Signalling power supplies into Victoria Station

Hi Ian,


I'm sure there are people here who do, but this is a public forum so it would be inappropriate (and probably in breach of their contracts of employment) for them to post here. It is a challenge for these forums that those who do know the answers to complex issues are often not in a position to discuss them. It's a nice idea, "let's have a forum for engineers to share ideas and improve the world", but most of the stuff we do as a day job - and hence have real expertise in - we really can't discuss here. Ideas welcome as to how we solve that. (You'll notice that - except on the Wiring Regs forum - mostly when anyone posts a "does anybody know?" question the only replies are from those who start "it's not my field, but...") 


And this is an extreme case - if I knew how to bring down a large proportion of London's railways by generating a single failure, and I posted it on an open forum, I suspect several people in long overcoats would be paying me a visit!! Quite reasonably. Also, of course, there are rather more professional ways that I should be notifying and working with the relevant bodies.


In this case, having worked in the UK rail industry for 25 years I can be pretty confident that an awful lot of people will be spending an awful lot of time trying to make sure that (as far as possible) this doesn't happen again. What happens in practice, which works reasonably well, is that the infrastructure owner (presumably Network Rail in this case) works with the relevant suppliers of this and any similar equipment to ensure that it doesn't happen again. Been there, done that, in a past life I was regularly on the receiving end of those visits: "your competitors' products failed like this, how can we be sure that yours won't fail in the same way?"


In theory there is an interesting general point here: when an accident occurs on the railways the results of the investigation are made public, but when a major availability failure occurs the results are not. Which does miss the advantage of letting new suppliers offer innovative solutions to prevent reoccurrence (which I guess was the point of your post). However, it's a fine balance. If you publicly report exact equipment and system failures, and hence give competitors an advantage, then equipment and system suppliers will consider whether this is really a market they want to enter - and it's hard enough getting them to enter it as it is due to the very high technical standards (relative to cost) expected. Any thought of how other industries manage this which the rail industry could learn from?


Finally, of course, what the newspapers report relating to engineering failure tends to be grossly oversimplified. And equally, absolutely nothing to do with a signalling system is ever simple. So personally I reserve speculation until I hear the facts. (And when I do know the facts I usually then can't discuss them. D'oh.)


Sorry if that all come across as a bit bland. Oddly enough, despite spending Thursday travelling into, across, and out of London I seemed to be the one person not affected by any of the various delays! (This was only one of several things that went wrong that day.) But I saw plenty of the side effects, and fully agree that we don't want it to happen again. 


Interesting point,


Cheers,


Andy

Independent Safety Assessor (rail) / MIRSE

Hi Andy,

This statement released to the BBC gives sufficient detail, I think
http://news.bbc.co.uk/1/shared/bsp/hi/pdfs/06_07_18_victoria.pdf 

It seems that the changeover panel, which selects one of various power supplies, itself became a single point of failure.

The statement is referenced from this story:
https://www.bbc.co.uk/news/uk-england-london-44733968 


There's no further detail, but at least we know that there was provision of redundancy in the power supply design.


You make some interesting points about the difficulty of entering the railway market. I spent 6 years in the rail industry recently (not in signalling!). 

My thoughts about the effect of this typically would be that the relative lack of competition, the small size of the market, and the cost of recertification would be disincentives to product development. However, having a single customer has its own issues, such as intermittent order flow, and the limited competition does provide some reassurance to vendors. 

Thanks for the link, I hadn't seen that. Yes, this exactly the sort of common mode failure that gives us nightmares. (I spend my life, and indeed spent many hours this morning, saying "yes I know it's very unlikely, but what would happen if it did, and what could you do to stop it?" Followed by going back to the end customer to see if they can afford the more robust solution...)


You have summed up the rail market very well...the secret is to supply to the UK first, and then use that certification to gain entry to the world wide market. Other than (in my experience) Russia, the UK is usually the toughest market to get into because of the most rigorous standards, hence if you can achieve that you can achieve anything. And of course be prepared to supply bespoke equipment to military levels of robustness and reliability for the price you'd usually sell off-the-shelf commercial equipment


Innovation (product development) is possible, and when you do it well it is popular for precisely the reasons of continuous improvement we're discussing, but it does take a very serious management approach to be successful. Personally I enjoy that process, handled well it makes you produce very well engineered designs, but it drives many innovation managers completely mad!!  


Incidentally, my often expressed personal view is that one of the biggest technical challenges and opportunities facing the UK Rail Network is accelerating the widespread roll-out of condition monitoring infrastructure (although I wouldn't like to say whether it would have helped in this case). This issue has been rumbling on for years - a very large proportion of system outages are detectable in the early stages, the problem is the cost of the infrastructure to effectively monitor and react to them. It's a phenomenally complicated issue - for years the IET ran two day conferences just on condition monitoring in rail - with no simple solutions (or more accurately they've already been done) but lots of opportunities.


Cheers,


Andy