Historic number of passenger trains in the timetable

[Malcolmffc]

These days on a typical weekday there are about 22,000 passenger trains in the timetable (24,000 odd pre-Covid). Does anyone know roughly what the equivalent figure would have been at the pre-Breching peak?

 

[steamybrian]

I did read somewhere that pre- Covid there were more trains per day than in the 1960s.
Many (not all..!) branch lines that closed had a sparse service of 3 or 4 trains per day.
The frequency of many main line services have improved following electrification or dieselisation.

 

[zwk500]

I did read somewhere that pre- Covid there were more trains per day than in the 1960s.
Many (not all..!) branch lines that closed had a sparse service of 3 or 4 trains per day.
The frequency of many main line services have improved following electrification or dieselisation.

Much lower frequency but of course a much higher number of lines to serve, it would be interesting to see the number but you'd need either a lot of time or some good OCR to interrogate a Bradshaw's!

 

[Fazaar1889]

I mean, just from looking at the graph on Wikipedia, its obvious that there would be around half as much just before the beeching cuts.

 

[Dr Hoo]

Some of the BR lines with the largest number of trains have ‘transferred’ - Tyneside Electrics, Waterloo & City, Bury, Altrincham for a start.

 

[zwk500]

I mean, just from looking at the graph on Wikipedia, its obvious that there would be around half as much just before the beeching cuts.

Passenger numbers are not strictly tied to the number of passenger trains. Pre-Beeching a fairly substantial number of passenger trains were carrying single figure passengers for significant lengths.

 

[midland1]

The number of passenger train miles might be a better look i.e. 6 trains each doing 10 miles = 60 miles 1 train doing 100 miles = 100. There are far more long distance trains now than the early 1960s.

 

[Fazaar1889]

Passenger numbers are not strictly tied to the number of passenger trains. Pre-Beeching a fairly substantial number of passenger trains were carrying single figure passengers for significant lengths.

ah fair enough

 

[Roger1973]

I'm sure there has been a similar question in the past, but the 'number of trains' can be misleading.

Three examples that come to mind (there must be many more) -

For example - on the South Eastern suburban lines, at some times, there have been trains which are Cannon Street to Cannon Street via (for example) Crayford and Slade Green. At other times, an identical train working would have been one passenger train from Cannon Street to Crayford, then an empty train round the curve to Slade Green, and then another passenger train from Slade Green to Cannon Street.

I also have a vague memory from the late 1980s of regional railways (or was it still provincial sector then?) knitting journeys together so that there were some strange workings on the Matlock branch, like a through train (from memory) Llandudno to Matlock via Chester and Stoke on Trent, which at one stage would probably have been 3 or 4 separate trains.

And then there's the Thameslink routes where a Bedford - Brighton train may have replaced a Bedford - St Pancras train and a London Bridge - Brighton train, but the cross London bit has added to mileage and probably generated more passengers.

 

[Mogz]

I'm sure there has been a similar question in the past, but the 'number of trains' can be misleading.

Three examples that come to mind (there must be many more) -

For example - on the South Eastern suburban lines, at some times, there have been trains which are Cannon Street to Cannon Street via (for example) Crayford and Slade Green. At other times, an identical train working would have been one passenger train from Cannon Street to Crayford, then an empty train round the curve to Slade Green, and then another passenger train from Slade Green to Cannon Street.

I also have a vague memory from the late 1980s of regional railways (or was it still provincial sector then?) knitting journeys together so that there were some strange workings on the Matlock branch, like a through train (from memory) Llandudno to Matlock via Chester and Stoke on Trent, which at one stage would probably have been 3 or 4 separate trains.

And then there's the Thameslink routes where a Bedford - Brighton train may have replaced a Bedford - St Pancras train and a London Bridge - Brighton train, but the cross London bit has added to mileage and probably generated more passengers.

I remember a Grimsby-Chester that was usually a 156 in late RR/ early Central Trains days.

Was that one of them?

 

[D6975]

There were others too,
Wolves-Wakefield (via Chester)
Cardiff-Cleethorpes
were 2 of them.

 

[Rescars]

I'm sure there has been a similar question in the past, but the 'number of trains' can be misleading.

Three examples that come to mind (there must be many more) -

For example - on the South Eastern suburban lines, at some times, there have been trains which are Cannon Street to Cannon Street via (for example) Crayford and Slade Green. At other times, an identical train working would have been one passenger train from Cannon Street to Crayford, then an empty train round the curve to Slade Green, and then another passenger train from Slade Green to Cannon Street.

I also have a vague memory from the late 1980s of regional railways (or was it still provincial sector then?) knitting journeys together so that there were some strange workings on the Matlock branch, like a through train (from memory) Llandudno to Matlock via Chester and Stoke on Trent, which at one stage would probably have been 3 or 4 separate trains.

And then there's the Thameslink routes where a Bedford - Brighton train may have replaced a Bedford - St Pancras train and a London Bridge - Brighton train, but the cross London bit has added to mileage and probably generated more passengers.

Further to the point about knitting trains together, how should trains which split into portions for different destinations be accounted for? The Atlantic Coast Express as an example. Also, should slip coach workings be accounted for as separate trains?

 

[zwk500]

Further to the point about knitting trains together, how should trains which split into portions for different destinations be accounted for? The Atlantic Coast Express as an example. Also, should slip coach workings be accounted for as separate trains?

Personally, Slip coaches wouldn't be counted as an extra train unless they form a 1-coach train onwards on their own. If they terminate then they're not really forming anything, and if they're attached to a waiting branch train then they just move from one to the other. Similarly with through coaches - if the two trains are running anyway and a coach is moved from one the other it's not formed an extra train.

Portion workings are harder to decide on - e.g. is the Caledonian Sleeper 2 trains or 5 each way? I'd personally lean towards including portions as part of the same train, although I'd also be tempted to count them separately.

 

[Bald Rick]

And then there's the Thameslink routes where a Bedford - Brighton train may have replaced a Bedford - St Pancras train and a London Bridge - Brighton train, but the cross London bit has added to mileage and probably generated more passengers.

Whilst that feels correct, in this example it isn’t. In the train counts each Thameslink service through the core counts twice, once either side of Blackfriars. This is because they are of different ‘service groups’ either side, as per the track access contract.

There are other similar examples, but Thameslink is by far the biggest.



The numbers the OP quotes are “SX”, ie weekdays, but they vary day to day quite a bit and of course are lower on Saturdays and much lower on Sundays.

But some very approx numbers:

2010: 21,500
2011: 22,000
2012 - 2017: a gradual increase up to 22,500
2018: 23,000
2019: 24,000

 

[zwk500]

Whilst that feels correct, in this example it isn’t. In the train counts each Thameslink service through the core counts twice, once either side of Blackfriars. This is because they are of different ‘service groups’ either side, as per the track access contract.

There are other similar examples, but Thameslink is by far the biggest.

Thameslink trains are the same Headcode and schedule through from Brighton to Bedford (and so on) though, so it really depends what defines a 'train'.

 

[Bald Rick]

Thameslink trains are the same Headcode and schedule through from Brighton to Bedford (and so on) though, so it really depends what defines a 'train'.

In the train counts, it’s Headcode by service group.

 

[zwk500]

In the train counts, it’s Headcode by service group.

Which would be an interesting exercise to attempt to apply retrospectively.

 

[Surreytraveller]

Thameslink trains are the same Headcode and schedule through from Brighton to Bedford (and so on) though, so it really depends what defines a 'train'.

Yes, but each Thameslink train that goes through the core is two trains in the figures. So, if one train is cancelled, the figures actually show two trains cancelled.
Same with portion working. A train which splits is two trains. The second portion has its own Train ID anyway.

 

[Bald Rick]

Which would be an interesting exercise to attempt to apply retrospectively.

yes it was!

 

[DynamicSpirit]

Portion workings are harder to decide on - e.g. is the Caledonian Sleeper 2 trains or 5 each way? I'd personally lean towards including portions as part of the same train, although I'd also be tempted to count them separately.

I would also count portion workings in the same way if you are counting train-miles, so using your example the Caledonian sleeper would count as one train from London up to the point it splits, then two trains after it splits. That of course only works if you are (sensibly) counting train-miles rather than just how many trains are in the timetable. It also presumes you're not trying to take account of train length in your calculations - so a 12-carriage train counts the same as an old 1-carriage DMU. That's probably also sensible since there just won't be sufficient historical data on train lengths to do otherwise.

 

[Rescars]

I would also count portion workings in the same way if you are counting train-miles, so using your example the Caledonian sleeper would count as one train from London up to the point it splits, then two trains after it splits. That of course only works if you are (sensibly) counting train-miles rather than just how many trains are in the timetable. It also presumes you're not trying to take account of train length in your calculations - so a 12-carriage train counts the same as an old 1-carriage DMU. That's probably also sensible since there just won't be sufficient historical data on train lengths to do otherwise.

How should situations like Wick and Thurso be counted. Trains used to call at Georgemas Junction before going on to one terminus with a branch service to the other. AIUI trains now run to Thurso, then reverse and go on to Wick before heading back south towards Inverness. More economic use of rolling stock, but how many trains?

 

[nw1]

I mean, just from looking at the graph on Wikipedia, its obvious that there would be around half as much just before the beeching cuts.

Interesting to note that the biggest trough (aside from Covid) was in the early 80s, perhaps around 1982.

Strangely that was the year I started using the railways. The southeast had it pretty good at that time (and peak services were much better than today on many lines) but it seems apparent that other parts of the country had to make do with infrequent and/or irregular services.

The "NSE peak" around 1988-90 is apparent; I do recall the 1987-90 period as a general time of improvement. Sadly 1991 marked a decline, due I believe to the recession.

I would also count portion workings in the same way if you are counting train-miles, so using your example the Caledonian sleeper would count as one train from London up to the point it splits, then two trains after it splits. That of course only works if you are (sensibly) counting train-miles rather than just how many trains are in the timetable. It also presumes you're not trying to take account of train length in your calculations - so a 12-carriage train counts the same as an old 1-carriage DMU. That's probably also sensible since there just won't be sufficient historical data on train lengths to do otherwise.

If it were me I'd (for the sake of argument) count the front portion as a through train from the origin, and the rear portion as a separate "shuttle" (even if it isn't).

So for the old Basingstoke/Altons, one could perhaps consider that as a Waterloo-Basingstoke and a Woking-Alton. I believe that was what happened with train codes; the front portion would determine the train code from the origin, and then the rear portion would get its own train code after the split.

 

[DynamicSpirit]

How should situations like Wick and Thurso be counted. Trains used to call at Georgemas Junction before going on to one terminus with a branch service to the other. AIUI trains now run to Thurso, then reverse and go on to Wick before heading back south towards Inverness. More economic use of rolling stock, but how many trains?

To my mind, by far the best thing to do is to count train-miles rather than trains. That means the new arrangement would give you a small increase in miles for each Inverness-Wick train because of the extra mileage worked between Georgemas Junction and Thurso. And that makes sense because there actually are more trains per day between those two stations as a result of the reversal at Thurso en route to Wick.

If you're just counting how many trains there are in the timetable and ignoring miles, then it becomes a bit ambiguous and you'd have to make an arbitrary decision whether the split between Wick and Thurso portions that used to happen counts as one train or two trains. But to be honest, I don't really care because, as discussed by several people upthread, I don't see number of trains in the timetable as a sensible statistic to measure if you're not accounting for miles travelled.

 

[Irascible]

Passenger numbers are not strictly tied to the number of passenger trains. Pre-Beeching a fairly substantial number of passenger trains were carrying single figure passengers for significant lengths.

The population was about 51m, less urbanised ( national census data, not my conjecture ) and ( more my conjecture ) less likely to travel, which is why I don't really like these absolute comparisons - if I have spare time I must dig out some data from the late 70s-80s & see if there's any changes in urbanisation since then though, given that was pretty much the nadir of rail travel.

Passenger miles/hour would be an interesting stat.

 

[eldomtom2]

Perhaps the best solution would be to make maps showing frequency and end-to-end journey time.