LNER 1943/46 renumbering

[norbitonflyer]

I am looking at how this worked.

The original numbering was based on the pre-grouping railways, with post-1923 builds filling the gaps
NER 1-2542
GNR 3001-4770
GCR 5001-6494
GNoSR 6800-6915
GER 7001-8900
NBR 9001-10481
The original pre-grouping companies did not, in general, group similar locomotives together, but simply allocated the lowest number currently available at the time the lovcomotive was built.

The post-1946 numbers were based on type
1-999 (and 10000) large six-coupled passenger tender (mainly 4-6-2)
1000-1999 medium six-coupled tender (mainly 4-6-0)
2000-2999 four-coupled tender
3000-3999 eight-coupled tender
4000-5999 six-coupled freight tender (0-6-0)
6000-6999 electric
7000-7999 passenger tank
8000-8999 shunting (including petrol engines - later also used for diesel)
9000-9999 freight and mixed traffic tank

Within each group each class had its own number block, allocated generally in order of physical size

Note that one locomotive based at, say, Aberdeen could be waiting for a number to be released by a locomotive based in Stratford - or outshedded at some remote branch terminus. How was this co-ordinated? In other cases a locomotive might take on a number previously allocated to another member of the same class, which could lead to misunderstandings.

As in many (probably the majority) of cases a locomotive's new number in the new scheme had been allocated to a different locomotive in the old scheme, it was deemed necessary to organise the renumbering in stages - first renumbering those locomotives whose new numbers were vacant in the old scheme, and then those whose new numbers had been relaeaed in the previous stage and so on. An added complication was that, because inadequate space had initially been made for some new classes, some locomotives were renumbered twice - most famously "Flying Scotsman" (from 4472, first to 502, and then to 103 when the 5xx block was needed for the new A2 class)

Some of the chains could be quite long - more than a dozen steps - so how was it actually organised?

Was an edict sent out (every week? month?) to each depot with instructions on which locomotives should be renumbered? How long did this all take?

Were there any circular chains, and if so how were they handled?

Update - the longest chain I have found so far involved 23 renumberings
7211 - number vacant: previously a GER class F4
GER class F5 7110 to 7211
GCR class F2 5782 to 7110
NER class J27 836 to 5782
LNER class V2 4807 to 836
LNER class J39 2781 to 4807
GER class E4 7463 to 2781
NBR class C15 9164 to 7463
NBR class N15 9223 to 9164
NBR class N15 9225 to 9223
GCR class N4 5513 to 9225
GER class J16 8163 to 5513
NER class D23 No 23 to 8163
LNER class A4 4482 to No 23
NBR class J35 9195 to 4482
NBR class N15 9240 to 9195
GCR class N4 5636 to 9240
NER class J24 1948 to 5636
LNER class K2 2447 to 1948
NBR class D32 9886 to 2447
NER class H1 1527 to 9886
GER class B12 8527 to 1527
GER class J69 7350 to 8527
GNR class C12 4009A to 7350

No 4009A had been added to the GNR duplicate list in 1921 (as 1009A) to make way for a new K3 locomotive. It was one of two on the GNR duplicate list to make it through to renumbering - the other was J53 No 155A (LNER 3155A), trasferred to the duplicate list in 1906 to make way for one of the relatively short-lived L1 (LNER class R1) 0-8-2Ts (all withdrawn by 1934)

EDIT - longest chain corrected

 

[norbitonflyer]

I am looking at how this worked.

The original numbering was based on the pre-grouping railways, with post-1923 builds filling the gaps
NER 1-2542
GNR 3001-4770
GCR 5001-6494
GNoSR 6800-6915
GER 7001-8900
NBR 9001-10481
The original pre-grouping companies did not, in general, group similar locomotives together, but simply allocated the lowest number currently available at the time the lovcomotive was built.

The post-1946 numbers were based on type
1-999 (and 10000) large six-coupled passenger tender (mainly 4-6-2)
1000-1999 medium six-coupled tender (mainly 4-6-0)
2000-2999 four-coupled tender
3000-3999 eight-coupled tender
4000-5999 six-coupled freight tender (0-6-0)
6000-6999 electric
7000-7999 passenger tank
8000-8999 shunting (tank and diesel)
9000-9999 freight and mixed traffic tank

Within each group each class had its own number block, allocated generally in order of physical size

Note that one locomotive based at, say, Aberdeen could be waiting for a number to be released by a locomotive based in Stratford - or outshedded at some remote branch terminus. How was this co-ordinated? In other cases a locomotive might take on a number previously allocated to another member of the same class, which could lead to misunderstandings.

As in many (probably the majority) of cases a locomotive's new number in the new scheme had been allocated to a different locomotive in the old scheme, it was deemed necessary to organise the renumbering in stages - first renumbering those locomotives whose new numbers were vacant in the old scheme, and then those whose new numbers had been relaeaed in the previous stage and so on. An added complication was that, because inadequate space had initially been made for some new classes, some locomotives were renumbered twice - most famously "Flying Scotsman" (from 4472, first to 502, and then to 103 when the 5xx block was needed for the new A2 class)

Some of the chains could be quite long - more than a dozen steps - so how was it actually organised?

Was an edict sent out (every week? month?) to each depot with instructions on which locomotives should be renumbered? How long did this all take?

Were there any circular chains, and if so how were they handled?

Update - the longest chain I have found so far involved 23 renumberings
7211 - number vacant: previously a GER class F4
GER class F5 7110 to 7211
GCR class F2 5782 to 7110
NER class J27 836 to 5782
LNER class V2 4807 to 836
LNER class J39 2781 to 4807
GER class E4 7463 to 2781
NBR class C15 9164 to 7463
NBR class N15 9223 to 9164
NBR class N15 9225 to 9223
GCR class N4 5513 to 9225
GER class J16 8163 to 5513
NER class D23 No 23 to 8163
LNER class A4 4482 to No 23
NBR class J35 9195 to 4482
NBR class N15 9240 to 9195
GCR class N4 5636 to 9240
NER class J24 1948 to 5636
LNER class K2 2447 to 1948
NBR class D32 9886 to 2447
NER class H1 1527 to 9886
GER class B12 8527 to 1527
GER class J69 7350 to 8527
GNR class C12 4009A to 7350

No 4009A had been added to the GNR duplicate list in 1921 (as 1009A) to make way for a new K3 locomotive. It was one of two on the GNR duplicate list to make it through to renumbering - the other was J53 No 155A (LNER 3155A), trasferred to the duplicate list in 1906 to make way for one of the relatively short-lived L1 (LNER class R1) 0-8-2Ts (all withdrawn by 1934)

EDIT - longest chain corrected

I have now been through all the renumberings - it gets complicated because some were renumbered twice. My sources are Longworth (which only covers locos which survived to nationalisation a few years later) and Casserley & Johnston (which only covers locos inherited at the Grouping) so there may be some gaps, but they would have to be relatively short-lived : built in LNER days, and withdrawn between 1943 and 1947.
The string of 23 renumberings listed above is, I believe, the longest.
Three locomotives kept their original numbers - J4 No 4125, J3 No 4126 (both ex-GNR 0-6-0s) , and W1 4-6-4 No 10000

Sixteen pairs of locos did straight swaps of numbers: ex-GNR N1s 4574 to 4589 exchanging numbers with ex-NBR J37s 9454 to 9469 respectively, and there were just four larger loops, again involving the N1s and J37s (two of each):
4590-9470-4596-9476-4590
.......
4593-9473-4599-9479-4593

Loops would have been difficult to manage, as in general I understand each loco could not be renumbered until its new number had been relinquished by the original user. The relatively small number of loops looks to have been deliberate.

Some of the temporary renumberings look to have broken what would otherwise have been loops, although others were a change of plan (originally the A3 and A4 Pacifics were to have been numbered in build order, starting at 500, but the 500 block was then needed for Thompson's new Pacifics so the older ones were numbered starting at "1", but this time with the A4s first and, with a few exceptions, both classes renumbered in the same order as their original numbers.

It is interesting to note that the "B12" 4-6-0s revereted to their original Great Eastern railway numbers 1500-1570 under the 1946 scheme, having been 8500-8570 for most of their LNER careers.

 

[Gloster]

You have presumably considered the former Metropolitan H2, L2 and M2 classes.

 

[norbitonflyer]

You have presumably considered the former Metropolitan H2, L2 and M2 classes.

Most grateful for pointing these out. As the LNER only acquired them in 1938, and only four survived to nationalisation, most of them did slip through the gap between my two reference sources.

From Longworth I had those which had survived to nationalisation. The "LNER Register" identified a fifth, but it does not give the 1946 numbers for the H2s and L2s. Can anyone help?

M2 6154-57 The first three became 9075/6/7 (the "LNER register" 9075 as allocated but never carried. None ever carried their allocated BR numbers

L2 6158-63: Longworth has 6158/60 becoming 9070/71. The others did not survive to 1948 but did any of them get new numbers before withdrawal? The gap between the M2s and L2s is suggestive.

H2 6415-22: None survived into 1948. I have found photos of two of them carrying the numbers 7511/12, so at least two were allocated new numbers - unfortuntaley the picture captions do not identify the original LNER numbers.

Their original (1938) LNER numbers were all in the 6xxx range, which was reserved in the 1946 scheme for electric locomotives. None of the LNER's fourteen elcetric locomotives took numbers that had previously been used by ex-Met locos.

 

[Gloster]

Met. 94-97 became 6154–6157 and the first three became 9075-9077 in the 1946 scheme, although 9075 may not have carried its number. 103-110 became 6415-6422; 6415-6420 and 6422 became 7510-7516, although only 7511 and 7512 may have carried their numbers. 111-116 became 6158-6163; 6158, 6160, 6162 and 6163 became 9070-9073, but only the first two may have carried the number. All were in order.

Source: Locomotives Illustrated 65, May-June 1989, Ian Allan. (Compiler Peter Rowledge.)

 

[norbitonflyer]

Met. 94-97 became 6154–6157 and the first three became 9075-9077 in the 1946 scheme, although 9075 may not have carried its number. 103-110 became 6415-6422; 6415-6420 and 6422 became 7510-7516, although only 7511 and 7512 may have carried their numbers. 111-116 became 6158-6163; 6158, 6160, 6162 and 6163 became 9070-9073, but only the first two may have carried the number. All were in order.

Source: Locomotives Illustrated 65, May-June 1989, Ian Allan. (Compiler Peter Rowledge.)

Many thanks. I will update my spreadsheet.

Some statistics

Number of chains - 2582
Average length of chain 3.35 (i.e 2.35 renumberings)
Longest chain 24 (23 renumberings)

Loops - 4 (each with four numbers)

Straight swaps - 16

Number Unchanged - 3

Numbers not involved 1266 (although many of these relate to locomotives withdrawn before renumbering or built new with 1946 scheme numbers, such as the diesel shunters numbered from 8000)

Number of locomotives involved 6108 (Although not all actually received their new numbers)

Longworth gives the number of ex-LNER locos inherited by British Railways as 6548. The difference is largely accounted for by ex-War Department locos (classes O7 and J94), which came into LNER hands after WW2.

 

[oldman]

There is information in the RCTS Locomotives of the LNER (available free at https://archive.rcts.org.uk/locomotives-of-the-lner/), both a summary of the renumbering in Vol. 1, and sections in the chapters for each class. As well as the Metropolitan Railway, another oddity is the small number of Midland and Great Northern locomotives taken over in the 1930s - some (class J93) got new numbers.

 

[Clarence Yard]

And for details of the individual locomotives, the relevant volume of Yeadon will give you the dates for each renumbering.

It would appear that in 1946/7 the majority were done on shed on a designated day. I would imagine that someone centrally kept tabs on all this and, when the number became available and the time was right, the appropriate order to stop and renumber would have been sent down to the shed via the appropriate DMPS for that area.

The work was carried out between 13/1/46 and 18/1/47. Everything that existed in July 1943 or after seems to have been allocated a number.

Yeadon 21 gives full details for the Met tanks and of these only 7511/2, 9070/1 and 9076/7 were so renumbered. All others were withdrawn with their “1938” numbers.

 

[Gloster]

As far as I can see the only M&GN locos that were renumbered were of LNER classes J4 and J93. However, both were in course of withdrawal, so I can’t help as to which ones actually gained the new numbers. It should be noted that before renumbering they just carried the old M&GN number with a zero at the start: 083, 094, etc.

 

[norbitonflyer]

As far as I can see the only M&GN locos that were renumbered were of LNER classes J4 and J93. However, both were in course of withdrawal, so I can’t help as to which ones actually gained the new numbers. It should be noted that before renumbering they just carried the old M&GN number with a zero at the start: 083, 094, etc.

According to Longworth, M&GN J4 locos 83, 85, 87, 88 and 92 became LNER 4158, 4160, 4162, 4163, 4167 in the 1946 scheme, and J93s 16, 94 and 96 became LNER 8489, 8488 and 8484 respectively. (Note the reverse order - with the exception of the Gresley Pacifics, the 1946 scheme was particular about numbering locos in each class in order of age)

Longworth only records those M&GN locos which were still in service in 1948, and the gaps in the sequence suggests there were others that were renumbered but didn't make it to 1948. Most were withdrawn shortly after nationalisation, and only No 4160 actually carried a BR number (64160), surviving until the end of 1951.

 

[norbitonflyer]

It's taken a while, but I think i now have a definitive list. The main complication was that several locomotives were renumbered more than once, with most such revisions to the scheme being initiated before the previous one was complete, and my sources were not always clear whether a renumbereing was from origional to intermediate, intermadiate to final, or original to final (disregarding the intermediate one). the very short-lived numbers given to the ex-War department O6s and O7s in 1946/47 (reclassified by BR as LMR 8Fs and Austerities) added to my confusion.

1941 - 122 locos in the 8xxx range was given new numbers in the 7xxx range to make space for Thompson's new B1 class - only 39 locos were renumbered before the more comprehensive 1943 scheme was introduced

1943-1946 - the 1943 scheme was not implemented until 1946, but some locomotives were renumvbered as they went through works or where their old numbers were required for new-build O6s and diesel shunters. 22 locomotives were given temporary numbers as their new numbers were still in use by their original owners

The 1946 scheme involved a major revision to the numbers in the range 1-999, fifty of the 397 locos in that range had already been renumbered, so they had to be re-renumbered. To make way for the rapidly-expanding B1 class, the thirteeen members of Class B5, two of which had already been renumbered, were also allocated new numbers.

The O6 class 31xx were renumbered a second time (to 35xx) to make way for ex-WD Austerities (LNER class O7). Neither class would acquire BR numbers in the 63xxx range, being numbered by BR respectively in the 48xxx range (as LMS 8Fs) and the 90xxx range.

I was mistaken in my previous post - the order in which classes were numbered in each block was not by size but by age. Classes still enetering service usually came first, then a gap, then other classes in ascending order of age. It was thus planned that new classes would fill the gap and then take over numbers as the oldest classes went out of service. Of course, this never came to pass as the scheme was overtaken by nationalisation a few years later.

Thus 138 locos were renumbered twice

6,532 locomotives were involved - this includes some which were allocated numbers in 1943 but were withdrawn before the number could be applied

The total number of chains is 2,355 (this includes straight swaps, loops and the three locos that kept their original numbers. It follows that 2,355 numbers were used in the old scheme but were spare in the other, and likewise there were 2,355 spare numbers in the old scheme that were used in the new one

1,253 numbers (between 1 and 9999) were not used in either the old or new schemes

It follows that 4037 numbers were used in both schemes, and the average number of renumberings in a chain is 2.83

I have not found a longer chain than the one of 23 I mentioned in my previous. There is another chain of 20, two of 18 and one of 17.

There were a few numbers in the old scheme that were outside the range 1-10000: these included:
- former North British Railway class Y9 numbered between 10083 and 10102 (NBR 1083-1102): the LNER had planned to renumber these in the 99xx range but never got around to it. In the new scheme they became 8092-8112 (not in the same order)
- former Midland & Great Northern class J3 and J93, numbered on acquisition by the LNER as 081-092 and 016 - became 4156-4167 and 8482-8489
- former GNR duplicate list class J52 3155A and class C12 4009A, became 8804 and 7350 respectively
- former GER J92 Crane tanks "numbered" B, C and D, became 8667-8669