Yes, that’s exactly what I was trying to describe, using Google satellite view I was able to find the exact arrangement where they change between single and double track cantilevers. But I’m not so sure there are redundant routes between the ATFS and the eventual trackside connections, they may be in a completely different area, eg as in the 6 taller devices in the background of the picture in post#4.
While the London Underground has used cables with low smoke emission and more resistant to ignition for some while, I suspect this is not true of the surface lines, so their cables probably are more susceptible to fire.
It's a less intensive service on a Sunday, but checking Maidenhead to London, both Elizabeth and semi fast GWR services appear to be running normally.
As you can kinda see on the Google Earth 3d view, I think these cable runs likely carry all the circuits between Kensal Green substation ("A") and the point at which they connect to the OLE ("B") which is not a very long distance at all, so a very unfortunately located fire. I'd imagine at least 8 +25kV circuits (4 for the lines to Paddington, 4 for the lines going west) and maybe 2x -25kV AT feeder lines. There may also be dedicated circuits that go to Old Oak Common depot as well which is also unfortunate. The tracks immediately against the feeder station are actually the approach tracks for the XR depot. I know there is also a 50kV cable from Kensal Green to Westbourne Park substation (which only feeds the XR core tracks) which may also have been damaged depending on how Kensal Green ATFS organises its cables. This interconnector is not often used as the XR core is normally fed from the eastern end
Of course but highly unlikely we will ever get to know. Also what was the source of ignition? Potentially a overheating cable/failed joint although more likely a discarded cigarette on the adjacent canal path.
This location is well clear of the main lines and even accepting the supposition that all these cables provide power to the OLE that wouldn't have prevented the movement of diesel powered trains along the fast lines. GWR though ran a very scant service maybe that was more to do with crew displacement but even so with an all bimode IET fleet they had the opportunity to do better. The net impact was severe overloading on SWR and Chiltern which is creating the potential for another catastrophe elsewhere. GBR has plenty to do but it would be wise if it took up some medium term challenges like how we manage and recover from such incidents.
Someone mentioned last night that the mains were already under possession yesterday?
The entire line was clearly blocked for the emergency services to respond. It isn’t that far away. There was nothing blocking the lines towards Reading at any point, and nothing towards London after around an hour. There were two stranded trains but they were in the 6 track section and easy to work round.
I don't think there are redundant cable routes. But what I was trying to explain is that they will have alternative feeding arrangements so the OLE can be supplied from feeders elsewhere on the network, even if this feeder is completely out of use.
Yeah most likely Didcot ATFS extended past it's usual boundary at Maidenhead to feed as far east as you could go (does anyone know how much was switched back on?). It might theoretically be possible to pull some switches around the Westbourne Park portal to feed Paddington from the XR core but I doubt that would ever be considered, especially in the short term.
No way of me properly knowing this but I'd imagine all the OLE circuits for GWML, OOC depot are within that cable route apart from the interconnector to Westbourne Park
There's no neutral section at Kensal Green. Paddington to Maidenhead is one feeding section. There are 2 grid supplies at Kensal Green but one of them just does nothing and either fills in for the other in N-2, or can be a backup for the XR core if Pudding Mill Lane is in N-2.
Given comment from @GingerRH it appears that this was the case which is rathers goes against basic electrical principle of providing diversity.
Thanks for clarity too grounded on classic 25kV from my ECML days!
Well, that’s what I was cautiously thinking, but based solely on the pictures as they appeared. But NR & TOCs seem to play down the risk to infrastructure, and we get news items in the mainstream media referring to it as a ‘grass fire’.
The Distribution operators and National Grid would insist on a lot more separation
No control of the circuit breakers in Kensal Green FS?
So the fire cant be tackled until earths are up?
XC to Bristol TM, then GWR to Paddington arrive 01:15. At least, that's what I assume happened.
This part is planned engineering work, unrelated to yesterday's fire.
Longer feeding distances is both an advantage and an disadvantage of the AT system, neutral sections are normally considered a hindrance for all sorts of reasons, but in the case of emergency isolations you suddenly want more of them. I think in some niche cases, neutral sections have been put in purely to allow for strategic isolation.
Diversity of feeding doesn't ever seem to be considered at this low of a level since we're talking about at most a 200m long cable route. I think even under BR practice in a similar situation the cables would likely be grouped together as well. But BR schemes tend to put substations directly at the OLE connection site as often as possible. Equally though, the fire could equally happen within the substation which would have a similar result. I wonder if best practice might change from this, maybe having two different cable routes, but you'd have to choose which circuits get which routes: perhaps all the Down lines on one and the Ups on another meaning you could throw alt feed switches on the OLE structures to group down and ups together.
The platonic ideal written in textbooks of an uninterrupted run of dual supply feeder stations and midpoint TSCs almost never happens in practice anywhere, there's always something more confusing going on. MML is the closest to consistency. Feeder stations with no neutral sections was sometimes a move made on BR era systems, sometimes but not always because they only had one grid supply incoming. That latter point is never repeated in post BR systems interestingly. For ECML: Nene, Newark, Benton and Ulgham Crossing come to mind - although the latter has 2 grid supplies, they're just in phase! (Even more strange!).
This is much more damming in my opinion. On older systems, ECRs operated their own pilot line networks that ran within S&T cable routes. But nowadays, and I'd assume anything GWEP follows this practice, SCADA communications are handled by the general use FTN network, so it's surprising to me that all the fibre lines to the substation were within that cable route.
