Gaps in the locomotive class numbering scheme.

Grumpy said:

Why was there no D6000?
We had D1000, D2000 etc up to D9000 but no D6000. Was this intended for a class that was cancelled?

Click to expand...

In the pilot scheme there were seven different Type B/Type 2 locos, these were:

D5000
D5300
D5500
D5700
D5900
D6100
D6300

If the Baby Deltics, starting at D5900, had gone on to be a fleet of more than 100 locos, then the 101st would have been D6000.

The oddity is that D5000 were allocated 300 numbers, but the others only 200. Both the BR and Brush Type 2s burst through these barriers, hence the former restarting at D7500 and the latter at D5800.

xotGD said:

I've always thought it was strange that in pre-TOPS days class numbering always started with number zero.

Click to expand...

The BR Type 4s started at D1.

Falcon1200 said:

Is there not a slight difference in the meaning of the numbering systems though? The SP loco mentioned is number 3000, as was the first Class 08 before renumbering, and although it did not apply to that particular loco, 08000 would actually have been Class 08 number 000, which might be why such numbers were not used.

Click to expand...

Although that would depend on whether the system identified the class/subclass from the running number, which given some of the number ranges for subclasses seems less likely?
So, is the number just a number with the class number derived from a different field?

In any case 55000 for D9000 makes little sense in it's suggested nullity and would be open to confusion with an error?

TheHovisKid said:

In any case 55000 for D9000 makes little sense in it's suggested nullity and would be open to confusion with an error?

Click to expand...

Why is that, when 45000, 75000, 70000, etc were quite acceptable for steam locos - the latter being presented as the start of a new generation of locomotives.

Taunton said:

Why is that, when 45000, 75000, 70000, etc were quite acceptable for steam locos - the latter being presented as the start of a new generation of locomotives.

Click to expand...

Because, although used as a colloquialism in some parts of the country, the first two numbers didn't constitute a class element, it's just a serial number starting off from a round number for convenience.
Exactly the same as with the unrenumbered diesels, not the renumbered ones... Chalk & Cheese.

You could design excel to have a row 0, but it'd cause more confusion than it solves.

There's a difference between "cannot" 000 and "should not because it seemed more intuitive not to" in the renumbering, and throwing up examples of "well, they did it" isn't relevant without that context. We don't know if it can or can't, just that they didn't.
Well, I don't know, do you?
But that's not changed by saying that my library card or the barcode on my lunchtime sandwich ends in multiple zeroes.

And if it's a policy matter, it ought to be easy enough to apply validation to prevent it.

Taunton said:

Why is that, when 45000, 75000, 70000, etc were quite acceptable for steam locos - the latter being presented as the start of a new generation of locomotives.

Click to expand...

If, for example, Britannias as a class had been deemed to be class 70 thousand, loco number 0 in that class might well have become loco number 1.

I have come across (on line) a locomotive number 0 - an 0-4-0ST of the Tangxu Railway in China, preserved at the Datong Railway Museum, see https://www.steamlocomotive.info/vlocomotive.cfm?Display=27924 or https://www.flickr.com/photos/lhoon/191839921 .

But in general, when things (not just locomotives) are numbered, it is customary to start at 1.

And traditionally that is what railways did - in general having multiple series for different types of vehicle: the concept of a mixed train being formed of loco no. 1 hauling 1st class carriage no. 1, 3rd class carriage no. 1, van no.1, open wagon no.1 and brake van no.1 may be apocryphal, but I believe there were railways with the appropriate series of numberings... Even if such complexities were reduced, separate series for locos, passenger and freight vehicles remained normal.

When there came to be a desire - at different times on different railways - to have locomotives of the same class within a block (often many such!) of numbers for ease of identification, the numbers were still within the one sequence of loco numbers, and it was logical to start 'blocks' at ~0 or ~00, and that is what the GWR (extensively), LMSR (to a lesser extent) and later BR (for standards and diesels) - sometimes, and to varying extents, using the number of the class leader to identify the class - and, at least to some extent, for the BR standards, the first two digits as a class or type indicator.

But abroad, when the Austrian kkStB decided in 1886 to have a system of class numbers, with locos of each class then starting from ~01, to give three or four digit loco numbers starting with 101 (first loco of class 1) they soon found an issue when classes came to have more then 99 members. So in 1904 they changed the system to put a '.' between the class and individual loco numbers - to now start with 1.01, so 200 could be renumbered to be with its class mates as 1.100, and 11.01 could be distinguished from 1.101 [note - random numbers to demonstrate system]. But this did consider the two parts of the number as being separate sequences, i.e. class number + individual number within class.

Trivia question: was the Austrian kkStB the first railway to adopt this now widely used system?

Eventually, when in the early 1970s BR did decide on a renumbering, based on the class numbers, to give what we call 'TOPS Numbers', they did work on the basis that the second part (individual loco id) would be a separate number sequence, starting from 001...

I suppose a good example of the difference is the extention on LNWR numbering through to the Patriots. The LNWR numbered their locomotive assets upwards from 1. If an engine wore out and the cost of it was written off, it'd be replaced by a new locomotive bearing the same number. But not to a superceded design, something current...
On some railways, a superannuated engine fit for use would be retained in a duplicate series, sometimes with a zero prefix, by adding a round number of thousands to it, or an A suffix. Hence some specific LSWR retainees having 0xxx numbers, and the LMS adding 20000 to lingering oddments to free up ranges for new engines.

The same practice happened later with the Patriots which took the number of the Claughton they nominally replaced, presumably to satisfy the accountants- new is from capital, rebuild is from revenue, and there's a difference to the tax man.

But, as the area of one-engine-one-crew passed it was more useful to be able to tell what kind of engine something was by it's number range, hence the renumbering of Patriots into a single run for operational convenience shortly after.

But those numbers were still just a serial number on an asset register and to try and treat them as a classification in the same way as the diesel scheme is folly.

TheHovisKid said:

But, as the area of one-engine-one-crew passed it was more useful to be able to tell what kind of engine something was by it's number range, hence the renumbering of Patriots into a single run for operational convenience shortly after.

Click to expand...

It is easier for people to remember blocks of numbers than individual seemingly random numbers. More recent examples include BR swapping numbers between Class 76s to keep the air brake fitted locos in blocks and the four Westerns that didn’t get air brakes being D1017-20. In the latter case, it was a deliberate decision by the shopping controller at Paddington to nominate four consecutive numbers for easier operational identification.

TheHovisKid said:

I suppose a good example of the difference is the extention on LNWR numbering through to the Patriots. The LNWR numbered their locomotive assets upwards from 1. If an engine wore out and the cost of it was written off, it'd be replaced by a new locomotive bearing the same number. But not to a superceded design, something current...
On some railways, a superannuated engine fit for use would be retained in a duplicate series, sometimes with a zero prefix, by adding a round number of thousands to it, or an A suffix. Hence some specific LSWR retainees having 0xxx numbers, and the LMS adding 20000 to lingering oddments to free up ranges for new engines.

The same practice happened later with the Patriots which took the number of the Claughton they nominally replaced, presumably to satisfy the accountants- new is from capital, rebuild is from revenue, and there's a difference to the tax man.

But, as the area of one-engine-one-crew passed it was more useful to be able to tell what kind of engine something was by it's number range, hence the renumbering of Patriots into a single run for operational convenience shortly after.

Click to expand...

The LNWR - somewhat infamous for its higgledy piggledy numbering system, although it's probably fair to say its was no worse than many other railways', it was just larger - did have a duplicate list, formed by adding 3000 to the former number.

After the grouping, and it became part of the LMS, the new regime undertook a renumbering, on the same principles as the Midland's 1907 renumbering, to produce some sort of order, with LNWR (LMS Western division) locos being renumbered between 5000 and 9999; parallel schemes applied to the Central (ex L&Y - from 10000) and Scottish (from 14000) divisions. But all of these failed to leave large enough gaps for the future... While the Claughtons were able to retain their neat group (5900 to 6029) of numbers, and those 'converted' to Patriots kept theirs, once it was decided to recognise the reality of the Patriots being new locos in a new class, and the large number of new locos coming under Stanier, there had to be some mass renumbering (by adding 20000) to provide enough space.

TheHovisKid said:

But those numbers were still just a serial number on an asset register and to try and treat them as a classification in the same way as the diesel scheme is folly.

Click to expand...

Yes: the problem with loco - and other rail vehicle - numbers is that are both asset ids (which can be essentially random), and 'working' ids for operators and engineers, who like to have some idea of what they imply (and who are horrified by my example of a six vehicle train, each with the same number, above). (see the discussion last November regarding EVNs and URVIS numbers). There have been periodic attempts over the last two centuries to keep everybody happy...

Clarence Yard said:

It was to put all vehicles in a common number range with locos not being altered. So passenger stock had to fit around the locos.

The pre-nationalisation coaching stock existing in around 1983 gives you the guide as to what numbers were left blank for the subsequent issue of numbers. Reuse was done under common rules for all types of stock.

Click to expand...

That’s not the issue though - it’s why in some cases numbers were reused almost immediately whereas in others they apparently couldn’t be until 7 years had elapsed.

etr221 said:

And traditionally that is what railways did - in general having multiple series for different types of vehicle: the concept of a mixed train being formed of loco no. 1 hauling 1st class carriage no. 1, 3rd class carriage no. 1, van no.1, open wagon no.1 and brake van no.1 may be apocryphal, but I believe there were railways with the appropriate series of numberings...

Click to expand...

The metropolitan railaway was one such, with electric locomotives, Driving Motor cars, 1st class Driving trailers, 3rd class trailers, and composite trailers numbered in five separate series all starting at No 1 in 1904, together with two more separate series already existing for steam locomotives and hauled stock. Yet another series was added by the takeover of the GN&CR in 1913, but those cars were renumbered into the appropriate main series (plural) in 1923.

Met EMU vehicles were renumbered by London Transport into the 2xxx, 6xxx and 9xxx ranges, GN&C stock being in the x9xx ranges.

The exile said:

That’s not the issue though - it’s why in some cases numbers were reused almost immediately whereas in others they apparently couldn’t be until 7 years had elapsed.

Click to expand...

I can’t now remember when the 7 year “rule” came in but they would have started by reference to a database as at a certain date. I seem to recall that renumbering to free up a number was easier than re-allocating a number of a previously scrapped vehicle, presumably because of data retention issues.