Thames Valley infrastructure problems causing disruption elsewhere - 2019

[whole diary]

[IndustryInsider]

Possession now taken to repair the damaged track.

[TonyK]

On the final signalling hiccup.  Is it my imagination or are these now becoming very frequent?
Going back in time, surely failures occurred but we didn't notice? or trains ran without a problem?
Since centralisation, like most centralisation projects, they seem very susceptible to major outages.
Whilst putting all your levers into one big box, you save on staff costs, infrastructure maintenance, etc; but it seems a high price to pay for so many outages.
Which leads me to wonder - does the East & West Coast centralisation suffer similar outages as we do?

I for one share your sense of outage.

[IndustryInsider]

On the final signalling hiccup.  Is it my imagination or are these now becoming very frequent?
Going back in time, surely failures occurred but we didn't notice? or trains ran without a problem?

I'm not sure whether exact figures on signalling failures per year, per route, are published - one for Stuving, our resident Google sleuth? 

Given the track defects last night at Langley and yesterday and today at Slough you might think those were on the increase, but I think generally the opposite is true.  'Broken rails' were down a whopping 90% from 952 in 1998 to 109 in 2015 for example.

[bobm]

Following the repairs to the track at Slough the unit for the Windsor shuttle is trapped on the branch. As a result some services have been cancelled to conserve fuel.

[broadgage]

Following the repairs to the track at Slough the unit for the Windsor shuttle is trapped on the branch. As a result some services have been cancelled to conserve fuel.

How much fuel is needed ?
No question of a drum of diesel fuel in the back of a van and a portable electric pump worked from the van battery I suppose.
Easier just to take the default option of mass cancellations.

[bobm]

It means two return trips an hour are running instead of three.

I’m not sure how you could get a road based vehicle to access the train.  It might be possible if the train is shunted to the siding near Slough station I suppose.

[Henry]

 From where I am in South Devon, it seems that signal/track problems always seem to happen in a
 relatively small area with regular frequency, correct me if I'm wrong.
 Are the new signalling systems not as 'robust' as the one's they replaced ?
 Do they have the same problems in other area's, i.e. South West Railway ?

 From a local point of view, Network Rail seem 'to fix' problems but not necessarily repair.
 I suppose with exception, the well publicised sea wall at Dawlish.

[GBM]

 From where I am in South Devon, it seems that signal/track problems always seem to happen in a
 relatively small area with regular frequency, correct me if I'm wrong.
 Are the new signalling systems not as 'robust' as the one's they replaced ?
 Do they have the same problems in other area's, i.e. South West Railway ?

 From a local point of view, Network Rail seem 'to fix' problems but not necessarily repair.
 I suppose with exception, the well publicised sea wall at Dawlish.

Exactly my question, which interestingly, hasn't had an answer as yet.
I do wonder if in bygone days before the tinternet, you would only hear about problems directly affecting you.
Also you were not so interested in how other areas were doing.

It does also seem to be the case with signalling issues that a quick fix is done.  Guess that's because there is no budget to look into historical failures/common issues and find a long term solution.?
More questions than answers!

[TonyK]

From where I live in mid-Devon, the railway seems a relatively simple thing, becoming a much more complex organism the closer you get to London. A quick count on RTT^» (Real Time Trains - website) shows 17 trains between Taunton and Exeter, with 18 in the opposite direction, between 0600 and 1200 today, with only 7 signals (if I got that right) between the two stations. Slough, although I didn't try counting, seems to get that in an hour. Between there and Paddington, the signalling is much more intensive and the track layout much more complicated. The effects of a fault are therefore much more severe than on the lightly-loaded line. You could maintain a reasonable service with drivers telephoning from every failed signal in Devon, but never in the Thames Valley.

[broadgage]

It means two return trips an hour are running instead of three.

I’m not sure how you could get a road based vehicle to access the train.  It might be possible if the train is shunted to the siding near Slough station I suppose.

So the combined resources of network rail (whose defective track caused the problem) and GWR^▸ (Great Western Railway) whose train needs fuel, are not able to deliver a modest supply of diesel fuel to a train in suburban London. Several easy ways of doing this exist.

1) Put drum of fuel in back of a van. Pump fuel into train at someplace with suitable access.

2) Put drum of fuel in van. Unload and roll the drum/use a trolley to the platform by means of the ramp or lift intended for wheelchairs. Use a hand pump to fill tank on train.

3) Fill a number of 20 liter jerry cans with diesel fuel use a van to transport these to the station. Carry the jerry cans to the train. Use a large funnel attached to a short piece of hose to pour from jerry can into fuel tank on train.

4) Contact a fuel delivery firm with a small tanker that incorporates a pump and delivery hose. These are very common and readily available. They deliver to small tanks at the far end of a domestic back garden without problems, so a parked train should be easy.

[bobm]

While I don't disagree with the sentiment I am not sure that refuelling a train while it is stood on a public platform is likely to be countenanced.  Refilling water tanks is one thing, fuel - albeit it diesel - is another.

[Oxonhutch]

While I don't disagree with the sentiment I am not sure that refuelling a train while it is stood on a public platform is likely to be countenanced.  Refilling water tanks is one thing, fuel - albeit it diesel - is another.

And yet it is done every day, with members of the public including children and babes in close proximity - diesel [and petrol!] with lots of moving vehicles and often food and drink sales close by. My local has a sign saying Tesco and sometimes offers a discount of 5p per litre.

[IndustryInsider]

Probably a more realistic idea would have been to re-time one of the trains back 10 minutes so the 2tph at 30-minute intervals that used to operate would run.  Maximum delay of just 10 minutes to passengers, and in fact the way the connections work, some through journeys would actually be 10 minutes quicker!  Possible implications with crew diagrams and breaks though.

[broadgage]

This is how they do it in Romania, fuelling arrangements at the very beginning.
Perhaps a few of the vehicles could be procured to replace Pacers ?
https://www.youtube.com/watch?v=0zeBIxI7n1I

[stuving]

On the final signalling hiccup.  Is it my imagination or are these now becoming very frequent?
Going back in time, surely failures occurred but we didn't notice? or trains ran without a problem?

I'm not sure whether exact figures on signalling failures per year, per route, are published - one for Stuving, our resident Google sleuth? 

Given the track defects last night at Langley and yesterday and today at Slough you might think those were on the increase, but I think generally the opposite is true.  'Broken rails' were down a whopping 90% from 952 in 1998 to 109 in 2015 for example.

I don't think Network rail have ever published (i.e. externally) details of the failures that cause "incidents", though the Route Plans and similar documents have contained a lot of mentions of performance and how it's going to improve. But they are much fuller of jargon than comprehensible facts. Mostly you can get PPM^‡(resolve) and the like, telling you nothing about who did what do whom.  I've seen promises of more useful performance information, but not anything specific.

Network Rail (NR^» (Network Rail - home page)) do quite a bit of work in asset monitoring and the like (and promote it), aimed at predicting failures before they happen (the best kind of prediction, I'm sure). Some of that gets into the railway technical press. Suppliers are also busy doing this kind of stuff; even my ex-colleagues at Thales are at it now - though we didn't have any rail projects in my time. I saw a presentation of their work based on detailed analysis of the current drawn by a point motor, looking for patterns that show it meeds servicing or is about to grind to a halt. That and similar tricks may be very clever and potentially useful, but of course it all has to work, and make a difference, when rolled out on a large scale. And that I've not heard anything about.

What I have found is that NR now have a page of links to data (search for statistics and pick "Our information and data").

There is an item for Network Performance, bit its "related link" isn't (a link). The only other thing that might be of use, if anyone has the time to make use of it, is "historical delay attribution". That's not as historical as it might be (just 2018-19 and on), but is a complete list of all delay attribution incidents. There's a short explanation of its content and the column contents, which start with:

The file attached is a standard data extract from our Performance System database (PSS^▸ (Provincial Sector Services)). The
information within the file contains all “attributed” delays to passenger train services in line with the
guidance in the Delay Attribution Guide. The information is structured for industry systems and
understanding, the below provides a few key insights into the data contained.

The data contains both delay and cancellation events (denoted by the performance event code) and a
user should be careful when summing delay minutes (pfpi minutes) to exclude cancellation events.

Each (4-week) period's data comes in a large zipped CSV file - large as in up to 500,000 rows x 39 columns. You'll need a copy of the DAPR to hand, not just the explanation, to process that lot into anything useful! These are the column heads:

FINANCIAL_YEAR_AND_PERIOD
ORIGIN_DEPARTURE_DATE
TRUST_TRAIN_ID_AFFECTED
PLANNED_ORIG_LOC_CODE_AFF
PLANNED_ORIG_GBTT_DATETIME_AFF
PLANNED_ORIG_WTT_DATETIME_AFF
PLANNED_DEST_LOC_CODE_AFFECTED
PLANNED_DEST_GBTT_DATETIME_AFF
PLANNED_DEST_WTT_DATETIME_AFF
TRAIN_SERVICE_CODE_AFFECTED
SERVICE_GROUP_CODE_AFFECTED
OPERATOR_AFFECTED
ENGLISH_DAY_TYPE
APP_TIMETABLE_FLAG_AFF
TRAIN_SCHEDULE_TYPE_AFFECTED
TRACTION_TYPE_AFFECTED
TRAILING_LOAD_AFFECTED
TIMING_LOAD_AFFECTED
UNIT_CLASS_AFFECTED
INCIDENT_NUMBER
INCIDENT_CREATE_DATE
INCIDENT_START_DATETIME
INCIDENT_END_DATETIME
SECTION_CODE
NETWORK_RAIL_LOCATION_MANAGER
RESPONSIBLE_MANAGER
INCIDENT_REASON
ATTRIBUTION_STATUS
INCIDENT_EQUIPMENT
INCIDENT_DESCRIPTION
REACTIONARY_REASON_CODE
INCIDENT_RESPONSIBLE_TRAIN
PERFORMANCE_EVENT_CODE
START_STANOX
END_STANOX
EVENT_DATETIME
PFPI_MINUTES
TRUST_TRAIN_ID_RESP
TRUST_TRAIN_ID_REACT



Edit: VickiS - Clarifying Acronym