deliver enough power,
c) it won't go at 125mph and
d) your rolling stock becomes heavy and complex

35/

Many many of you asked about earthing the wire through the bridge. Claims were made that this would mean a much lower bridge height. This needs serious examination, as it is a widely-held view that this is a solution.

36/

We need to start by examining the smallest air gap that allows live OLE to be squeezed through. GL/RT1210 specifies 270mm above live parts (to the bridge) & 200mm below (to the train). But it allows you to go below those values as long as you undertake a risk assessment.

37/

The lowest you would normally go without special measures would probably be the old "Reduced Clearance" level, which is 200 above/150 below. So 350mm of air in total.

38/

Now lets look at an earthed wire. You no longer need an electrical clearance air gap; so is it OK for the wire to touch the train or the bridge? No of course not. You still need a mechanical clearance.

39/

As the train passes through the bridge it will still push the wire up. Allow 70mm for that, plus 50 mechanical air gap. Below the wire, I wouldn’t imagine the gauging engineer would be happy with less than 50. So 180 total.

So our earthed wire saved us 180mm. Hurrah!

40/

Now don’t get me wrong; 180mm is sometimes the difference between recon and no recon. But we havent yet understood how we get that wire to be earthed.

You can't just pass a pantograph from live to earth - you will get an arc drawn, just like in this video.

41/

So you need a neutral section either side of the bridge. Neutral sections are a pain to set up/maintain, have their own signal restrictions, and are a known weakness & failure point in OLE.

You also need the HV cabling we mentioned way back in /24 to get the power through.

42/

Neutral sections are sometimes used - as @sween5ter points out, there are 4 in Scotland - but you certainly wouldn't do it lightly.

In any case this is all moot, as there is now a much more attractive option that allows you to keep the wire live & have v small clearances.

43/

I discuss that option - surge arresters and insulated paint - below. It is under trial at present but allows a total clearance almost the same as an earthed wire, with none of the risks of a neutral section

https://twitter.com/25kV/status/1169698106268684288?s=19

44/

I'll finish this thread extension with two excellent comments:

""The West London line was a fine example… when you changed from overhead to third rail driving south there was an indent the size of the pantograph in the Westway flyover."" - @waterbillway, train driver

45/

"If wouldn’t consider switching your engine off at 70mph on the M6 then why would you specify a railway to do exactly that?"" - @don_dapper

/ENDS