A very informative and detailed article.
A traditional relay interlocking is relatively immune to voltage variations. What I think caused the problems was the electronic remote control systems that connect the interlocking to the control centre - their power supplies would have tripped and needed resetting.
In addditon to that. How is it that one station has its screen in disruption mode but the other doesn't?
In a power switch over scenario relay interlockings and basic TDM field units usually happily restart fully functional after signals drop to red or go dark. It is complex processor-based systems that sometimes don't properly reboot automatically, and in the case of SSIs and similar they sometimes blow their software fuses, so UPS systems are neccessary to keep these going through a problem. SSIs are usually all grouped together at the control centre though so there's usually a tech nearby to intervene, unlike relay installations which are distributed around the network at their respective junctions and stations.
Ideally some kind of power conditioning is required somewhere in the system that could isolate critical complex systems (including the UPS) from the incoming overvoltage supply before switching to the UPS. I doubt if that could be truly continuous though, which could be a problem.
Interestingly (perhaps!) our freezer went into alarm mode just before midnight a few days ago, which I believe was the result of major voltage fluctuation (according to the information I could glean on line about the fault code). It was the first time in the nine years that we've had it that it's done this, and it's been fine again since. We live near Dorking, Surrey. I wonder if there have been some quite widespread voltage blips across Surrey/London during the last week, one of which caused the signal system failure(s)? Probably completely unrelated, but...
Sorry to reply to my own post but I've thought about that some more and all the extra voltage monitoring device needs to do is to switch off the supply downstream which would be just like a regular power off event as far as the UPS is concerned so it should automatically cut over to battery as usual, subject to its own dependability of course. The excess voltage over timescale threshold of the new device needs to be set correctly to trip at a lower volt-time function than the systems protected downstream (like fuse discrimination) and the input side needs to be significantly overrated voltage-wise so as not to be damaged itself by extreme events. One has to wonder why the supply from the grid itself doesn't come with such protection built in.
Yep as I said, all trains were showing as On Time. As their departure time arrived they just disappeared off the screen one by one
Ditto. 20 seconds is an incredibly long duration for such an event.
You are effectively proposing what alegatta proposed previously. As you say, this kind of over-volt trip on the incoming main high-voltage power supplies would need to be faster and more sensitive than that on all the down-stream internal electronic power supplies - no point in switching after those have shut down. This would be technically very difficult, to say the least. Worse, such a sensitive trip would likely operate quite frequently, leading to the supply being far less reliable than it is now.
Just another thought - Would the remote control systems have redundant power supplies(Like in a server, for example)? And could both power supplies be fed from separate supplies, whether that be two DNO supplies or a traction/signalling supply with over volt protection on the primary side shutting down the individual PSU rather than losing the whole piece of kit? (I am aware that this might already be the case, and that both power supplies might have shut down reacting to the same fault on the same DNO circuit)
In my experience a whole equipment room can be switched from one supply to another, but all systems within are connected only to the active supply. If an individual system had a second power module, I suspect it would just switch the overvoltage to that on failure and of course that would trip as well. If it was already 'hot' it would have blown at the same time anyway. Also we don't know whether the overvoltage was present on only one or all of the available supplies.
Most modern remote control systems do have a second PSU, but this is in case the first one fails. As MarkyT says, they are both fed from the interlocking power supply, on the basis that if the interlocking hasn't got any power then there is no point in having any remote control.
