Great Western Coffee Shop

From Journey Check ...


and


Faresaver are reporting their x31 service (from Chippenham to Bath) full and standing on leaving Chippenham Station - that will be fun for over an hour - and I've got reports of the 1 carriage TransWilts service turning people away at Chippenham on the way into Swindon.

Astonished that bus routes 265 (Westbury - Trowbridge - Bradford-on-Avon - Bath) and 267 (Frome - Bath) aren't listed at taking rail tickets, and that the 15:06 Westbury to Swindon is cancelled due to "Train Crew Shortages". 

I know things aren't instantly organised, but I would have thought that with all the trains unable to get through Bath something longer could run on the TransWilts ... and with all the strain crews unable to drive their trains through Bath, a driver and conductor could be found for all the TransWilts services.   Or am I being naive?

P.S. Reports of excellent staff work at Trowbridge.  Probably the case elsewhere too but the reports I hear tend to be a bit random.

Looks like quite a "beefy" powerline is down Guessing from some tweets it was feeding the Signalling power.

https://twitter.com/wpduk/status/802467474444812288

"#Urgent works required in Westerleigh Road #B37 due to a line down. The road will be closed for most of today. Apologies for the disruption."

Hmm.  Did NR take away all their signalling power supply standby generators then? ::)
If so then there is always one portable one usually available for emergencies such as this (well there used to be in the past :P).

There appears to be a lot of trains in the area cancelled due to traincrew shortages

I thought all the signals had recently been upgraded in this apart of Network Rails GML upgrade project.

What happened to the resilience part of this project?

he answer to that depends a bit on whether and if so how it is related to the earlier failure of the 132kv public electricity supply that blacked out large areas of S Glos. 

The journey check message says "Owing to a problem with the external power supply which has affected signalling between Bristol Temple Meads and Bath Spa we are currently unable to operate any train services between those points. " and it would be an amazing co-incidence if there was no relationship between the S Gloucs issue and the railway's external supplier problem.

But surely there should be resilience if an external supplier fails to deliver?  Or are there some contingencies where the decision is "We'll take the risk"?

I suspect that the failure MAY be related to the earlier series of power
 failures in and around Bristol, I believe that there were several related faults.

The main question though in my view is what happened to the backup or standby electricity supplies for signaling equipment ?
Surely there should be generators for this sort of entirely foreseeable failure ?

Or is the new signalling less resilient than the old ?

is there not a contingency for running a very limited service without electrical signalling?  Padlocking the points and using a token or pilotman or some such?     

IIRC resilience from failure of the public power supply can be provided either by a standby generator or by dual supplies from the public power supply network.  Where dual supplies are used, the level of resilience is expressed in terms of how separated the supplies are in the network.  

So the best possible protection would come from supplies fed from two different national grid connection points (e.g. Iron Acton and Melksham), a lower level of protection from different substations fed from the same national grid connection point.  

AIUI this morning's failure was on a 132kv line from the Iron Acton national grid connection point. So it was quite high up the chain which would increase the risk.  

However, it is not clear whether the failure of the signals around Bath relates to a continued the loss of power from the grid, or whether the manner of the failure caused some damage to the signalling system.  

Until we know the facts we cannot make a judgement whether the failure is truly exceptional or should have been foreseen by NR.  

In response the earlier question as why no generators.  Standby generators are notoriously unreliable as they stand idle for almost all of the time.  That is why many users who need standby capacity use dual supply.

I have seen them at signalling centres, but I do not recall seeing them at remote locations. 

However I do not know what NR's policy is on these. 

All lines were reopened just before 8pm.

GWR tickets from yesterday will be valid today.

Thank Goodness.

---

Yesterday's failure was "exceptional" ... this sort of failure is "very rare indeed".   Except that these exceptional failures add up to make for unexceptional service failures.    From the Chippenham / Trowbridge view in the last 4 weeks:
a) Sunday 30th October - no service due to points failure
b) Monday 31st October - virtually no service due to points failure
c) Monday 15th November - virtually no service until later afternoon due to track problems
d) Monday 22nd November- severeley impacted service due to weather conditions and flooding
e) Saturday 26th November - severely impacted services due to signalling problems Bath to Bristol
and that's before we look at the minor stuff like Melksham to London journeys extended by 90 minutes due to planned engineering on Sundays, and overcrowding to the extent of discomfort on the 17:36 off Swindon on 16th ...

None of the five exceptional days we've had in four weeks can be 'blame attributed' to Great Western Railway who have for the most part been let down by the failure of Network Rail to supply the safe, operational infrastructure and signalling that GWR need to run their trains.  However, the failure of trains to run (especially at short notice), the lack of realistic and trustworthy information in many cases, and the absence of adequate alternative provision whoever is to blame does and will put people off using the train, and perhaps questions need to be asked with a view to seeing if the underlying systems can be made more robust and customer friendly.

Readers will note that I'm full of praise for the staff on the ground who deal with the situations as best they can without being provided with the means to get people to where they want to go in a timely manner, and with information feeds to pass on to the customer which are often over-optimistic and sometimes plain crazy.

So - have the last four weeks been just a run of bad luck, and it will get better without any interventions, or should someone rethink something?

This was my thought. It would at least ensure that trains were still being used, and reduce the necessity for road replacement services.

With the Bath Christmas market, this was not a good time for this failure to happen.

Appreciate that this was a fairly major failure of the supply outside NR's control, but it does feel like the contingency planning for such events is not particularly robust. I would have thought there should be a plan with various scenarios listed, so that in such circumstances it can be dusted off and put into action.  I appreciate that you could dream up a lot of scenarios, but at least the thinking would be done in advance as to what is practical and sensible under the circumstances.

Of course, I wouldn't expect the plan to cover such specifics as the fact that the Xmas Market was in full swing, but it could nevertheless have suggested a Bath - Bristol shuttle with a policeman. An 8 car unit or HST would be able to shift a reasonable number of people in and out of Bath I would have thought. 

Another thought, given it happened at around 9am I expect many people just gave up and didn't travel to the market.  It would probably have been much more difficult if the failure had occurred mid afternoon when Bath was full with people who then needed to return home.

A generator seems a wise precaution, or possibly a battery bank if the load is small enough.

Generators do not have to be unreliable. It is infrequent use that tends to reduce reliability.
Weekly testing is advisable, preferably on load, this can be automated in order to reduce the labour requirement. If the generator fails to start for the weekly test then the cause can be investigated and rectified hopefully before the generator is needed in anger.
With a mains supply available normally, it is a simple matter to keep the engine warm and the starter batteries fully charged.

A single generator if of ample size could also supply the nearest station, and possibly other railway equipment.

Even with a weekly test on load my experience is that they are still not that reliable. 

The market did seem to me relatively quiet compared to what I expected (and on a sunny day as well).

Generators are indeed not always reliable, even if tested weekly.
All too often the installation is done incorrectly and permits of a brief test but not a prolonged run on load.

I have considerable experience of standby generators and the testing thereof, and assisting in finding out what went wrong after a failure. IME the most common faults are, in approximate order of popularity as follows.

1)Defective or discharged starter batteries-------the remedy is obvious.
2)Low water in cooling system, overheat protection stops engine-----check regularly
3) Engine starts but fails to take over the load due to defective, or misunderstood, changeover switchgear.

4)Engine runs but stops for want of fuel before power is restored.
5)Engine runs but stops due to electrical overload/unauthorised extra loads.
6)Engine runs but stops due to low lubricant oil level.

Or finally, which is not truly a fault at all, it all works perfectly but the standby generator only supplies ESSIENTIAL loads as was the design brief. The elfansafety then send everyone home because they feel that the workplace is unsafe without every light and electric heater working.

The first 3 reasons given account for the great majority of failures IME. Learning points include use of good quality starter batteries, preferably of twice the recommended capacity. Duplicated battery chargers with alarm in case of failure. Replace starter batteries every 4 years.
Check coolant levels regularly, general maintenance to prevent leaks.
Test not just the engine but all related automatic or manual changeover switchgear.

Including 2M12, the 15:06 Westbury to Swindon. Interestingly, for the one day only an Empty Coaching Stock train called 5M12 ran from Westbury to Swindon scheduled at - err - 15:06.   

That's the one I remember - having inherited the backup generator from the previous users of the factory, no-one had gone through the process of splitting the circuits between essential and the rest, and making sure the essential power demanded was available. So the generator struggled up to a lot less than 50 Hz, and overheated all the fluorescent ballasts, one of which (oil-filled) burst and produced an atomised spray of ... something. We hoped it was silicone - but there was a rumour it was a PCB.

NR don't have those all over the place - but they gave a 5-year contract to Aggreko in May 2016 (http://blog.aggreko.co.uk/aggreko-wins-five-year-rental-power-contract-with-network-rail), for mobile generators to cover both planned and emergency needs. So why didn't they get one of those in?


Well, it would depend on how much power that much signalling needs (though it was only a few miles). And, as reported, it wasn't the loss of power that mattered - that was restored within an hour. It was some consequential damage (which might be just one of those things, or might be one of those things that was NR's fault). I wonder whether the damage was to the switchgear where the external power feeds in - in which case connecting a generator might not be quick enough to help. More generally, the response time of a mobile generator might always be a few hours - more suitable for longer-term emergencies.

Well in my time in NR maintenance they were all tested 'on load' every 4 weeks and every time they were used 'in anger' they worked perfectly.  I can only recall having to bring in a mobile generator once and that was to cover the rebuild of the permanent one over an extended period.

At the risk of drifting well off-topic...
We had a UPS (uninterruptible power supply - just a load of batteries and stuff to turn it into 220V AC) to feed four servers, at least long enough to shut them down in an orderly fashion if we lost mains power.
One day, the supply to the buiding failed.  Our four servers kept going, and the display on the UPS said that it could keep them going for another hour or two.  So we decided to shut them down, as no-one could use their desktop PCs to acess them anyway.
But all the terminals (screens and keyboards for logging in to the servers) had no power! So we couldn't type in the shutdown commands.  And the UPS only had 4 power sockets.  So there was a frantic search for a laptop with a) terminal emulation software and b) enough battery power left to be able to power off one of the servers, so we could unplug that one from the UPS, and then plug in a terminal to shut down the other 3.

Moral: rehearse your emergency plans.

Very true. And, of course, the same also applies to data recovery off backup.

Purely for the record - the story from the BBC (http://www.bbc.co.uk/news/uk-england-bristol-38117848):

"Several trains were cancelled, and one person reported on social media that "hundreds" of passengers had been "left stranded" in Bristol until Sunday."

Really? Funny how I fairly easily got a train from Bristol to Bath at around 8pm.
I am sure there would have been longer waits for some people (I was pretty lucky in that I managed to jump on a service that was 34 minutes late just before it actually left), but I seriously doubt people were "left stranded".

... but that story was on the BBC!  What could possibly be inaccurate about it??  :o ::) ;D

I guess the most likely inaccuracy would be that more than one person reported that on social media - if only by passing it on - is that what you meant?

That would have been considered by GWR. If it ws thought that the power outage was going to last all day at that stage (and it did), would you risk transporting many thousands to the Christmas market & not be able to get them back again easily, or would you not attempt to carry them in the first place & your lift easier?

I think they made the right decision. Those few *needing* to travel, I'm sure got away somehow.