Train Punctuality Needs Review, says Passenger Focus

[ANorthernGuard]

Just seen this on BBC News

http://www.bbc.co.uk/news/uk-12358860

I understand the need of punctuality (of course I do), but the timetable is set out for catch up times in numerous places, with wheelchairs/bikes/High volume of passengers alot of services will "lose" a few mins between same stations and then gain between others.

Personally I think the railway does very well most of the time, different TOC's Have different problems, my TOC's problem is mainly ancient stock and being held at Stations/Junctions/Signals for express Stock to pass.

Sometimes I do feel that no matter what the railway does, it will never be good enough for some.

 

[pemma]

The problem is where the recovery time is added in. Passengers for 8/10 stations on a line can experience a late arrival of around 15 minutes but the service is declared as "on time" as it reached it's destination within 5 minutes of it's scheduled time due to recovery time added in at the end.

 

[ANorthernGuard]

The problem is where the recovery time is added in. Passengers for 8/10 stations on a line can experience a late arrival of around 15 minutes but the service is declared as "on time" as it reached it's destination within 5 minutes of it's scheduled time due to recovery time added in at the end.

The trouble with that is where do you put the recovery time? almost every route has longish gaps between certain stations whereas others are only a minute apart, so where do they do it?

 

[pemma]

The trouble with that is where do you put the recovery time? almost every route has longish gaps between certain stations whereas others are only a minute apart, so where do they do it?

Simple answer is not between the final two stations on a service. Having it at that point means passengers for the final station will expect the train to arrive at it's destination early 90% of the time so when it does arrive "exactly on time" there's passengers who have very tight connections, while a lot of other passengers don't benefit from the recovery time as they've alighted before the final destination, so the only real beneficiare is the operator.

 

[ANorthernGuard]

Simple answer is not between the final two stations on a service. Having it at that point means passengers for the final station will expect the train to arrive at it's destination early 90% of the time so when it does arrive "exactly on time" there's passengers who have very tight connections, while a lot of other passengers don't benefit from the recovery time as they've alighted before the final destination, so the only real beneficiare is the operator.

on most of my routes there is make up time about half way through the route (varies quite alot tho) and always near Mcr Picc/Sheffield/Chester/Liverpool/Stoke/Crewe as most main stations have atleast7-8 mins before a 1st stop.

 

[Zoe]

If you don't put the recovery time between the last two stations then you are going to be extending overall journey times.

 

[Failed Unit]

For me it needs reviewing to let trains run late at times. On branchlines maybe give 10 / 15 minutes late start allowance from the mainline. The amount of times the 1924 Newark - Cleethorpes service has left the majority of it's passengers behind at Newark because it us not allowed to wait 5 minutes for a mainline connection is beyond belief. This is why everyone wants direct trains, because local trains can't be held. Yes sometimes this may cause conflicts but BR could sort these out. If a local service is held to connect to a mainline don't record it!

 

[Fred26]

I like the idea of padding, ...ish. I can see that it's good, in that if the train has a minute or so (depending on service) wait at a busy station, then passengers will have longer to board and thus not rush so much. However, it must be frustrating for passengers who sit there wondering why they're not moving for one, two or three minutes or even longer.
I've sat on trains a few times, wondering what we're waiting for, now I'm more aware, but it doesn't make me any less frustrated.

I don't know what the rules are, or even if there are any, for padding, but I can't see the overall harm in some padding. That is presuming that the 'catch-up point/s' are at busy stations, where it makes sense to wait. I think there needs to be a maximum amount of padding though, so that TOCs can't just take the Mickey.
--- old post above --- --- new post below ---

For me it needs reviewing to let trains run late at times. On branchlines maybe give 10 / 15 minutes late start allowance from the mainline. The amount of times the 1924 Newark - Cleethorpes service has left the majority of it's passengers behind at Newark because it us not allowed to wait 5 minutes for a mainline connection is beyond belief. This is why everyone wants direct trains, because local trains can't be held. Yes sometimes this may cause conflicts but BR could sort these out. If a local service is held to connect to a mainline don't record it!

Some branchline trains can be held. I'm not the most knowledgeable, but I'm fairly sure the Sudbury-Marks Tey line can be held for the train to/from Liverpool Street. I'm not sure of time-scales, but there must be other instances across the country. I'm sure there's someone out there who can correct me.

 

[Oliver]

Here is the original article.

http://www.passengerfocus.org.uk/news-and-publications/press-release.asp?dsid=4967

 

[pemma]

If you don't put the recovery time between the last two stations then you are going to be extending overall journey times.

Overall journey times were extended around 10-12 years ago when franchised operators started padding out the old BR times so that more trains ran on time. Since then ORR has set out a maximum allowed journey time for each service, which is significantly longer than the old BR times, although very few services use the maximum permitted time.

 

[ANorthernGuard]

Overall journey times were extended around 10-12 years ago when franchised operators started padding out the old BR times so that more trains ran on time. Since then ORR has set out a maximum allowed journey time for each service, which is significantly longer than the old BR times, although very few services use the maximum permitted time.

most of the services that I work if they took away padding would be late, so called padding can be used to benefit TOC's but to be fair they give a more honest account of how long a journey "really takes". A wheelchair takes about 2-3 mins to get the ramo out help pax on put ramp away etc. which as stations are becoming more accesible is happening alot more frequently (a good thing may I add), if TOC's stuck with the old BR rules every train would be about 5 late.

 

[ChrisCooper]

Holding branch trains tends to vary depending on circumstances. Low frequency trains or last trains will often be held, particularly where passenger numbers are high. At the end of the day the TOC's decision will be based on which is cheaper, hold the train, cause passengers to miss a train and therefore end up being delayed enough to claim refunds, or (particularly for last trains) have to provide alternative transport. It's quite common for the conductor to establish how many passengers wish to make a connection. Oviously if it's one or two then booking a taxi is easier, yet if it's a considerable amount then it's best to hold the train (especially as in some cases the train may end up pretty much empty). Length of the delay, crew hours and knock on effects will also come into it. Remember that by delaying the branch train you are delaying it's return and may cause passengers coming off the branch to miss their mainline connection, infact branch trains often have such short turn arounds that any delay would be very hard to recover from. Desisions are going to be made by controllers in conjunction with the train crew (who should have an idea of the numbers who wish to connect).

 

[mallard]

There are some obvious things that can be "fixed" with the punctuality reporting, such as making it based on the public timetable rather than the working one (which can be changed more-or-less at will).

Personally, I think that the times at all stopping points should count - something like the time a train takes from one station to the next should not exceed 105% of the timetabled time. This gives an extra 3 minutes per hour that the train takes, so on a long-distance journey, the overall allowance can be greater than the 10 minutes allowed now, but it can't just make up all the time at the end and be counted on time.

 

[Failed Unit]

on most of my routes there is make up time about half way through the route (varies quite alot tho) and always near Mcr Picc/Sheffield/Chester/Liverpool/Stoke/Crewe as most main stations have atleast7-8 mins before a 1st stop.

But looking at it another way, the 1750 Kings Cross - Leeds and the 1600 Edinburgh - London service both have a 7 minute connection at Newark North Gate with the 1924 Newark - Cleethorpes service. (last train to Cleethorpes and over 1 hour wait for Lincoln). If these trains are 8 minutes late people heading to Lincoln and Cleethorpes are over 1 hour late as the miss the connection (EMT refuse to hold) so to the passengers that have just missed the connection so what if it arrived at the destination on time!

However to be honest that problem would easily be solved by departing the Cleethorpes service later, it has to wait 7 minutes outside Barnetby waiting for a TPE service to clear its path, so if they moved the departure to 1931 then the problem would be solved, as in my experience the East Coast services are normally only about 10 minutes late max. (or on time )

 

[Fred26]

But looking at it another way, the 1750 Kings Cross - Leeds and the 1600 Edinburgh - London service both have a 7 minute connection at Newark North Gate with the 1924 Newark - Cleethorpes service. (last train to Cleethorpes and over 1 hour wait for Lincoln). If these trains are 8 minutes late people heading to Lincoln and Cleethorpes are over 1 hour late as the miss the connection (EMT refuse to hold) so to the passengers that have just missed the connection so what if it arrived at the destination on time!

However to be honest that problem would easily be solved by departing the Cleethorpes service later, it has to wait 7 minutes outside Barnetby waiting for a TPE service to clear its path, so if they moved the departure to 1931 then the problem would be solved, as in my experience the East Coast services are normally only about 10 minutes late max. (or on time )

Surely if the delay is over one hour and not due to the fault of the passenger, alternate transport should be offered.

 

[chuckles1066]

Sometimes I do feel that no matter what the railway does, it will never be good enough for some.

I am reliably informed by someone at my work, who commutes the same route that I used to before I switched back to road, that the 17.24 Bath Spa to Temple Meads service was 16 minutes late (count 'em) tonight.

How can a "high speed" 84 minute service (Paddington - Bath) be almost 20% late?

And, in my experience when I used the railways, that service is consistently late into Bath.

 

[Bald Rick]

There are some obvious things that can be "fixed" with the punctuality reporting, such as making it based on the public timetable rather than the working one (which can be changed more-or-less at will).

Punctuality stats do use the public timetable, and not the working timetable.

Practically every railway in the world uses some 'padding', or more accurately recovery time, just in different ways. The French plan TGVs to operate at approximately 10% below linespeed on the high speed lines, and run at linespeed to recover any delays. Most european railways have generous station dwell times on long distance services allowing them to catch up time there if delayed earlier. Long station dwell times eat up platform capacity, and unfortunately in this country we don't have enough of them where we need them (think York, Reading, Birmingham and, err, Ipswich).

In this country, maximum journey times and stopping patterns are set by the DfT in the franchise specification of each TOC, and then enshrined in the Track Access Contracts between TOC and Network Rail. The maximum journey times permitted are, believe it or not, little changed since privatisation. However what has happened is that some long distance trains call at more intermediate stations, therefore extending the headline end to end journey times of many long distance services. The classic examples of this are London - Norwich and London - Cardiff. The other thing that has happened is that as the network has become busier (60% more passengers in 15 years), station dwell times have had to be extended simply to get everyone on and off the trains. An extra 10 seconds on a commuter train with 12 stops is another 2 minutes.

Finally, of all the railways in Europe, we measure punctuality more consistently, more rigidly and in much more detail than any other. Which is why you will rarely see comparisons of punctuality between countries.

 

[mallard]

Punctuality stats do use the public timetable, and not the working timetable.

Practically every railway in the world uses some 'padding', or more accurately recovery time, just in different ways.

Apologies, some of the coverage of the modifications to the East Coast timetable last year gave the impression that time was added only to the WTT and that this was used for stats.

I have nothing against padding, however, it should be added evenly along the route, rather than all at the end. It's silly that a train that's 15 minutes late at its penultimate stop can be "on time" at the terminus. This is why I suggested a "105% rule". With this system, no matter what time a train leaves the previous station, it may only take a maximum of 105% of the timetable time to reach the next station.

e.g. Train is timetabled to call at station A at 10:00 and station B at 10:20. The train actually arrives at A at 10:05. Since the timetabled time between the stations is 20 minutes, the train has to arrive at station B within 21 (105% of 20) minutes to avoid being counted late at B (regardless of weather it is counted late at A). Thus, it must arrive by 10:26.

This way, a trains' lateness is recorded along its entire route rather than just the end, but it is not penalized multiple times for a single delay.

 

[Failed Unit]

Surely if the delay is over one hour and not due to the fault of the passenger, alternate transport should be offered.

That seems to be a bit hit and miss to be honest. It is claimed that by the time they get a coach organised and to Newark North Gate (and then onwards to Lincoln) then it is quicker to just wait for the next train. At that time of night very few taxi's are around, not enough for the amount of people making the connection. The Taxi's normally are waiting at Lincoln for the Market Rasen passengers. I have done it all to often and although the delay repay is nice, I would much rather get home on time. I was amazed that East Coast didn't change the arrivals of the trains to 1918 to break the connection to the EMT service as it must be costing them a fortune in Delay Repay, but credit to them they take it on the chin!

 

[chuckles1066]

e.g. Train is timetabled to call at station A at 10:00 and station B at 10:20. The train actually arrives at A at 10:05. Since the timetabled time between the stations is 20 minutes, the train has to arrive at station B within 21 (105% of 20) minutes to avoid being counted late at B (regardless of weather it is counted late at A). Thus, it must arrive by 10:26.

This way, a trains' lateness is recorded along its entire route rather than just the end, but it is not penalized multiple times for a single delay.

You ARE a politician and I claim my five pounds.

If it's late at stops A, B, C, D, E and F then it's late 6 times surely?

 

[Bald Rick]

Apologies, some of the coverage of the modifications to the East Coast timetable last year gave the impression that time was added only to the WTT and that this was used for stats.

Ah - I think we're saying the same thing. For many services, a couple of minutes are added to the WTT at the terminating station to provide the timetable that is published for passengers to use. The difference is known as the 'public differential' and makes it slightly easier to be punctual against the public timetable than the WTT. Ultimately it is the actual arrival time compared to that published which the passenger cares about.

Forgot to mention, the industry is looking at measuring intermediate station punctuality officially, albeit for the busier stations only. Punctuality is not recorded at every intermediate station, particularly on the more rural lines.

 

[mallard]

You ARE a politician and I claim my five pounds.

If it's late at stops A, B, C, D, E and F then it's late 6 times surely?

I'm thinking more like this:

Train is delayed between A and B (by 10 minutes beyond allowance say), but loses no extra time over the rest of the route to F. Train is counted as "late" for stations B, C, D, E, and F, with 10 "delay-minutes".

Train is delayed between by the same amount between A and B, but recovers all this time between C and D, arriving at F on time. Train is counted as "late" only at B and C, with 10 "delay-minutes".

This means that two stats would be produced: % station callings on-time and average delay minutes.

(Yes, this is a slight modification of the original idea.)

 

[The Planner]

There are some obvious things that can be "fixed" with the punctuality reporting, such as making it based on the public timetable rather than the working one (which can be changed more-or-less at will).

The PTT is skewed compared to the WTT. PTT times are often earlier than the WTT as a performance buffer aswell as it allows earlier departure from stops. All reporting systems work off the WTT anyway.

 

[Gareth Marston]

The PTT is skewed compared to the WTT. PTT times are often earlier than the WTT as a performance buffer aswell as it allows earlier departure from stops. All reporting systems work off the WTT anyway.

Alas passengers the customers don't have access to the WTT. Whilst we can all go blue in the face debating the technical intricacies of recovery time,padding/ WTT, PTT, whose fault it is etc etc - all the passenger sees is their train consistently being late for their journey - as they go by what their sold the PTT and then ATOC banging on about record punctuality. The wider consensus is were being conned.

 

[brianthegiant]

I'm thinking more like this:
% station callings on-time and average delay minutes.

These sound like much better indices than what we have at present.
What galls me at present is that a train can be late enough mid journey that you could miss an 'official' connection from NRE but it can still make up enough time to be counted as on time, that just doesn't seem right.

I actually think a BS/Euro standard for train punctuality would help. Then we would be able to actually compare how our punctuality rates compare with those abroad, my anecdotal impression is that our punctuality is actually better than a lot of our neighbours, and as a result this would help to quash those national stereotypes about how unreliable Britains railways are

 

[tbtc]

most of the services that I work if they took away padding would be late, so called padding can be used to benefit TOC's but to be fair they give a more honest account of how long a journey "really takes". A wheelchair takes about 2-3 mins to get the ramo out help pax on put ramp away etc. which as stations are becoming more accesible is happening alot more frequently (a good thing may I add), if TOC's stuck with the old BR rules every train would be about 5 late.

I agree with A Northern Guard

I'd rather have a timetable that is realistic and doesn't fall to pieces at the first bit of disruption.

 

[ANorthernGuard]

I agree with A Northern Guard

I'd rather have a timetable that is realistic and doesn't fall to pieces at the first bit of disruption.

Well those hope for us all yet lol myself & tbtc agreeing

 

[chuckles1066]

I agree with A Northern Guard

I'd rather have a timetable that is realistic and doesn't fall to pieces at the first bit of disruption.

From memory, I'm sure the 17.24 Bath - Bristol journey time is deemed to be 17 minutes.

On the sort of train used (a HST as we used to call them in my day) it's only going to take 9-10 minutes tops.

So it can get into Bath at 17.31 and still arrive in Bristol on time?

I assume that's the "padding" you refer to?