Southern suburban services slower than 50 years ago.

[Bald Rick]

Yes, but how often are main line services crush loaded, apart from at their London termini when dwell time doesn't matter, and at how many places do they have a substantial crush load interchange on and off, like the Jubilee Line at Waterloo?

Regularly crush loaded. And with 2 doors per coach (and narrow doors on the 455s at that), it only needs one way flow fir there to be long dwells. For big two way flows, off the top of my head:

Clapham J, Wimbledon, Richmond, Balham, E Croydon, Denmark Hill, Lewisham, and London Bridge (6-9).

Notably the Elizabeth line has 60 second dwells throughout the core, even with 3 sets of door per coach side, and they are often needed.

 

[Craig1122]

I think the same is true across much of the South East. I’m pretty sure the Hastings - Charing Cross line is slower than it was.

I can compare 1972 for Hastings (Possibly not fair because that's pre electrification). Saturday times because they're more consistent.

It was 101 minutes every hour in 1972

The current service is either a fast at 92 minutes or a stopper taking 104. Etchingham and Wadhurst had no service for most of the day in 72 and the trains were non stop at London Bridge which would be unthinkable now. High Brooms has also gained an all day service to London. There's also now a substantially earlier start on weekdays and later finish.

So overall the current timetable is very much better. I think it's easy to forget that higher frequency is often effectively a journey time reduction for anyone who can't time their arrival at the station exactly

 

[Peter stanley]

I can compare 1972 for Hastings (Possibly not fair because that's pre electrification). Saturday times because they're more consistent.

It was 101 minutes every hour in 1972

The current service is either a fast at 92 minutes or a stopper taking 104. Etchingham and Wadhurst had no service for most of the day in 72 and the trains were non stop at London Bridge which would be unthinkable now. High Brooms has also gained an all day service to London. There's also now a substantially earlier start on weekdays and later finish.

So overall the current timetable is very much better. I think it's easy to forget that higher frequency is often effectively a journey time reduction for anyone who can't time their arrival at the station exactly

Yes I would much rather have say, two trains per hour slightly slower, than have just one per hour fast.

 

[Taunton]

Notably the Elizabeth line has 60 second dwells throughout the core, even with 3 sets of door per coach side, and they are often needed.

As a daily Liz user I can report how tedious one finds such timewasting stops in the core, sat there (well, standing, actually) at stop after stop, way after passenger movement is completed with nothing going on.

 

[stuu]

Notably the Elizabeth line has 60 second dwells throughout the core, even with 3 sets of door per coach side, and they are often needed.

As a regular user of the EL from CW to Paddington, I have never experienced a time when the dwells seemed like they were needed. Off peak especially it feels glacial. Once I was on a train when there was disruption and the train left as soon as possible and it made up about 3 minutes running time

 

[Snow1964]

What do you mean by commercial reasons?

Originally there was a bit of spare (not scheduled to 100% of acceleration and braking), something like 5% extra (it's so long ago I forget).

Then from late 1980s or 1990s statistics on percent of late of trains came in, initially when data was input manually it was only taken at final stop so got padding out of timetable (Waterloo-Vauxhall was say 4 minutes, but Vauxhall+Waterloo jumped to 9 or 10 minutes), so could be 5 minutes late all intermediate stations and still be on time when arriving at terminus.

Later as more automatic logging of times became available was extended to doing stats at all stations. Subsequently delay repay came along, and guess what more slowing to avoid payouts.

Recently all the operating companies have to publish on their website performance data for previous 4 week period, and now changing from 5 or 10 minute late being deemed on time to 3 minutes.

The commercial part is the slowing of trains and having extra long station stops to avoid delay payouts, which seems to have become more important than fast service, or increasing revenue by selling fast efficient journeys

 

[Harpo]

Subsequently delay repay came along, and guess what more slowing to avoid payouts.

General refunds for poor performance began after John Major’s 1991 Citizen Charter, translated on BR into the Passenger Charter.

 

[Magdalia]

Regularly crush loaded. And with 2 doors per coach (and narrow doors on the 455s at that), it only needs one way flow fir there to be long dwells. For big two way flows, off the top of my head:

Clapham J, Wimbledon, Richmond, Balham, E Croydon, Denmark Hill, Lewisham, and London Bridge (6-9).

In my experience long dwells are a significant issue at stations with big two way flows especially Clapham Junction, Wimbledon and Richmond. The refurbish of the South Western class 455s did help to manage the issue with the narrow doors.

 

[jfollows]

Not southern, I know, but I compared Manchester-Macclesfield stoppers (Stockport, then all stations) today - 31 minutes - with my regular travel back in 1977 - 31.5 minutes. Dwell times are certainly longer, so presumably better acceleration and higher speed compensates.

Today’s peak service frequency is virtually non-existent in comparison, though, one extra train in the day to the otherwise hourly service.

 

[williamn]

I can compare 1972 for Hastings (Possibly not fair because that's pre electrification). Saturday times because they're more consistent.

It was 101 minutes every hour in 1972

The current service is either a fast at 92 minutes or a stopper taking 104. Etchingham and Wadhurst had no service for most of the day in 72 and the trains were non stop at London Bridge which would be unthinkable now. High Brooms has also gained an all day service to London. There's also now a substantially earlier start on weekdays and later finish.

So overall the current timetable is very much better. I think it's easy to forget that higher frequency is often effectively a journey time reduction for anyone who can't time their arrival at the station exactly

Interesting, thanks! I think I was thinking since electrification. I’m pretty sure Wadhurst to London was under the hour, and now it’s always over an hour.

 

[Bald Rick]

The commercial part is the slowing of trains and having extra long station stops to avoid delay payouts, which seems to have become more important than fast service, or increasing revenue by selling fast efficient journeys

It really isnt a commercial decision ‘to avoid payouts’. Longer dwell times are there to reflect the reality of train running and large numbers of passengers. We want the trains to run on time!

 

[GordonT]

Comparing the working timetables from 1975 with today, South London suburban services are about 20-25% slower.

I wonder how that compares with equivalent journeys by road in 2025 compared to 1975?

 

[RT4038]

The commercial part is the slowing of trains and having extra long station stops to avoid delay payouts, which seems to have become more important than fast service, or increasing revenue by selling fast efficient journeys

Therre is no point in trying to sell 'fast, efficient journeys' if they are not consistently achievable, and not being consistently achievable is unlikely to materially increase revenue but instead end up with disappointed customers.
Far better to sell an achievable most of the time but slower service.

 

[Horizon22]

Building resilience into the timetable to avoid having to pay out for delays.

Which is no bad thing. Greater Anglia is one of the best operators currently and has a timetable with plenty of allowances. People get to where they want to based on journey planner / timetable they have looked at. And maybe even slightly early as a nice bonus.

It’s not all a conspiracy by the sneaky TOC.

== Doublepost prevention - post automatically merged: 1 Aug 2025 ==

As a regular user of the EL from CW to Paddington, I have never experienced a time when the dwells seemed like they were needed. Off peak especially it feels glacial. Once I was on a train when there was disruption and the train left as soon as possible and it made up about 3 minutes running time

If the train is even slightly delayed, CBTC will automatically retime the dwells lower to reduce delay.

“Glacial” is a bit extreme - you could maybe chop 15-20 seconds off-peak. But a consistent journey time throughout the day also has benefits.

 

[The Planner]

Originally there was a bit of spare (not scheduled to 100% of acceleration and braking), something like 5% extra (it's so long ago I forget).

Then from late 1980s or 1990s statistics on percent of late of trains came in, initially when data was input manually it was only taken at final stop so got padding out of timetable (Waterloo-Vauxhall was say 4 minutes, but Vauxhall+Waterloo jumped to 9 or 10 minutes), so could be 5 minutes late all intermediate stations and still be on time when arriving at terminus.

Later as more automatic logging of times became available was extended to doing stats at all stations. Subsequently delay repay came along, and guess what more slowing to avoid payouts.

Recently all the operating companies have to publish on their website performance data for previous 4 week period, and now changing from 5 or 10 minute late being deemed on time to 3 minutes.

The commercial part is the slowing of trains and having extra long station stops to avoid delay payouts, which seems to have become more important than fast service, or increasing revenue by selling fast efficient journeys

Why is there such a conspiracy theory over schedules and delay repay?

 

[Peter stanley]

I wonder how that compares with equivalent journeys by road in 2025 compared to 1975?

Interesting point. I would guess long distances are quicker due to motorways but shorter distances such as between central London and Croydon with no motorway option certainly seems to take longer with more congestion. Maybe checking old bus timetables would give some indication.

 

[perstreperous]

Interesting point. I would guess long distances are quicker due to motorways but shorter distances such as between central London and Croydon with no motorway option certainly seems to take longer with more congestion. Maybe checking old bus timetables would give some indication.

You're right. In the early 1990s Herne Hill-Leatherhead was viable by car - I could leave at 0815 and get there by 0900. Now it is only viable if I set out before 0700, and coming back is unviable at any time. Which is why I got rid of my car. (I have worked on and off with that journey for 35 years).

Simple population increase is probably much of the reason (in my census area, 36 per cent in 30 years ... it is not obvious that the late 1980s/early 1990s were a post-war low point for London).

 

[Peter stanley]

Originally there was a bit of spare (not scheduled to 100% of acceleration and braking), something like 5% extra (it's so long ago I forget).

Then from late 1980s or 1990s statistics on percent of late of trains came in, initially when data was input manually it was only taken at final stop so got padding out of timetable (Waterloo-Vauxhall was say 4 minutes, but Vauxhall+Waterloo jumped to 9 or 10 minutes), so could be 5 minutes late all intermediate stations and still be on time when arriving at terminus.

Later as more automatic logging of times became available was extended to doing stats at all stations. Subsequently delay repay came along, and guess what more slowing to avoid payouts.

Recently all the operating companies have to publish on their website performance data for previous 4 week period, and now changing from 5 or 10 minute late being deemed on time to 3 minutes.

The commercial part is the slowing of trains and having extra long station stops to avoid delay payouts, which seems to have become more important than fast service, or increasing revenue by selling fast efficient journeys

Whether longer dwells are to build resilience or avoid delay payouts, my recollection (admittedly subjective and unscientific) is that punctuality was better in the 60s and 70s. I travelled regularly and lived close to a railway and recall that you could almost set your clock by the passing trains - being more than two or three minutes late was a rarity. Major disruptions very rare, only after heavy snow or a derailment eg freight train at Streatham common 1965.

== Doublepost prevention - post automatically merged: 1 Aug 2025 ==

Originally there was a bit of spare (not scheduled to 100% of acceleration and braking), something like 5% extra (it's so long ago I forget).

Then from late 1980s or 1990s statistics on percent of late of trains came in, initially when data was input manually it was only taken at final stop so got padding out of timetable (Waterloo-Vauxhall was say 4 minutes, but Vauxhall+Waterloo jumped to 9 or 10 minutes), so could be 5 minutes late all intermediate stations and still be on time when arriving at terminus.

Later as more automatic logging of times became available was extended to doing stats at all stations. Subsequently delay repay came along, and guess what more slowing to avoid payouts.

Recently all the operating companies have to publish on their website performance data for previous 4 week period, and now changing from 5 or 10 minute late being deemed on time to 3 minutes.

The commercial part is the slowing of trains and having extra long station stops to avoid delay payouts, which seems to have become more important than fast service, or increasing revenue by selling fast efficient journeys

Apart from less delays, I also also remember cancellations were almost unheard of .

 

[sh24]

Why is there such a conspiracy theory over schedules and delay repay?

As with most conspiracy theories, my hunch is they provide a simple answer to explain rather than acknowledging that the world is complex, difficult and hard work!

 

[185]

At least these services don't have the "ceremony of the doors" a la Northern. If they did each stop would take 2 mins!

To test this theory:

Just looked at a train at Guide Bridge, 2G12, Northern EMU. It had a dwell time of 30 seconds. 1141.45 departing 1142.15.

And to add balance 2H61 currently at Woodley, Northern DMU. It too had a 30 second dwell time. 1151.30 departing 1152.00.

This Northern door ceremony you mention. It seems pretty good.

Those DOO services down south should be faster and more efficient. But are they? Simple fact, is that longer trains can take longer to dispatch.

 

[Horizon22]

Whether longer dwells are to build resilience or avoid delay payouts, my recollection (admittedly subjective and unscientific) is that punctuality was better in the 60s and 70s. I travelled regularly and lived close to a railway and recall that you could almost set your clock by the passing trains - being more than two or three minutes late was a rarity. Major disruptions very rare, only after heavy snow or a derailment eg freight train at Streatham common 1965.

== Doublepost prevention - post automatically merged: 1 Aug 2025 ==


Apart from less delays, I also also remember cancellations were almost unheard of .

No offence, but this seems like the epitome of rose-tinted glasses!

 

[Deepgreen]

Building resilience into the timetable to avoid having to pay out for delays.

Exactly that - thanks!

== Doublepost prevention - post automatically merged: 1 Aug 2025 ==

No offence, but this seems like the epitome of rose-tinted glasses!

Agreed - on a good day, journeys were quicker in general, and most people were fine, but slam doors opening in motion (for boarding and alighting!) were fundamentally unsafe. However, humans deal with their existing reality well and accidents were far less frequent than might be expected - no platform yellow lines; people just knew to stand back, and passengers opening doors in motion knew to check for people on the platfom before opening wide. I recall the norm of unlatching doors on entering the platform at Waterloo, etc., and stepping off at 3-4mph every day without incident, but it was a young man's game!

== Doublepost prevention - post automatically merged: 1 Aug 2025 ==

There is also the signifiant impact of door release times - the Desiros, for example, are dire, and on a long suburban run this makes a significant difference.

 

[Snow1964]

I happen to have the South West division timetable public booklet for 1979 and virtually everything has been slowed down in SWR land.

It was probably the fastest timetable, as it followed the increase in speeds 65mph to 75mph in New Malden area, Brockenhurst area resignalling, Feltham area resignalling.

Of course at the time much of SW inner London still used the lower 660v (not the 750v nominal after 4SUBs went), and 12% lower voltage was not going to help dc motored trains reach top speed.

Talking of top speed it was only 90mph, and a lot was only 60-80mph then. And they had very low acceleration as they approached top speed often taking many miles to get to max speed. braking from 90mph took longer as trains were tread braked. Now you have trains that can rapidly accelerate to 100mph and are disc braked.

So basically have quicker trains, but slower dwell times, and lots of hanging about. Which explains why some stations are as much as 35% slower to/from Waterloo than 46 years ago.

 

[LLivery]

It's noticeable over the years and as well as vs Kent services

The Sydenham stoppers are now timed to take 10 mins from NXG to London Bridge, so 5-6 mins extra on the up vs down. The London Bridge service has just got more and more unattractive.

 

[Peter stanley]

It's noticeable over the years and as well as vs Kent services

The Sydenham stoppers are now timed to take 10 mins from NXG to London Bridge, so 5-6 mins extra on the up vs down. The London Bridge service has just got more and more unattractive.

Yes it's often the last part of the journey that takes the longest - gives more chance for late-running trains to "catch up" thus making the stats look better. Another factor may be platform availability at London bridge on the Brighton (terminating) side.

== Doublepost prevention - post automatically merged: 1 Aug 2025 ==

It's noticeable over the years and as well as vs Kent services

The Sydenham stoppers are now timed to take 10 mins from NXG to London Bridge, so 5-6 mins extra on the up vs down. The London Bridge service has just got more and more unattractive.

You may be interested to know that in 1975 the Sydenham to London Bridge journey took average of 15 mins with New Cross Gate to London bridge taking 5 to 6 minutes.

 

[Harpo]

I get the safety aspects of increased dwell but not the poor design of door operation and despatch that has engineered these inefficiencies, both in terms of build and ops process.

The underground has always been hugely focused on journey cycle reductions to increase frequency. If 2 mins can be saved on the cycle of a 2 minute headway line, either one less trainset is needed or 1tph can be added from the same resources.

The ex-BR network has gone in the opposite direction.

 

[Bald Rick]

Yes it's often the last part of the journey that takes the longest - gives more chance for late-running trains to "catch up" thus making the stats look better.

the stats are measured at every station, not just the terminus. There is no incentive* to stretch out the last part of the journey for the stats.

*except for viewers in Scotland and on TfL rail servcies

 

[Peter stanley]

the stats are measured at every station, not just the terminus. There is no incentive* to stretch out the last part of the journey for the stats.

*except for viewers in Scotland and on TfL rail servcies

So why do it? I can understand it for an operating reason eg lack of platform availability, but it does also reduce delay repay claims, given that most people on those services (Sydenham to London bridge) will be alighting at London Bridge not intermediate stations.

 

[3141]

The longer dwell times may just be because most trains only have 2 doors per carriage now (that open slowly) instead of something like 10 (as seen on the middle cars of an EPB or VEP)

Yes, and the design of a modern train, with large vestibules where the doors are, enables a larger number of passengers to be accommodated, so naturally they take longer to get on and off through the smaller number of doors.

I don't think the fact that these services take longer than they did 50 years ago is anything to get excited (or disappointed) about. More passengers, and more trains on the network, inevitably has consequences.

Comparing the Southern Electric area with the Underground isn't useful. The Underground itself is busier than it used to be, and dwell times in the peaks are often longer. But there'll be another train in two minutes, so passengers can choose not to get on this one. Also, efforts are made to speed things up: "This train is now ready to leave, please stand clear of the closing doors, stand clear of the closing doors...." Actually it's not ready to leave, people are still trying to get on, but it's coming to the end of its scheduled dwell time and so people are being encouraged to stop trying and wait for the next one.

 

[Bald Rick]

So why do it?

Because….

for an operating reason eg lack of platform availability

…and getting the timetable to work.

If you look at the timetable in Real time trains, you can see that almost all the allowances in addition to the base running time is pathing time. That is what is used to make the timetable fit together. It is not to make it easier to achieve performance stats / delay repay / etc.

For reference, of the 49 Southern service trains timed to arrive at London Bridge terminating platforms between 0600-1000, only 11 have performance allowances inbound of New Cross Gate, and the total for those 11 trains is 11.5 minutes. Performance allowance is what is used to account for minor operational issues that we know will happen - on average - but not where or when. For example passenger assistance, very heavy loadings on certain services, ‘passenger action’, etc.