Delays through the central operating section again this morning. No movement at Farringdon Eastbound for at least 15 minutes thus far. The Westbound is also delayed due to congestion at Paddington with all reversing roads occupied.
-
Our booking engine at tickets.railforums.co.uk (powered by TrainSplit) helps support the running of the forum with every ticket purchase! Find out more and ask any questions/give us feedback in this thread!
Elizabeth line reliability issues
- Thread starter mrmartin
- Start date
- Status
Sponsor Post - registered members do not see these adverts; click here to register, or click here to log in
R
RailUK Forums
Delays at Abbey Wood due to driver displacement and a depot train that isn't moving to the depot. Causing a small tail back towards custom house.
3 Mins Later:
Tail back has reached Whitechapel but now the depot bound train is on the move. No clue why the 9T38 was withheld longer than need be once it's driver arrived.
3 Mins Later:
Tail back has reached Whitechapel but now the depot bound train is on the move. No clue why the 9T38 was withheld longer than need be once it's driver arrived.
Last edited:
redreni
Member
Pleasantly surprised to have working escalators (up and down) at Abbey Wood this morning to help me transfer off a westbound Southeastern service to the Elizabeth line and to hear PA announcements on the Elizabeth Line platforms telling me the first departure would be from Platform 4. It makes a real difference when the equipment at the stations works. The transfer I made would have been difficult if not impossible without escalator assistance (especially since the staircase isn't immediately adjacent to the escalator, so when the escalator is closed and blocked off you lose 60s walking down the platform to get to the stairs).
Nothing eastbound at Canary Wharf for about half an hour right in the peak 0815-0845 I learned. Then everything seemed to come through after 0900 on minimum headways and so things got disorganised at Abbey Wood. Meanwhile there seemed a major hangup at the same time down the GWML.
It does seem there has been a gross disruption in the peaks every day this week, sometimes two at once. One hopes that the reasons for these are analysed and addressed, rather than just a "well, we fixed it" coming from the sharp end without further ado.
4V42 failed just past Twyford, still there but things are moving better now, with no service to Twyford.
Yeh sadly those escalators are the wrong type for outdoor so prone to breakage at times. PA system indeed helps the staff alert SE side if things about any delays before they detain at least for aslong as it works.
I think that was a different problem again.
Is it just me or has the central operating section been relatively well behaved over the past week? They’ve had a few minor incidents on the Network Rail sections such as a failed freight train at Langley last night, but these have been fairly contained.
Has there been some further system updates?
Has there been some further system updates?
iphone76
Member
No updates as far as I'm aware. But it has been a joy this past week seeing it mainly work like clockwork. Long may it continue!
It was perfect for me today. As was Thameslink.
Even the large freestanding signs in the Core about the platform indicators showing nonsense have been turned to face the wall ...
New protocols regarding issues in/out of Westbourne Park. Most of the issues can seemingly now be rectified within 5 minutes once diagnosed, not causing the massive sit-downs of the past few months.
The train fleet had become by far and away the biggest detriment to performance so naturally a lot of attention seems have been put towards it.
Even the large freestanding signs in the Core about the platform indicators showing nonsense have been turned to face the wall ...
Also the same. Just more work getting to the root cause of a number of issues across the board. Not to say this wasn't already happening, but a lot of the systems issues have been isolated over time and fixes resolved.
Despite ELR400 supposed to be last update and Crossrail is formally over, there's already been a few "mini patches" and another larger one due to happen in 2024...
Because this is my local line, I was keen to track its performance. I created a simple dashboard using the RealTime Trains API to show an approximate daily PPM for the central core. I thought people on this thread may find it interesting, so sharing the link:
https://zhibek.github.io/crossrail-ppm/
https://zhibek.github.io/crossrail-ppm/
That's very interesting, thank you, though as I guess it is all services it would be interesting to look at performance at peak travel times, say at 9am. Because such statistics (the official ones, too) can hide the customer experience by taking into the averages all the 'easy' services, at quieter mid-day times, giving them equal standing to peak services with standees stuck in the tunnel.
In passing, even I didn't expect the overall figures to be as poor as portrayed.
In passing, even I didn't expect the overall figures to be as poor as portrayed.
good stuff.
Just a small offer of advice: ‘On Time’ has a specific meaning in railway performance terms, it means being On time to the minute (strictly, to 0-59seconds) at every station stop.
PPM is only measured at destination, but cancelled trains count as an automatic failure.
I think your figures are 0 - 4m59sec, but I’m assuming this is only for trains that run, and is measured at the station shown?
greatkingrat
Established Member
- Joined
- 20 Jan 2011
- Messages
- 2,793
Although the high frequency means the PPM figures don't tell the whole story. If all the trains are running 15-20 mins late, but there is still a train every few minutes, then from the perspective of someone waiting in the core, that is still a good service.
Any reason it's 0800-2000 only? Also it only seems to be Abbey Wood and not the rest of the line? For example PPM whilst good on Friday certainly wasn't 100%!
Indeed. Headways (gaps between trains) are much more important with an intensive metro operation. That is also a performance metric that MTR / RfLI are measured on.
True of any railway performance metric at any time, for any company over the past 20+ years though.
The figures aren't as poor as portrayed although Thursday 29th was awful for the reasons mentioned in this thread (signal failure TCR and the swan suspension).
This may be true within the Core, but for example I do various journeys from Canary Wharf to Heathrow or to Slough, and it then becomes more significant that the full service, and range of final destinations, is presented. Even mismatches between Core services to Abbey Wood or Shenfield can be an issue leading to platform overcrowding.
That's just because a railway-centric portrayal of figures is concerned with the trains rather than the passengers being delayed. There are a range of calculation techniques to convert the former to the latter, unfortunately all they do is depress the reported figures further so it doesn't get done.
Last edited:
redreni
Member
In my experience of commuting between Abbey Wood and Farringdon two days a week since the line opened, the good service / minor delays / severe delays information TfL puts out is actually useless. There is no correlation between what they warn me to expect and the extent to which I'm actually delayed.
On two of my last six evening journeys a good service has been advertised but I've waited between 10 and 14 minutes for an Abbey Wood train. Conversely, as you say, there have been times when they've reported severe delays and I've turned up, boarded a train and the journey time has been completely normal.
If putting out the "severe delays" message serves any purpose at al, it would only be to try to reduce passenger numbers to facilitate service recovery. It's not helpful to regular passengers because you know there's every chance your journey will be fine, ergo you're not going to divert to a slower route.
Bridge strike at Hanwell causing severe delays between Whitechapel and Reading/ Heathrow. Meant for some reason there were no trains at most of the COS stations more so on the Woolwich Branch.
Whilst it has nothing to do with RFLI infrastructure one still wonders why some shuttle services or terminations weren't done to keep at least some service shuttling between. Paddington and Abbey Wood.
Whilst it has nothing to do with RFLI infrastructure one still wonders why some shuttle services or terminations weren't done to keep at least some service shuttling between. Paddington and Abbey Wood.
Indeed. One of the stated reasons why auto-reverse "had" to be implemented, at considerable cost and delay, and despite the reversing per hour at Westbourne Park being no more frequent per hour than on the simple Abbey Wood section when the Core first opened, was that if there was a hangup westwards on the GWML then the entire Core service from both Abbey Wood and Shenfield would have to be turned round at Westbourne Park, which was "impossible" without auto-reverse.
And yet, as in this morning's peak (again) eastbound at Canary Wharf, when some problem arises on the GWML the whole service on the Abbey Wood line is dropped.
I suppose it depends partly on how many trains were in that section at the time. Creating shuttles is a huge endeavour, so they would eventually come back to start at Paddington, but at the booked time. So that would cause gaps, which should have been evened out over time (ideally as soon as possible) but I guess depends how long the suspension would be.
And as always with these things, the impact of crew displacement.
Edit: Indeed between about 0815-0900 there were no available trains at Abbey Wood having all the stock stuck on the GWML. This is why the Abbey Wood branch is disproportinately affected. The same thing can happen on the GEML, but of course that serves the "core" section, so the Abbey Wood trains at least provide some frequency in that area.
Last edited:
There must surely be some predetermined broad 'game plan' worked out, so that in the event of any blockage on any section of the system it can immediately be put into action, reversing trains, diverting some Shenfields to Abbey Wood or vice-versa, etc. Is this what years of planning was not meant to have ready on the shelf to be taken down as required?
Diverting Shenfield trains to Abbey Wood causes massive crew displacement issues. You could maybe do one or two and then you're importing disruption to the GEML. All trains left Abbey Wood ran as booked (to Paddington) until there were no more to run. Trains were reversed and put back into platforms on the GWML. So opt
And no plans cannot "immediately be put into action" when a chunk of your trains are the wrong side of the blockage (nor can plans be put in anywhere "immediately"!). Years of planning is all well and good, but still dealing with legacy infrastructure owners on GEML/GWML where things go wrong. GWR were also in major disruption and contingencies are all well and good but every incident is unique. Would expect a shuttle was looked at and deemed not possible for several reasons.
It's one of those things when a key part of the route is blocked for ~45 minutes in the morning peak; serious disruption is inevitable. It actually seemed to recover not too badly.
Last edited:
Thank you for the advice.
The resources I'm fetching from the RTT API only provide times a half minute resolution, but yes, I've classified anything under 5mins late as punctual (and labelled it "Ontime"). If all services arrive within 5mins of their booked time, then the quasi-PPM I've genrated for the day would be 100%. I appreciate Wikipedia is not an authoritative source, but I based this on the article here - https://en.wikipedia.org/wiki/Public_performance_measure.
Is there a term used within the industry for a service meeting PPM requirements, other than "on time"? Possibly "punctual" would avoid mis-using the "on time" term and be useful here?
I created the tool based on my own usage. I mainly travel at those times and on the "core" section between Paddington & Abbey Wood. It was only after finding the results over time fascinating I thought I'd share it with the forum. I think the data is still enough to give a representative overview of performance.
By the way - if you want to check the services that ran on a particular day, you can click on the down arrow next to the date to see a full list.
PPM is not later than 5 minutes at destination and serves all booked stops.
Right Time is within 1 minute at each station stop.
Snow1964
Established Member
TfL Board Elizabeth Line Committee papers for next weeks meeting are now published
Usage data is available. Lots of graphs of times of day and days of week etc
Plenty in this paper (section 4) on reliability
This is link to all the papers of the meeting
Usage data is available. Lots of graphs of times of day and days of week etc
Plenty in this paper (section 4) on reliability
This is link to all the papers of the meeting
Lots of interesting usage data, of which I found this interesting. Shows the slower alternatives available which people have now ditched (although I think some people have moved to certain places along the line in advance of it opening delayed!)
mr_moo
Member
Very useful and interesting summary Thanks!
- Status