The substations cost a small fraction of the cost of 25kV capable substations though. At the traffic levels prevailing on the remaining non-electrified portion of the railway network, it is unlikely that substation costs for DC would be significantly greater than they would be for AC.
You still need a substation at the points where the grid feeds each group of third-rail substations, to take the grid voltage down to 33kV (I think it is, and it makes sense) that the railway uses to take the power on to the substations themselves. The spacing of these is similar to that of the 25kV ones, so really what you end up needing is the same-ish number of medium-voltage substations
and a bunch of third-rail-feeding ones
as well.
Where you save is that you
don't need to carry around the works of a third-rail-feeding substation on every train, whereas with 25kV you do. And you also get to keep them in a sheltered clean environment that doesn't vibrate and doesn't suffer such extremes of temperature.
Note that you do not suffer full-on third-rail levels of losses all the way from grid to train. For most of the distance the power is running from grid connection point to low-voltage substation, suffering about the same level of loss that it does over the same distance with 25kV overhead. It's only the last couple of miles from low-voltage substation to train that the heavy losses kick in.
If the sections of third rail a few hundred metres long, which at present are daisy-chained, were instead fed individually from small low-voltage bidirectional substations (with or without demand switching), the heavy losses would only apply over the last couple of hundred
metres, and you could more or less ignore them. You would need more substations than the present convention, of course, but on the other hand they would only under exceptional circumstances need to handle more than one train at a time, and their duty cycle would be a lot lower in most cases, so you wouldn't need to make them so big. Essentially what you'd be doing is equivalent to taking the on-board substations off 25kV units and sticking them at the trackside instead.
A Class 345 at max draw requires something like 4.4 Megawatts
[snip]
OHLE
4400000W / 25000V = 176A
DC 3rd Rail
4400000W / 750V = 5867A
The amount of current needed by the DC system to supply the same amount of power / work as the OHLE is nearly 34x more
What you are missing is that the loss in the track conductor (overhead wire or third rail) and return is proportional to the
square of the current. The current is (25k/750) = 33.33... times greater, so the transmission loss is (25k/750)^2 = 1111.11... times greater.
Hence low-voltage systems only step down to the track supply voltage for the last couple of miles, to keep it bearable. And as said, if you cut the spacing down to a few hundred metres you could more or less stop worrying about it.
If you used a four-rail system with +750V on one rail and -750V on the other, you could get down to only 277.77... times more for that last couple of hundred metres.
It's already been mentioned how this [sections of live rail going live on demand]would not be suitable for a high speed main line railway.
No, it was explained how it wouldn't work if you did it using unsuitable technology, which would have been a valid objection when that was all we had, but that was a long time ago now.
If any part of the derailed train comes to rest in contact with the third rail, or it gets shunted into the running rails, it will trip out.
Well, it might do. A lot of the time it doesn't. Random wreckage doesn't usually make a very good contact, and even if it did, the fault current can be so little greater than the maximum normal load from several trains simultaneously being in the same electrical section that the circuit breakers can't tell the difference.
This is another point in favour of having lots of substations feeding very small electrical sections - they don't have to feed several trains simultaneously so the trip current can be set lower.
But it's not the breakers operating that you want to be relying on anyway. Even when OHLE comes down it occasionally manages not to trip them out.
PS How easy is it to cut 3rd rail and get it in the back of a transit?
It's a similar weight per length to the running rails, so, not very. Unless you're planning to leave the rails there for 20 years while you dither on failing to decide whether to reopen the line or not, you probably don't need to worry.
I will probably be corrected and unfortunately I have not been able to find it within these forums, but I do remember someone pointing out within one of the threads, that you will not be able to get much more than 100mph on 3rd rail routes
Oh,
everyone knows third rail can't do more than 100mph...
There's not much call for that round here, sir. We have hardly any third rail routes where it's even worth thinking about trying, and for most of their mileage it would need a lot more than the power supply upgrading to get them rated for over 100mph and/or the trains rarely get a chance to get up to the maximum linespeed as it is. So there's never been any point in anyone trying to develop that pickup method to the same state that overhead pickup has been developed, and it's basically frozen in the Cretaceous. But that does not mean that it
can't be developed to handle higher speeds if someone did put the effort in. If overhead pickup had received the same lack of development as third rail we wouldn't be doing 100mph with that either.
But in any case, it hardly matters. Pretty well all the lines where it might be worth considering making them capable of over 100mph are in regions which have already been geographically bagsied by 25kV. The possible applications for third rail are pretty well all lines that don't call for exceeding 100mph anyway.
