Solar storm railway operation disruption

[St. Simon]

3 minutes ago, adb968008 said:

As your a signaller, I would be interested to know your thoughts on why it happened.

 

As I said another service was bearing down on that signal and deaccereleratd considerably. Nothing crossed the DF to warrant it, and it was cleared after… as you can hear on the video, I wasnt the only one who saw it and thought it odd.

 

Hi,

 

A momentary Track Circuit, Power Supply, TPWS, Point Detection or Lamp failure, the signaller might have either needed to (obstruction or report of a tresspasser) or accidently cancel the route, many reasons why a signal revision would happen. 

 

It could be to do with the 92, but I can't believe that the emissions are that bad it would affect a equipment housed in a Loc 6 tracks away!

 

Simon

[jjb1970]

1 hour ago, Nearholmer said:

If you read what the uni guys actually said, they talk about it in very sober terms; don’t fall into the trap of having your buttons pressed by newspaper headline writers, or get trapped into a Mexican Wave of reaction to something you haven’t read (unless, of course, you have been able to access and read the base work).

 

 

 

I agree, if you are interested iin any paper or document then there is no substitute to reading the actual documents.

However  in this case I don't think it is the paper people are reacting to, it is how it is being presented by the BBC. I did read the paper, though as with any paper based on modelling you can only form a real opinion if you analyse the model in use, which I didn't. 

The media are terrible for stuff like this, the headline is just click bait and the version I have doesn't link to the paper which is a pet hate of mine. 

[KingEdwardII]

3 minutes ago, jjb1970 said:

the version I have doesn't link to the paper which is a pet hate of mine. 

The BBC are notorious for not providing links to the stuff they discuss in their online news articles. They did it again yesterday with respect to the publication of the latest ORR data for passenger traffic through stations.

 

Yours, Mike.

[jjb1970]

1 minute ago, KingEdwardII said:

The BBC are notorious for not providing links to the stuff they discuss in their online news articles. They did it again yesterday with respect to the publication of the latest ORR data for passenger traffic through stations.

 

Yours, Mike.

 

It's something I find annoying and really rather questionable as it begs the question why they don't make it easy and help readers do their own reading and form their own conclusions. 

[KingEdwardII]

3 minutes ago, jjb1970 said:

why they don't make it easy and help readers

I can't decide whether it is some weirdly misplaced idea relating to impartiality (etc) or whether it's because the folk at the BBC simply don't understand how the world wide web actually works. Some of the decisions they make about (not) providing links between some of their own web pages tends to make me think the latter...

 

Yours, Mike.

[jjb1970]

A bit like ignorance of VPNs. In the case of the news I suspect part of it is an assumption that readers/viewers need somebody else to tell them what documents say and mean. 

[The Stationmaster]

16 hours ago, phil-b259 said:

You are getting your track circuits confused!

 

TI21s are very much still in use through the network (including where the 373s, class 92s, NSE networkers were employed) although they have now been ‘digitised’ (rebranded as EBI200 and EBI400) where the receivers now set themselves up rather than techs having to adjust the gain setting manually plus having the ability to transmit all the perimeters to data loggers (and there hence to remote condition monitoring techs who can view the data in real time).

 

The track circuits which had to be removed en mass from routes used by the 373s etc are ‘Reed’ type track circuits - and their existence on the Brighton main line was the reason the 92s were banned from using the route via Redhill even though RfD paid for the electrification of the Redhill to Tonbridge section.
 

 

Although Reed tracks intended to be able to be used without IBJs such a configuration was problematic and never employed on the ex Southern region where all Reed track circuits came with IBJs. In recent years though Network Ra8l has run programmes to get rid of them entirely (including on the BML) with the replacement being the EBI (or TI21 in old money) type.

As I said I wasn't sure if it was the T121.  However there was definitely one 'jointless' track circuit which was very prob=ne to 'noise' interference and was being rapidly removed oin the mid '90s.  If I can delve out any of the old test programme and 373 route clearance stuff i might be able to identify it.  

 

Reed tcs were definitely a problem as you say.

 

 

 

 

 

[KeithMacdonald]

3 hours ago, jjb1970 said:

It reminds me of risk assessments where absolutely everything is identified as a potential fatality and dire threat to the world. Especially qualitative assessments, someone in a mess room stumbles, cracks their head and dies, and as they fall they grab hold of someone else and they die from a head injury too. In a sense it's correct, just about anything we do in life can end up in death but it doesn't really help risk management and it's unhealthy to go from a sensible awareness of risk to being scared of shadows.

 

I've seen far too many of those that consistently frighten the pants of everyone with "the sky is falling down!" fear-porn messages, and consistently fail to provide any sense of proportion, or proportional risk. Yes, a mega-asteroid* might hit our planet tomorrow, and wipe out life as we know it (Jim). But what are the actual odds of that happening? And should normal life grind to a panic-stricken halt because of it?

 

Far too many of the "predictions" that are parroted by the MSM are woefully inaccurate computer models that border on the Mystic Meg. As modellers ourselves, we know that a model can be as accurate or inaccurate (and hideous) as we care to make them. A truly good accurate large-scale model can take years to develop, while our managers/politicians want simplistic sound-bite quick answers that usually turn out to be wrong. But only a few realise that in hindsight, while most people are too distracted by some other piece of more-recent fear-porn to ever go back and do their own checks on the validity of the previous models/guesses.

 

* I was careful to choose a mega-asteroid as an example, and not one of the other "usual suspects" fear-porns that we (and especially our children) are being bombarded with.

[jjb1970]

Modelling is a great tool, but it relies on the input assumptions, veracity of data and relationships and an awareness of limitations. That said two of my pet hates are model related:

 

-the increasing tendency to equate modelled analysis with measured data and lose the distinction; and

-presenting modelled analysis and claiming the 'data' proves a hypothesis without giving any details of the model.

 

And this is before the whole issue of corellation and causation.

[david.hill64]

I was in the cab of an HST leaving Reading towards the west when our green signal momentarily went to red before returning to the green. The drivers blamed the 37 in the triangle causing loss of detection on a set of points. They weren't at all fazed by it.  However, for me, seeing a loco on an apparent conflicting route cause the events was an exercise in sphincter control.

 

As Simon says, lots of things can cause a signal to revert to red. Getting a transient wrong side failure would be a different animal.

[lapford34102]

Nice timing

 

https://www.bbc.co.uk/news/av/world-67734790

 

Stu

 

 

[AY Mod]

1 hour ago, lapford34102 said:

Nice timing

 

https://www.bbc.co.uk/news/av/world-67734790

 

Which shows that it is a relevant concern. We haven't enough experience of our own technologies to accurately state what the effects, impacts and reach of any damage could or would likely be.

 

And that one wasn't even really pointing this way.

 

[billbedford]

2 hours ago, lapford34102 said:

Nice timing

 

https://www.bbc.co.uk/news/av/world-67734790

 

Stu

 

 

 

Has anyone else noticed how odd the numbers quoted in that article are?

 

This time -- X2.8

2017 -- X8.2

2003 -- X28

 

No info about the scale either. 

[TheSignalEngineer]

On 15/12/2023 at 01:38, adb968008 said:

I caught a 92 tripping a signal at Finsbury Park once…

 

9mins in, look where the 92 is in relation to the trackside cabinets

 

On 15/12/2023 at 09:38, St. Simon said:

Except, the Loc cases that that loco passes aren't associated with any of the equipment that controls that signal, so the reversion is unlikely due to the 92.

We had no end of trouble with signal reversion and mysterious unexplained failures during the SSI testing on the DC Lines in 1988. It was all new technology then, the first application of Long Distance Data Links. During one session I has a scope on the location power supply 110V busbar at Harrow and Wealdstone. After about two hours I could tell the class of any electric loco or unit approaching on the main lines by the way the waveform was distorted.

We could pick up a trace on the power supply waveform presumably from the RAF radar when a plane was flying into Northolt.

 

Class 325 units could put spurious signals into the rails when being towed via Salford Crescent. On Thameslink we had a lot of strange experiences of how traction return currents behave. A long story but I hope the earthing on the structure of the old Smithfield Market is still good.

[KeithMacdonald]

2 hours ago, lapford34102 said:

Nice timing

https://www.bbc.co.uk/news/av/world-67734790

Stu

 

1 hour ago, AY Mod said:

Which shows that it is a relevant concern. We haven't enough experience of our own technologies to accurately state what the effects, impacts and reach of any damage could or would likely be.

 

 Users of GPS have been reporting some interruptions in availability. Marine GPS positions would pause at the last-good location. Friends using "smart watches" for running route logs report strange gaps or jumps in the recorded route. Fortunately we don't need GPS for railways do we?

 

[The Stationmaster]

Marine GPS positions reported on the 'net don't need solar activity to 'freeze' or get lost for hours.  I've seen it at times ona vessel I keep regulatr track of but I never saw any sort of navigation information failure while aboard that vessel (for one or two weeks of they year and nobody ever mentioned such losses of information.  

 

Maybe they do happen but has it ever been raised as an issue that needs to be dealt with?  Don't forget that most ships now rely entirely on electronic systems and GPS, or similar, style inputs for navigation purposes so any issues there would be far more serious than bobbing track circuits or signals on a railway.

[KeithMacdonald]

41 minutes ago, The Stationmaster said:

I never saw any sort of navigation information failure while aboard that vessel (for one or two weeks of they year and nobody ever mentioned such losses of information.  

 

In UK waters, it usually happens in specific sea areas (for one or two days) during NATO exercises, with GPS jamming as part of the exercise. But they usually and kindly post Notices To Mariners to let people know in advance. Whether boat crews are aware enough to read those Notices To Mariners is a seperate issue.

[TheSignalEngineer]

One of our ex forces staff was sailing a private boat round part of the English coast during a the first Gulf War. When he was passing between Lands End and Longships the SatNav was telling him he was in the Bay of Biscay. Fortunately he knew what was going on and wasn't relying on it.

[adb968008]

 are railways much more vulnerable to single points of failure… ?

 

if a signalling centre went pop, is there redundant back ups the signallers can just move into and carry on or does the job stop long term in a large area, if a whole location, or IT facility just went up in smoke ?

BA came a cropper once by not spending on its IT.

 

Edited December 16, 2023 by adb968008

[MarkC]

19 hours ago, adb968008 said:

 are railways much more vulnerable to single points of failure… ?

 

if a signalling centre went pop, is there redundant back ups the signallers can just move into and carry on or does the job stop long term in a large area, if a whole location, or IT facility just went up in smoke ?

BA came a cropper once by not spending on its IT.

 

This was the big concern when the centralized signalling centers were first proposed - what happens if, for example, somebody accidentally puts a *JCB scoop through a cable run? I don't recall any proper answer to this ever being offered...

 

* other excavating machinery is available...

 

Mark

[jjb1970]

I don't know what the railways do, but there's a suite of tools to analyse failure modes and identity vulnerability to single point failure and common failure modes. Common failure modes get less attention in the media as they're less obvious than a single point (it doesn't matter how much redundancy you have if a common failure mode can take it all out). Tools like FMEA, FMECA, Hazop, FTA etc are excellent,  but only as good as inputs.

[billbedford]

20 hours ago, adb968008 said:

 if a signalling centre went pop, is there redundant back ups the signallers can just move into and carry on or does the job stop long term in a large area, if a whole location, or IT facility just went up in smoke ?

 

 

If we are talking about solar storms, I expect the damage would be more widespread than a single system. 

[Michael Hodgson]

On 16/12/2023 at 12:41, The Stationmaster said:

 

Maybe they do happen but has it ever been raised as an issue that needs to be dealt with?  Don't forget that most ships now rely entirely on electronic systems and GPS, or similar, style inputs for navigation purposes so any issues there would be far more serious than bobbing track circuits or signals on a railway.

 

All very well in normal operation, but I would hope that they still have charts and somebody on board who knows what a sextant is for?

[jjb1970]

3 minutes ago, Michael Hodgson said:

 

All very well in normal operation, but I would hope that they still have charts and somebody on board who knows what a sextant is for?

 

The problem is the same as any other industry, if people rely on modern assistance systems and crewing levels/workloads are planned on the basis of all sorts of such systems to automate and manage systems then people are going to struggle if those systems go down and they have to open cupboards and get gear out they use once in a blue moon to make sure they still know what it is.

Even many years ago I remember stuff like being called out by the duty engineer because an exhaust gas alarm on a cylinder had gone insanely high, because the DCS was telling them it was something like 1200 degrees then it must be right. They weren't stupid or badly trained, they were just reliant on the control system because that's the way the industry evolved. I know people thought I was a dinosaur because I insisted everyone manually start, synchronise and load share generators regularly so that if the PMS went tango uniform they could still do it manually. There were similar stories upstairs, my favourite was a second officer calling the old man in a panic because an oil rig was approaching the ship at 27 knots because of an issue with the radar assisted plotting display.

We had the same issue in electricity generation, operators lost the knowledge of plant layout and operation to the point my then employer deliberately specified a lower degree of automation to force operators to retain the skills needed to do it the old way if necessary.

[TheSignalEngineer]

1 hour ago, jjb1970 said:

We had the same issue in electricity generation, operators lost the knowledge of plant layout and operation to the point my then employer deliberately specified a lower degree of automation to force operators to retain the skills needed to do it the old way if necessary.

In my early railway days I remember doing simulated mains power failures to test diesel generator or inverter supplies and manual/ automatic changeover and restoration circuits..