Train Departing 5 Minutes Early

[Urban Gateline]

Qs. If a train is booked to "set down only" is this listed in the timetable ? If said train sets down and all passengers alight can it then depart ahead of schedule ? What about if it was pick up only and no passengers were on the platform ?

Yes there is mostly always an "s" next to the Station name where a service calls to set down passengers only.

Yes those trains can depart when ready, I think my record was nearly 55mins early on the Caledonian Sleeper that calls to set down only at Watford Junction on the way to London Euston.

This is the same with many Virgin Trains services after 11pm when they seem to have very generous timings to get between Stations or passing points. We are told that if the signal is clear and there are no more passengers to come to dispatch the Train.

As for "pick up only", these are just normal passenger services and they wait till the normal time, ie 40 sec before departure they get dispatched if they happen to arrive early!

 

[455driver]

My FGW North Downs line train this morning was approaching Redhill when the guard proudly announced that we were arriving 2 and a half minutes early. He should not have been proud, though, as this was partly because we left the previous two stations early!

What train was that then?
I have just checked every FGW arrival at Redhill this morning and none of them left their previous station early!

Any proof greatfully received or failing that an apology for your unfounded accusation!

2O26 (0724 arrival) had arrived 2 early having left Reigate 1 late at 0719.
2O28 (0830 arrival) had arrived 3 early having left Reigate right time 0823.
2O30 (0930 arrival) arrived RT having left Reigate 1 late at 0925.

 

[Deepgreen]

It was 2O28, which, as you say, left Reigate at 0823, which is not right time as its scheduled departure time is 0824. You can't make up 3 minutes' journey time in the short distance between Reigate and Redhill!

 

[louis97]

It was 2O28, which, as you say, left Reigate at 0823, which is not right time as its scheduled departure time is 0824. You can't make up 3 minutes' journey time in the short distance between Reigate and Redhill!

2O28 departed Reigate a mere 6 seconds early, assuming the offset for the departure is correct. The 3 minutes early comes from the fact there is a 2 minute passenger/working timetable differential in the arrival time. RTT data suggests therefore that it arrived Redhill 2 1/4 minutes early (0827¾ - this would display as 0827 outside of RTT) by the passenger timetable (0830), or 1/4 minutes early by the working timetable (0828).

Although the 0827¾ arrival was only ¾ by 4 seconds, the offset stating the train stopped at around 082756, obviously this is by no means for definite, but a good estimate.

 

[A-driver]

It was 2O28, which, as you say, left Reigate at 0823, which is not right time as its scheduled departure time is 0824. You can't make up 3 minutes' journey time in the short distance between Reigate and Redhill!

It left a few seconds before 0824. That's certainly not counted as early as it comes down to how accurate the seconds on the clock you use are. Anyone missing a train by a few seconds has only themselves to blame.

 

[Deepgreen]

I had no idea so many forum members were there on the train with me!

 

[A-driver]

I had no idea so many forum members were there on the train with me!

No, but many forum members can look up the time the train actually departed. Chances are that the clock you were looking at was a few seconds out. Even platform clocks can be out with each other.

 

[Antman]

It left a few seconds before 0824. That's certainly not counted as early as it comes down to how accurate the seconds on the clock you use are. Anyone missing a train by a few seconds has only themselves to blame.

Early is early even if it is 'only a few seconds'

 

[embers25]

Early is early even if it is 'only a few seconds'

No its not as for example SWT close doors 30 secs before and Virgin at Euston is 2mins and so they often leave early but that is not considered early as it is advertised that doors close. Can't remember if fGW have 40 secs notices on north downs but even if they don't a few seconds is hardly worth complaining about and numerous trains arrive minutes early due to end of route padding. It's completely the norm. I'm guessing you think 455driver overreacted but he didn't because your comment could have gotten the on board team in trouble if it wasn't possible for us all to check the validity of your statement.

 

[Tomnick]

I don't think anyone would claim that the berth offsets (which are used to estimate the departure time from the actual time that the description steps into the next berth) are accurate within anything like a six second tolerance either.

 

[A-driver]

Early is early even if it is 'only a few seconds'

No it isn't. Doors tart to close 30secs before departure. I don't then sit in the platform for 15secs with doors all locked up, I go.

 

[Tom]

I don't think anyone would claim that the berth offsets (which are used to estimate the departure time from the actual time that the description steps into the next berth) are accurate within anything like a six second tolerance either.

I would. For the general case train at a given station, they'll be within ±5 seconds easily. I've tested this at a number of stations and the variance between, say, a 5 car and a 10 car train is about 4 seconds in general for arriving times. For departure times, the variance is slightly larger but normally within 10 seconds at most locations. This is why in the most part, RTT goes for quarter minute granularity where available.

The primary exception to this are stations where the speed through the platforms is sub 30mph.

Edit: in the instance about 2O28, as louis97 has pointed out (he has access to the Realtime Trains internal database), it appears to have left about 6 seconds early. The train passed the starter signal at Reigate at 082416, and past the home signal for Redhill at 082644. I've just looked at a signalling diagram for the area and RG5 (the starter) is near enough at the platform ramp so the time at which the step triggered (given the set of points in its rear) would be within about what I expect for that departure time.

 

[djjawuk]

It left a few seconds before 0824. That's certainly not counted as early as it comes down to how accurate the seconds on the clock you use are. Anyone missing a train by a few seconds has only themselves to blame.

Unless they miss the train because of a late connection?

 

[Antman]

Unless they miss the train because of a late connection?

Exactly, it might only be 'a few seconds early' but it can be the difference between getting the train and missing it.

 

[The Planner]

I would. For the general case train at a given station, they'll be within ±5 seconds easily. I've tested this at a number of stations and the variance between, say, a 5 car and a 10 car train is about 4 seconds in general for arriving times. For departure times, the variance is slightly larger but normally within 10 seconds at most locations. This is why in the most part, RTT goes for quarter minute granularity where available.

I would too, an awful lot of berth offsets are set manually by empirical data, ie: there is a bloke who goes out and times it.

 

[A-driver]

Unless they miss the train because of a late connection?

Exactly, it might only be 'a few seconds early' but it can be the difference between getting the train and missing it.

No, if you miss it by a few seconds, late connection or not then it's bad luck. Doors start closing 30secs before departure and won't reopen. So if you are hoping to connect with the 1231 and you arrive on a late train at 1230 then you are looking at missing the 1231 as you only have 30secs to join it.

 

[PermitToTravel]

I would. For the general case train at a given station, they'll be within ±5 seconds easily. I've tested this at a number of stations and the variance between, say, a 5 car and a 10 car train is about 4 seconds in general for arriving times. For departure times, the variance is slightly larger but normally within 10 seconds at most locations. This is why in the most part, RTT goes for quarter minute granularity where available.

The primary exception to this are stations where the speed through the platforms is sub 30mph.

Edit: in the instance about 2O28, as louis97 has pointed out (he has access to the Realtime Trains internal database), it appears to have left about 6 seconds early. The train passed the starter signal at Reigate at 082416, and past the home signal for Redhill at 082644. I've just looked at a signalling diagram for the area and RG5 (the starter) is near enough at the platform ramp so the time at which the step triggered (given the set of points in its rear) would be within about what I expect for that departure time.

Very interesting, thanks. Is there a difference caused by the separation between the signal post and the change in track circuit?

 

[Tom]

I would too, an awful lot of berth offsets are set manually by empirical data, ie: there is a bloke who goes out and times it.

Indeed, at minimum 6 trains for most offsets for those set by NR. For that are added into RTT over and above what NR measured (and we have remeasured a lot of NR ones too), it depends on the area but, in general, we do go to the ground and measure it at least once or twice. There's a piece of software on my end that goes and calibrates these offsets against at least 30-40 services to see if they make sense.
--- old post above --- --- new post below ---

Very interesting, thanks. Is there a difference caused by the separation between the signal post and the change in track circuit?

It's the track circuit that matters for the berth step. Depends if the signal has an overlap or not in general. For newer schemes, there is normally an overlap. For older ones, not always the case. There are some locations where the signal is right at the end of the platform but the offset can be upwards of 35 seconds. (and then of course, there are cases where the signal at the end of the platforms are repeaters...so you can see 120-150 second offsets - but thats another story!)

 

[Tomnick]

Thanks Tom, I'll take your word for it! I can certainly believe that the figures will be fairly consistent for departures (which is what matters here), but wasn't quite so sure about arrivals - especially those with quite large offsets - where different driving techniques (and the effect, potentially, of the signal at the end of the platform) can produce quite different braking curves. It's interesting to learn how thoroughly the offsets are measured - my perception's always been, for the single auto-reporting location that we deal with, that the offsets are quite a way out! Do they differentiate between stoppers and non-stoppers, and between freight and passenger?

 

[Tom]

There is a single offset for all traffic so they normally represent an aggregation of all traffic. For stations, it'll tend to be stopping passenger trains more than non-stopping freights, etc, but really depends on the platform and the traffic volumes through said platform. The other thing to note is that TRUST does not round times, it simply truncates the seconds. So an arrival of 131258 for instance will be 1312, not 1313.

I've got plans to put different offsets in for at least different traffic types (e.g. passenger vs freight) in the short term into RTT - I'd like to do based on train length too in the longer term but that requires the data to be available.