BR Numbering

[David Bigcheeseplant]

Probably  a simple answer but I wondered why B.R added 30000 to Southern locos, 40000 to LMS types and 60000 to LNER types why not 20000, 30000 and 40000?

 

David

[Gwiwer]

20000 was the series used for Southern Railway non-steam and some experimental locomotives.  Of which there were very few but the series was used.  Former CC1 and CC2 became 20001/2 and 20003 was numbered thus from new.  

[Wickham Green too]

The 10000 series was reserved for diesel locos and 20000 for electrics - and these were seen as the way forward ............ the question has to be why the WR were allowed to persist with their heavy numberplates which could have been melted down to pay for renumbering - without that anomaly a more logical scheme could have been implemented ( LMS loco retaining their numbers as the largest.

[Wheatley]

5 hours ago, Wickham Green too said:

the question has to be why the WR were allowed to persist with their heavy numberplates which could have been melted down to pay for renumbering - without that anomaly a more logical scheme could have been implemented ( LMS loco retaining their numbers as the largest.

Because they were the most labour-intensive to renumber as it required more than a signwriter and a couple of tins of paint, and were therefore the obvious fleet to leave alone. 

Edited April 30, 2020 by Wheatley

[JimC]

The flexible, highly idiosyncratic but logical when you understood it GWR system would have been a pain in the neck with a numeral prefixed.

[Phil Bullock]

5 hours ago, Wickham Green too said:

The 10000 series was reserved for diesel locos and 20000 for electrics - and these were seen as the way forward ............ the question has to be why the WR were allowed to persist with their heavy numberplates which could have been melted down to pay for renumbering - without that anomaly a more logical scheme could have been implemented ( LMS loco retaining their numbers as the largest.

 

So that meant 30000 for Southern and LMS starting at 40000 - but the LMS had already tried to address duplicate numbers on locos inherited from other Companies at the Grouping by adding 10000 so it made sense to use the 50000 series for them - although not consistently, as ex LNWR Claughtons were allocated the 459xx - 460xx series but were all withdrawn before new numbers were applied - likewise ex LNWR 0-8-0s which ended up in the high 48xxx - 49xxx series. Perhaps because their numbers did not duplicate other LMS built locos?

 

 

 

 

Edited April 30, 2020 by Phil Bullock

[JimC]

My understanding was that LMS numbers ran to in excess of 2,000, but whereas those below 2000 had 4,000 added making them 4xxx and 5xxx those above 2,000 were renumbered.

[The Johnster]

I think Jim is right about this; the 1xxxx series on the LMS which became the 5xxxx under BR was simply the allocation of numbers rather than specifying pregrouping types, though it is true that no LMS built loco was numbered in this series.  

 

Presumably BR could have applied the 1-9999 series to any big 4 company's locos except the LMS, but the GW was chosen, or was it simply picked out of a hat?  The surprise is that it was allowed to keep the numberplates, rather than having the 4 digit numbers painted on in the standard Gill Sans.  Incidentally, Caerphilly works did paint some numbers on 56xx and 57xx/8750s in the mid fifties, in yellow paint, but in the style of the brass number plates including a painted 'polished raised border' surround.  There are photos of these in E.R.Mountford's 'Caerphilly Works' if you can get a copy.

 

There was an underlying logic to the GW/WR system by BR days, based on the second digit, but this was never fully realised.  I believe it was Collett's idea, but the trend had started under Churchward,  0 was allocated to the Stars and their developmental successors, Castles and Kings, then the Hawksworth Counties, which were no more mixed traffic locos than the Merchant Navys.  The RODs were anomalous, taking numbers from withdrawn Dean Singles but before Collett was in the big seat. 1 was allocated to the Large Prairie family, 2 to the 8-coupled tanks, then the Collett Goods and Dukedogs, later to move to the 0 family, 3 to the Moguls, 4 to small tank engines and the Hawksworth taper boiler panniers. The 14xx 0-4-2Ts were originally 48xx and given 14xx series when 48xx was appropriated for oil fired 28xx.  5 was used on the small prairies, and appeared on the 15xx, but by 1958 when the last of the Dean Goods had gone was restricted to only those locos.  6 was originally allocated to the 0-6-2Ts, but later pressed into service for 8750s when the 7 series was exhausted.  The 47xx were an anomaly, as were the 16xx.  7 was originally allocated to the large panniers, and was exclusive to the 57xx/8750s and derivates after the last 2721s and 1854s went.  8 applied to the 28xx and 2884s, and was then adopted for both both 0-4-2s and Manors & Granges.  9 started with the Saints and when one of those was rebuilt into what became the Hall class stayed with them.  The 44xx were another anomaly in numbering terms.

 

The pregrouping absorbed and constituent locos did not fit the system, being simply sequentially numbered in a 1-1299 series, and the 1361/1366 were not really part of the sequence either.  The absorbeds and constituents were allocated blocks of numbers based on company, not class, and then numbered sequentially within classes.

 

The BR power classifications and Route Availability numbers were allocated to GW locos but the region kept painting the GW versions on them and the BR values were ignored.  This continued well into diesel days and blue livery, being replaced by TOPS in 1973.

 

So. 1-9999 ex GW, 10000-19999 diesel, 20000-29999 electric irrespective of system, 30000-39999 ex Southern, 40000-59999 ex LMS, 60000-69999 ex LNER, 70000-99999 BR standard and Austerity, subdivided into 70000-79999 for mixed traffic tender locos, 80000-89999 for tanks, and 90000-99999 for tender freight locos. 71000 is an unplanned anomaly as a one off 8P.  There were no 30000, 40000, 50000, or 60000 as these were sequences of the first digit being added to the previous numbering, and nobody had a number zero.  

 

Diesel number were re-allocated D1-D9999 in 1958, but pre- 1958 locos such as the Ivatt twins and Southern 1Co-Co1s kept their original numbers.  Only shunting engines were renumbered from the 10000-19999 series into D prefix numbers.  Prototypes not owned by BR were not part of this system, though Falcon became 1200 when BR bought her.  A similar E prefix block was introduced in 1958, starting with E2001, a rebuild of Gas Turbine 18000 for the 25kv Euston-Birmingham-Manchester-Liverpool electrification.  Again, locos previously in service with 20000-29999 series numbers kept them.  

 

Breaking down the pre-TOPs D prefix number shows the intent to have a structure.  D1-D1999 were to be Type 4s, D2000-D4999 shunting locos, D5000-D5999 Type 2s, D6000-D7999 Type 3s, D8000-D8999 Type 1s, and D9000-D9999 Type 5s.  It didn't quite work out like that, as Type 2 number blocks meant that D61xx and D63xx overlapped into the Type 3 block, and the D7000 series occupied by type 3s and type 2s.  The D95xx were Type 1s.

 

The E prefix numbers were E3000-E4999 for 25kv locos and E5000-E6999 for 750v 3rd rail locos.

[Phil Bullock]

Thanks both - that makes sense!

 

You have to wonder if the logic was documented as it did seem to get forgotten...

 

https://www.flickr.com/photos/27045884@N05/3256128268

 

[Wickham Green too]

7 hours ago, The Johnster said:

I ....... locos previously in service with 20000-29999 series numbers kept them  .......

.... as most didn't last long enough to receive their TOPS-allocated numbers - apart from some of the Woodhead Electrics.

[JimC]

15 hours ago, The Johnster said:

There was an underlying logic to the GW/WR system by BR days, based on the second digit, but this was never fully realised.  

The trouble was the GWR were averse to major renumbering, so there was the original sequential system 1 - 1300, Dean's block allocation by type, which was 1301 - 4400, and Churchward's 1912 scheme of class allocations by second digit, which managed to include a good percentage of the Dean groups too, especially as the particularly confused 4-4-0s were renumbered in 1912.
Something that gives an insight into conservatism of ideas is that the last new GWR classes, the 15s and 16s, as well as the renumbered 48s were given number ranges that matched the old Dean block system, with 14xx having previously icluded Metros and 517s, and 15xx and 16xx 0-6-0T.

[jim.snowdon]

An added complication, practiced widely pre-grouping, and which persisted on the LNER, was the practice of re-using numbers as they were vacated by scrapping 

 

Jim

[LMS2968]

I think that was widely practiced; the LMS certainly did it. Depending on the date of tyour list No. 8000, etc. might have been a Webb coal engine or an 8F!

[JimC]

GWR certainly. At the time of the grouping one number was issued to three different locomotives in the course of a year.

[Peter Kazmierczak]

Deltics were to be D1000-21, then D1500-21 before D9000-21decided upon. 

 

25kV electrics were to start with E1, before it was decided to incorporate the horsepower rang; hence E3001 etc

[jim.snowdon]

53 minutes ago, LMS2968 said:

I think that was widely practiced; the LMS certainly did it. Depending on the date of tyour list No. 8000, etc. might have been a Webb coal engine or an 8F!

Ah yes, but at least what the LMS did was to renumber the offending locomotives in order to clear a block of numbers for the 8Fs so that they were numbered en bloc from new. 

 

Jim

[kevinlms]

8 hours ago, jim.snowdon said:

An added complication, practiced widely pre-grouping, and which persisted on the LNER, was the practice of re-using numbers as they were vacated by scrapping 

 

Jim

The LNWR was a classic at that, randomly reusing numbers to such an extent, that few locomotives were numbered as a block.

[LMS2968]

The LNWR system, and that of other railways, was that new locos were ordered under one of two accounts: there was Capital, where the new engines were new in that they added to the number in stock and would be numbered in a block; and Revenue, where a new loco was built tp replace an existing one as its predecessor was withdrawn, the new loco taking the number - and sometimes name - of the withdrawn loco. The two might have no connection as to type of train working and the new class might replace engines of several different classes, hence the random numbering of LNWR locos within a class. It did give the LMS some headaches when it came to allocating new numbers within each class.

[Compound2632]

The BR renumbering provided regional consistency with the A-road numbering system...

[Compound2632]

1 hour ago, Lantavian said:

 

Only roughly.

 

LNER electrics were renumbered to begin with 2, but they weren't from Kent.

 

Indeed it falls down for any part of the Eastern and North Eastern Regions east of the Great North Road. It's also rather an Anlgocentric perception.

[Wickham Green too]

8 hours ago, kevinlms said:

The LNWR was a classic at that, randomly reusing numbers to such an extent, that few locomotives were numbered as a block.

Both constituents of the SECR were as bad - though on a smaller scale, of course ......... then, to add to the confusion, they added 459 ( the highest SER number ) to the Chatham locos, when the working union was formed, rather than, say, 460, 500 or a nice round 1000 ! 

[Peter Kazmierczak]

A bit like why did WR shed codes begin with 81A rather than 80A? The answer being that GWR shed districts began with 1 for London, 2 for Bristol and so on. So loco (or shed) numbering always seemed to accommodate the GWR's slightly different way of doing things.

[kevinlms]

7 hours ago, LMS2968 said:

The LNWR system, and that of other railways, was that new locos were ordered under one of two accounts: there was Capital, where the new engines were new in that they added to the number in stock and would be numbered in a block; and Revenue, where a new loco was built tp replace an existing one as its predecessor was withdrawn, the new loco taking the number - and sometimes name - of the withdrawn loco. The two might have no connection as to type of train working and the new class might replace engines of several different classes, hence the random numbering of LNWR locos within a class. It did give the LMS some headaches when it came to allocating new numbers within each class.

My understanding of the reuse of numbers, was so that the bean counters could go to the pile of numbers in the store and find the next number, say 2222, which would mean that the LNWR had 2221 locomotives. 

They would need to subtract any numbers from the selves from withdrawn locomotives.

 

An example of the numbers carried by a relatively small class (by LNWR standards), can be found here.

 

https://en.wikipedia.org/wiki/LNWR_Precursor_Tank_Class

 

Note 5 did carry a small block of numbers.

 

While the LMS generally went to a great deal of trouble to renumber locomotives in 1923 (usually based on construction date and power ratings). This often meant they were not numbered on build dates solely. 

However, if they were upgraded or downgraded (boiler types usually), they were NOT renumbered to match say their superheated sisters, but kept the same number. So you can't look at an LMS number to determine whether or not it's superheated for example.

 

Quite different to BR numbering since TOPS, where any significant variation to a loco, usually means a renumber to another batch. But I imagine that is due to computerisation. Although the LNER did a similar thing.

 

[LMS2968]

22 minutes ago, kevinlms said:

My understanding of the reuse of numbers, was so that the bean counters could go to the pile of numbers in the store and find the next number, say 2222, which would mean that the LNWR had 2221 locomotives. 

They would need to subtract any numbers from the selves from withdrawn locomotives.

 

An example of the numbers carried by a relatively small class (by LNWR standards), can be found here.

 

https://en.wikipedia.org/wiki/LNWR_Precursor_Tank_Class

 

Note 5 did carry a small block of numbers.

 

While the LMS generally went to a great deal of trouble to renumber locomotives in 1923 (usually based on construction date and power ratings). This often meant they were not numbered on build dates solely. 

However, if they were upgraded or downgraded (boiler types usually), they were NOT renumbered to match say their superheated sisters, but kept the same number. So you can't look at an LMS number to determine whether or not it's superheated for example.

 

Quite different to BR numbering since TOPS, where any significant variation to a loco, usually means a renumber to another batch. But I imagine that is due to computerisation. Although the LNER did a similar thing.

 

Unfortunately, there first part of that doesn't work as it would need an unbroken list of locos from, to use your example, 1 to 2221 to make the next one 2222, which didn't happen and the LNWR didn't have that many engines. But you got a wide variation of numbers within a particular class, from single to four digits. They just used the number of the last withdrawn loco, irrespective of its class, so there was no correlation between number, class, power rating or intended traffic.

 

The LMS did try to number the engines in sequence, first by class then by build date, but there were some that didn't appear in the correct chronological order. Given the complexity of the eight-coupled engines' rebuilds, this isn't surprising. You are right that an engine, once numbered, retained that number even if subsequently rebuilt to a different class, so once again the numbers didn't give an indication of class or year of rebuild, but it was a good guide to the number of wheels!

[The Stationmaster]

9 hours ago, Compound2632 said:

The BR renumbering provided regional consistency with the A-road numbering system...

I didn't think the A1 ran for London to Bristol.  But GWR engine A1, renumbered 100A1 definitely ran from Paddington to Bristol