What should the numbering layout of Edinburgh Waverley station be?

[Bletchleyite]

The circular numbering system worked reasonably well with the old layout, even if it did produce the odd 1/19 combination, as it was a simple clock-face. The “reopening” of the sub platforms was the only real complication. It’s the addition of extra through platforms with separate stairs etc that starts to make it less intuitive. Not sure what German style you mean - finding platforms 101 and 102 between 2 low-numbered ones always seemed weird (though I now know why).

The circular numbering always confused me, I'd have just done it linear like any other station to be honest. And having 8/9 East and West is just silly, nowhere else uses that, just have A and B. And if it's one continuous face just split it A and B, don't have the confusion of different numbers for different parts of the same physical platform.

Current map is here: https://www.networkrail.co.uk/wp-content/uploads/2023/08/Edinburgh-Station-Map.pdf

But my suggestion regarding German style is that you'd just group them into 3 groups based on roughly where they are in terms of access along the station i.e. from the New Town end:

Present 1/20: 101A/B
Present 2/19: 102A/B

Present 14-17: 201-204 (no A or B as they are just used singly)

Present 3 and 4: 301 and 302 (sorry, edited, I missed those!)

Present 5 and 6: 401 and 402
Present 12: 403
Present 7/11: 404A/B
Present 10: 405
Present 9: 406A/B
Present 8: 407A/B

That way the signing appears nice and clear, and similarly numbered platforms are near one another which the circular numbering fails on. It's one of those things that is a lovely system in principle but it doesn't help people navigate, as if they're on platform 1 they don't expect the other end of it to be platform 20.


You could just do that sequentially and it'd work though, aside from 3 and 4 which are a bit hidden - I had to edit this as I had missed them on the plan!

 

[mmh]

Hmm, East / West is silly because it's not used elsewhere (yes, it is silly), but Platform 406 isn't silly?

 

[The exile]

If all the through platforms are split then you could be even more radical (which is where my East and West comes in) - divide the station into an East and West end for numbering and start with 1 at each end - platforms 3 - 9 East and West wouldn’t actually meet in the middle, but that wouldn’t matter.
As an aside, I’ve just realised that there must originally have been 1 more bay at the East end than at the west - ironic considering the busy-ness of each end today.

 

[Bletchleyite]

Hmm, East / West is silly because it's not used elsewhere (yes, it is silly), but Platform 406 isn't silly?

The numbering is still sequential, that's the key. The circular numbering is a very rare concept and confuses people, as does the ridiculous situation of having platform 1/20 and 2/19 on the same faces, those should be one number and A/B ends and would be anywhere else with long platforms used in halves e.g. New St.

 

[The exile]

Hmm, East / West is silly because it's not used elsewhere (yes, it is silly), but Platform 406 isn't silly?

Sure north and south are used somewhere…

 

[Bletchleyite]

Sure north and south are used somewhere…

Is it York? I have seen it. But I would remove it. A and B are standard.

 

[The exile]

The numbering is still sequential, that's the key. The circular numbering is a very rare concept and confuses people, as does the ridiculous situation of having platform 1/20 and 2/19 on the same faces, those should be one number and A/B ends and would be anywhere else with long platforms used in halves e.g. New St.

I raise you Bristol TM !

 

[Bletchleyite]

I raise you Bristol TM !

At least the numbers are sequential there, but I'd agree it's equally silly. A standard system is best that works the same everywhere, and I'd advocate a platform face to only ever have one number, and if two trains use it often in defined places split it A and B.

 

[tbtc]

Hmm, East / West is silly because it's not used elsewhere (yes, it is silly), but Platform 406 isn't silly?

Exactly the point I was going to make

The status quo is as logical as the station layout permits

People always argue for clockface timetables (or whatever continental term you want to throw around to sound sophisticated)... Well, Waverley has a clockface layout

Looking north on a map, you start at 1 and 2 and go clockwise around to 19 and 20. What could be simpler? Platform 406 certainly isn't!

Announcements take long enough at it is, increasing their duration to specify Platform 407B doesn't help

The old 1/19 and 10/11 (2/19 and 7/11 in new money) make sense because they are each full length platforms for long distance trains - using an "A" suggests a half length platform - plus passengers are more likely to get confused between the similar sounding 101A and 101B than they are between 1 and 20

Considering the large number of passengers, including a lot of tourists, it seems to have worked perfectly well for longer than any of us have been alive

The OP's alternative would be incredibly confusing (AIUI the suggestion is that two of the current eastern bays - 3 and 4 - become 301 and 302 - but the other east end bays don't follow on with "303 and 304", instead 5 and 6 become 401 and 402?)... It would cost a lot to implement and not improve capacity whatsoever. Why bother?

 

[Tetchytyke]

It doesn't matter what you number a platform if the wayfinding signage is good enough. The problem at Waverley is that it isn't.

 

[Narrowgauge725]

Never had a problem so why change it

 

[Egg Centric]

It doesn't matter what you number a platform if the wayfinding signage is good enough. The problem at Waverley is that it isn't.

Yup.

If you're going to renumber platforms then I would instead rename them (perhaps to the names of Munros? I'd quite like to take a train from Schiehallion) so there isn't an implied order* or logic. It's then incumbent on someone who doesn't know the station to look for the name of the platform they want on the signs, rather than the current problem of reasoning based on the assumed relationships.

*yes I know you could do interpret them alphabetically or by height or something but most people wouldn't

 

[rmHawk765]

It doesn't matter what you number a platform if the wayfinding signage is good enough. The problem at Waverley is that it isn't.

This. As much as I don't want us having Platform 400s, signage is key.

 

[SuspectUsual]

And having 8/9 East and West is just silly, nowhere else uses that

Sure north and south are used somewhere…

Aberdeen platforms 6N/S and 7N/S if you want a Scottish example....

 

[Bletchleyite]

This. As much as I don't want us having Platform 400s, signage is key.

I think a logical order is as well, i.e. one end of the station to the other.

You could also do:

Present 1/20: 1A/B
Present 2/19: 2A/B

Present 14-17: 3-6 (no A or B as they are just used singly)

Present 3 and 4: 7 and 8

Present 5 and 6: 9 and 10
Present 12: 11
Present 7/11: 12A/B
Present 10: 13
Present 9: 14A/B
Present 8: 15A/B

3 and 4 are a bit oddly located for the sequence but provided the signage was good that isn't an issue, and you avoid the confusion of having two numbers for one platform face and of 20 being also 1.

I don't agree the A/B thing is wrong for a double full length platform. It works fine at say New St with two trains in at once - it is the standard approach for platforms with mid platform signals in the UK and so people will expect it. And if they're standing on 1A looking for 1B most will have a clue where to find it, but if they're standing on 20 looking for 1 they are going to be baffled that it isn't at the opposite side of the station.

 

[35B]

The order is entirely logical, and groups platform numbers together - Newcastle is exactly the same, and causes nothing like the same issues. While I'd be tempted to start the numbering scheme in a slightly different place (probably the current p3) to avoid the 1/20 & 2/19 framing (which is confusing*), otherwise the issue is a fundamental one about layout, compounded by bad signage.

I'd also draw a distinction between a normal full length platform sub-divided into 2 or 3 parts, and the the very long platforms like 2/19 which can take full length trains at each end - the significance of getting A/B wrong is much greater, and the unfamiliar will not necessarily realise the significance of their error until they have to start running

* Gloucester manages a similar very long split platform with 1/2; Bristol Temple Meads has also been mentioned.

 

[Russel]

The circular numbering always confused me, I'd have just done it linear like any other station to be honest. And having 8/9 East and West is just silly, nowhere else uses that, just have A and B. And if it's one continuous face just split it A and B, don't have the confusion of different numbers for different parts of the same physical platform.

Current map is here: https://www.networkrail.co.uk/wp-content/uploads/2023/08/Edinburgh-Station-Map.pdf

But my suggestion regarding German style is that you'd just group them into 3 groups based on roughly where they are in terms of access along the station i.e. from the New Town end:

Present 1/20: 101A/B
Present 2/19: 102A/B

Present 14-17: 201-204 (no A or B as they are just used singly)

Present 3 and 4: 301 and 302 (sorry, edited, I missed those!)

Present 5 and 6: 401 and 402
Present 12: 403
Present 7/11: 404A/B
Present 10: 405
Present 9: 406A/B
Present 8: 407A/B

That way the signing appears nice and clear, and similarly numbered platforms are near one another which the circular numbering fails on. It's one of those things that is a lovely system in principle but it doesn't help people navigate, as if they're on platform 1 they don't expect the other end of it to be platform 20.


You could just do that sequentially and it'd work though, aside from 3 and 4 which are a bit hidden - I had to edit this as I had missed them on the plan!

I've no idea what solution would be better, but I agree with you on this one, the numbering system used at Waverley, while on paper seems simple, in reality, is rather confusing if you're only an infrequent visitor such as myself.

I've been told on here before that it's not really an issue, but numbering platforms in a different way to almost every other station in the UK just seems passenger unfriendly.

Fortunately, whenever I'm catching a train out of Waverley, it'll be an XC service back down south which seem to use the same couple of platforms.

 

[Western Gaffer]

I've been told on here before that it's not really an issue, but numbering platforms in a different way to almost every other station in the UK just seems passenger unfriendly.

Surely Waverley has a fundamentally different layout from almost every other station in the UK, and hence the lack of consensus about what would be a better numbering scheme every time the subject comes up on here?

 

[35B]

Surely Waverley has a fundamentally different layout from almost every other station in the UK, and hence the lack of consensus about what would be a better numbering scheme every time the subject comes up on here?

Well said. I'm trying to think how many other stations have the same challenges

 

[Kingston Dan]

Well said. I'm trying to think how many other stations have the same challenges

It's essentially three stations in one (four if the old Canal Street tunnels were still in use) - an East terminus, a West terminus and a through station (or stations if you count the old platforms 20/21 - now 8/9? I think).

 

[YorksLad12]

It's the lack of consistency that bugs me. Some long platforms have two numbers (1/20, 2/19), others have letters (8E/W, 9E/W). It should be one or the other...

But then York does something similar with platforms 3 & 4 (with a huge gap between, admittedly) and A/B use on P5.

 

[35B]

It's the lack of consistency that bugs me. Some long platforms have two numbers (1/20, 2/19), others have letters (8E/W, 9E/W). It should be one or the other...

But then York does something similar with platforms 3 & 4 (with a huge gap between, admittedly) and A/B use on P5.

And in both cases that numbering tells you something really useful about the nature of those platforms that an A/B type option doesn't. From my limited (and sometimes confused!) experience of using Waverley, 8&9 are fundamentally different to 1/20 & 2/19 in their size and length - my Quail gives me the following lengths:
1 - 156m
2 - 261m
8 - 293m
9 - 289m
19 - 261m
20 - 156m

Even allowing that 1/20 is not that much longer than 8 or 9, it still has the crossover midway down the length - a major operating difference to 8&9

 

[Tetchytyke]

There's inconsistency all over the place.

Newcastle has the north/east bay is 1 then the through platforms are 2, 3, and 4. Until you get to the 'new' through platforms which are numbered 5/7 and 6/8 because those platform have crossovers halfway down them. Then the south/west bays are 9, 10, 11, and 12.

Bridlington has three platforms, which are numbered 4, 5, and 6.

Preston has platform 3C and 4C, both of which are bay platforms and not part of the through platforms 3 and 4. Same with platform 4C at Birmingham New Street.

As I've already said, I really don't think it matters what numbering system you use as long as you signpost it clearly. Edinburgh's issue is that the wayfinding signage is terrible.

 

[Bletchleyite]

Passengers however don't care about operational differences. They care about where their train is, and the confusion caused by having two systems is therefore inappropriate.

A and B is better, as unlike two numbers it allows you to deal with one train occuping both bits by not showing a suffix and thus it meaning the whole thing. Though it probably is better to add a note to the CIS as well saying "Train occupies both platform 12A and 12B" to avoid ultimate confusion.

== Doublepost prevention - post automatically merged: 18 Sep 2026 ==

Preston has platform 3C and 4C, both of which are bay platforms and not part of the through platforms 3 and 4. Same with platform 4C at Birmingham New Street.

Those are a bit like platform zeros though in that they're slot ins. The Preston ones I think used to just be parcel bays before they started reversing the Ormskirks and Colnes in them as the main station started getting busier with modern frequencies*, and New St was a new addition.

If you were renumbering you certainly wouldn't do that. But they are a lesser evil than having numbers not in sequence there, e.g. the Preston bays being 7 and 8 or something.

* Though Ormskirks did use 1 or 2 then pop off to the sidings to reverse for a bit before the whole hourly "Lancashire triangle" thing started, with Colne-Blackpool S being a through service using 1 in one direction and 2 in the other.

 

[Tetchytyke]

If you were renumbering you certainly wouldn't do that. But they are a lesser evil than having numbers not in sequence there, e.g. the Preston bays being 7 and 8 or something.

Perhaps, but look at Sheffield.

Sheffield achieves inconsistency even within the same station. The southern bay is numbered 2C even though it isn't part of platform 2, but the northern bays 3 and 4 are numbered separately.

 

[Bletchleyite]

Perhaps, but look at Sheffield.

Sheffield achieves inconsistency even within the same station. The southern bay is numbered 2C even though it isn't part of platform 2, but the northern bays 3 and 4 are numbered separately.

"Lots of stations have inconsistencies" != "if we renumbered a station it should have inconsistencies post renumbering".

It wouldn't surprise me given where it is if 2C was also a parcels bay later brought into passenger use like the Preston "C" platforms, but I don't know. You probably didn't need it when the Hope Valley stopper was a few a day and Intercities to London were about every two hours.

 

[35B]

Passengers however don't care about operational differences. They care about where their train is, and the confusion caused by having two systems is therefore inappropriate.

A and B is better, as unlike two numbers it allows you to deal with one train occuping both bits by not showing a suffix and thus it meaning the whole thing. Though it probably is better to add a note to the CIS as well saying "Train occupies both platform 12A and 12B" to avoid ultimate confusion.

But splitting a platform into A/B (say) implies a platform for a single train length, that may be sub-divided. The point about Waverley, or York, or Gloucester, is that the two "halves" of the platform are functionally separate to the passenger. Anyone who's endured the sprint at Doncaster when they've stood on platform 3 and found that this means the London end only will know what this is like. As a passenger, separate numbers make clear where you need to be in a way that A/B signage simply doesn't.

For that reason, I prefer the type of split at Waverley - I know which is p2 and which p19, and I can position myself accordingly. Put them into an A/B setup and that gets lost - and so do plenty of ordinary punters who know nothing about station ops.

I'm with @Tetchytyke in thinking the fundamental issue is signage; it might be better (but will never happen!) if the numbering scheme started in the bays rather than where it does, but that's the icing on the cake.

 

[Bletchleyite]

But splitting a platform into A/B (say) implies a platform for a single train length, that may be sub-divided. The point about Waverley, or York, or Gloucester, is that the two "halves" of the platform are functionally separate to the passenger.

It really doesn't matter. New St's A and B ends are mostly used separately other than double Voyagers and Euston trains - do you consider that confusing in and of itself (rather than the convoluted situation upstairs at the A end)? I would say it's not - it's certainly not more confusing than 1 and 20 being the same platform face!

I think here you're favouring the system over the eyes of the lay customer.

 

[35B]

It really doesn't matter. New St's A and B ends are mostly used separately other than double Voyagers and Euston trains - do you consider that confusing in and of itself (rather than the convoluted situation upstairs at the A end)? I would say it's not - it's certainly not more confusing than 1 and 20 being the same platform face!

I think here you're favouring the system over the eyes of the lay customer.

No, I'm reflecting my experience as an ordinary punter at different stations.

Personally, I'm not a huge fan of A/B where there's double ended operation at a platform but where the platform (like New St) is about the length of a full length train, it's probably the lesser of the evils. I'm defending the Waverley set up because, as a passenger with luggage, it is more of a help than a hindrance. If a train is in 19, then it's clear where 19 is; ditto 1. Number them as 1 and split into A/B, and it's less clear where to head - and if you take the obvious "stand in the middle" position, you'll be well away from both ends.

Consistency matters, but it's not all important - different places have different needs. Waverley is complicated because the layout is complicated in the first place - no scheme can be ideal. The important thing, which gets lost sight of in the attempt to define an ideal schema, is that it should be clearly signposted so that those not familiar with it can find their way round. Newcastle has a similar pattern, but when I used to use it regularly, it rarely caused problems because it was clearly signposted.

 

[Bletchleyite]

No, I'm reflecting my experience as an ordinary punter at different stations.

Personally, I'm not a huge fan of A/B where there's double ended operation at a platform but where the platform (like New St) is about the length of a full length train, it's probably the lesser of the evils. I'm defending the Waverley set up because, as a passenger with luggage, it is more of a help than a hindrance. If a train is in 19, then it's clear where 19 is; ditto 1. Number them as 1 and split into A/B, and it's less clear where to head - and if you take the obvious "stand in the middle" position, you'll be well away from both ends.

Where a layman would expect to find 19 is not however "the other end of 2". What you could do is make those 1/2 and 3/4 though I suppose.