Great Western Coffee Shop

If a train can make its journey in 100 minutes, should it be scheduled to take 90 minutes?  The 09:00 off Paddington to arrive into Bristol Temple Meads at 10:30 to give the fastest possible journey?  Hey - what a great idea.  Except:
a) It may get delayed by other trains on the way which aren't running to time
b) It may have to go slower some days because of engineering works
c) Passengers (awkward things at times) may take longer to board that usual
d) The train may not be performing to its best, or may be a slower model
e) On an ever-busier network, it may not be possible to schedule it cleanly
f) "Fastest possible" burns up much more fuel.
g) It may be sensible to add in stops to help fill the train and provide intermediate better services
h) If the train is fully loaded, will that mean it's going to be slower to accelerate?

It's probably a pretty rare occasion with so many trains and passengers around, and with an old and complex network yet one that has far fewer tracks than 60 years ago, for a train to be able to make it through in very best timing!   So schedules are eased back to allow longer end to end, but to be more reliable:
1. Some schedues include timings that are longer than really needed to run
2. Some station stops are generous
3. Extra minutes get added in as "recovery time" along the way
4. A "Working timetable" sometimes shows ealier planned arrivals and later planned departures than are in the "Public Timetable"

There is a tendency for allowances to be added towards the end of train's journey.   This allows the train to be delayed anywhere on its journey and keep on running, whereas "delay minutes" added in the middle of a journey are pretty useless if the delay's later on.   Addition at this point is, through, often frustrating to the traveller.  It means:
* Connections that could be made aren't offered / recommended
* People getting off short feel cheated at being told their train was on time when it was late at their station
* People making short end journeys (such as Keynsham to Bristol) have to allow proportionality much longer for their journey

Modern performance evaluation measures include timekeeping (as well as running at all), and that's usually (always?) at the final station.  [[Aside - is there a measure of train length, cleanliness, catering provision included?]].  And that leads to suggestions that additional time allowances in the final / latter legs of a journey are there to help produce good statistical results.  And it's also been suggested that those extra 5 minutes will sometimes reduce the delay to a train from (say) 33 to 28 minutes according to what can be normally achieved, and remove the need on occassions for compensation to be paid.

So - is that extra time there to produce a practical system, to produce something that looks good to the passengers, to allow more efficient / greener running, to increase performance figures, or to reduce conmpensation.  Is there too little or too much of it?  And does it matter in the end?

Edit to add a couple more reasons why a train can't run 'fastest possible' every time

.........judging by the statement given by the FGW manager on Dispatches last week, it's mostly about massaging performance figures and reducing compensation payments.

St erth to Penzance and vice versa the timetable shows clear padding.

St erth to Penzance is scheduled between 12 and 15 mins (except services that started from Plymouth or within Cornwall) but Penzance to St erth only between 7 and 9 minutes.

timetabled times need to be realistic and take into account likely delays en route etc.  A train that keeps to time and takes 90 minutes perceived as faster and better than a train that takes 90 minutes but is timetabled to take 85. 

But padding at the end of the journey is just about fiddling the figures and ought to be outlawed.

Yes I noticed them having a go at FGW about this, but other TOCs are worse, my last 2 trips into KX I have left Peterborough 9 late and arrived 2 early and left Stevenage 6 late and arrived 5 early making up 11 minutes on both occasions, the best one was my first trip on a pendolino, we left Rugby 4 late and arrived Euston 17 early!

Surely a few minutes allowance is realistic at the end of a 300 mile journey and gives a more dependable arrival time?

The reason it's added at he end is that if added at each stop, and it runs 'to time (ie without need of the padding) it will then sit occupying platforms each time iot stops. Add it all at the end, then just 'normal;' stops are required & catch up is all at the end of a journey.

XC adds it as they go, whereas most others just at the end. It's an industry issue, and as long as the times are shown in the public timetable, they aren't lying to passengers, so why not? With the compensation culture in this country (whether that's right or wrong is another topic altogether but worth having), I for one don't blame them & 10 minutes or so is neither here nor there to me - I would rather think I have a 90% chance of arriving at the time in timetable including padding than a 60% chance of getting there without padding, and needing to spend time claiming, frankly

I suspect it's an industry issue and they're all pretty much as bad as each other..........rather like the little kid getting caught in the orchard scrumping apples and saying "....but all my friends are doing it too!"  :D

Yes, it is an industry issue and some operators really do stretch it too far, one of the prime examples is the 11 minutes some Chiltern trains are allowed to get between Birmingham Moor Street and Snow Hill, a journey of half-a-mile which in reality takes about two minutes!

The best way to change it is to change the methods of how recording punctuality for compensation purposes are made.  If an average lateness at each station en-route was taken instead then that would stop the problem instantly, and we might see a proper balance between what's possibly journey time wise, what's sensible in terms of recovery time en-route, and what's commercially competitive.  Not to mention a fairer system for the travelling public who might not be travelling to the final destination of the train!

A couple of minutes might be fine.  But the problem with the current situation is that a train can be sufficiently late at intermediate stops for connections to be missed and yet still count as "on time" according to the industry definition thereof.  Such a train fails its customers but doesn't get a black mark in the stats and it is that disconnect that is the problem.  It makes the ToC look better than deserved. 

The airlines like padding aswell......Ryanair are the past-masters, but it does help their on-time stats to look very good.

You have to pay extra for padding with Ryanair, or sit on a bench with everyone else.

I don't object to airlines doing it because planes do not make intermediate stops.

they do, you know - ever flown to Australia, for example.

I have a tale to tell of my flight from Bristol to Los Angeles, via Schiphol, involving a DC10, a careless lorry driver. and a four-hour delay.

not generally though.  Would a LHR-SIN-MEL flight be recorded as late if it landed on time in MEL but was late into SIN?  I expect (but do not know) that each leg would have its performance recorded separately.

I doubt it's officially recorded at all.....simply because the data isn't needed, except for marketing

It is recorded on the "flight stats" website at least and made publically available there. 

I'd be surprised if the likes of the CAA, NATs and other official organisations didn't use the statistics as well.  You are of course entitled to compensation if your EU flight is delayed by more than 3 hours. 

Correct, although not if its only late at via point. If it was 3 hrs 1 late at the via point, yet only 2hrs 59 at destination its tough luck.

In the days of Plymouth based AirSouthWest I used the Newquay to Leeds Bradford service frequently. That service called at Bristol. A number of times it was late arriving at Bristol but on time in Newquay.

Like the trains then....

A commercial flight from Bristol to Newquay would file a flight plan, but would not necessarily be subject to having to negotiate a "slot" with air traffic control. It would not be discharging or embarking the whole plane-load of passengers at Bristol, so could, with luck, with no need for a fuel uplift, and with baggage handling at the ready, make up time on the turnaround there, taxy and be off the ground in just a few minutes. As it happened a number of times, the crew must know how to deal with it.

Many moons ago, there was a very obvious example of padding. A certain train (can't remember which or when) was Paddington to Plymouth on Mon-Thurs, but extended to Penzance on Fri.  Its Plymouth arrival time was something like 10 mins earlier on Friday, as that wasn't its terminating station that day.  So the additional 10 mins was likely to be only for making the final arrival more likely to be 'on time'.

The 1903 off Paddington by any chance?

You are thinking of the 1903 Paddington - Plymouth/Penzance, as someone has already mentioned, a few years back I think it was. Although the arrival time at Plymouth was the same throughout the week, you are thinking of the arrival time at Penzance. The arrival time into Penzance and at stations throughout Cornwall was the same throughout the week, whether it having been the HST through or the unit connection from Plymouth. Therefore as the unit was quicker stopping at all stations through Cornwall there was a high amount of extra time in the Public Timetable to bring it to the same time the HST on a Friday would arrive, which made St Erth to Penzance seem over double the normal journey time - frequently however it arrived early. This therefore mean't that on a Friday the HST was frequently made to look like it was late as it struggled with the unit public timetable, however then by Penzance it was back to the WTT arrival (and the unit public timetable arrival Monday-Thursday), making it arrive on time.

My train home this evening, the 2018 from Truro arrived in Truro at 2007. It is timetabled 26 mins from St Austell to Truro but a Voyager only needs about 15.
Its also then timetabled 12 mins from Truro to Redruth which a Voyager does in 9.

Allowing 3 minutes for passengers to get on and off and one minute per carriage for the Train Manager to walk along and close the doors  ;D ;D

Seriously, I'm always aware on some of the s*l*o*w*e*r Berks and Hants HST stoppers of just how painfully l*o*n*g they take.

This one doesn't fall into this category being a XC service.

Timetabled 15 minutes from St Austell. Departs SAU 1952, arrives TRU 2007.

Then booked to wait 11 minutes at Truro, I guess so it doesn't catch up the preceding service from London Paddington. There are, I believe, long signalling headways in that part of the world.

Timetabled 10 minutes from Truro to Redruth. Departs TRU 2018, arrives RED 2028.

The section between Camborne and chacewater is about 18 minutes I believe.

I saw it coming and ran oblivious to the timetables this evening, to find myself sat on the train for 10 mins!