How well represented the different railway companies are in preservation

I'm starting a new threat as I didn't want to divert the other thread on numbers of locos at the grouping.
Using those numbers I've done a basic assessment of how well represented the various pre-grouping companies are in preservation. Taking all locos ever owned by the company now preserved, and dividing by the number of locos at the grouping (to give a relative size of loco fleet for comparison) I've then ranked the "least well represented":

What was surprising to me personally here is the inclusion of the Midland! I never realised how badly represented that line is. Maybe those 2P folks are onto something after all (although a nie Johnson 4-4-0 would be much nicer and could represent the M&GN too with a livery change and some artistic licence).

I then excluded "older locos" in a slightly arbitrary manner (e.g. really old locos like Lion, Columbine, etc. up to about 1870ish) and the results were even more stark:


Interesting how the Welsh and Scottish companies come out badly, and most LMS companies, while the
Southern and most LNER do better.
The LNWR, Midland and GCR stand out at the large companies which are particularly badly representeded (if we ignore the very old LNWR locos especially).

The precise figures are debatable (both number preserved and number at grouping) but are intended only for relative comparison, and a few locos either way doesn't change the result in terms of the worst represented.

Perhaps this should give some direction to new-build projects....? The Barry and Rhymney had some nice and eminently useful tank locos, the Cambrian had some elegant 4-4-0s and 0-6-0s, the M&GN those 4-4-2Ts, GCR lovely 4-4-2Ts and 4-6-2Ts, the GSWR's 4-4-0s were nice, a Furness 4-6-4T might be fun, and the LNWR already has the George the Fifth projects (and the Bloomer). But based on my analysis, my ideas of usefulness, and personal preference I would say top priority is a Midland Johnson 0-4-4T! One was laid aside for preservation, but sadly later scrapped. It could also represent the Midland, M&GN and S&D with a livery change, I believe?

 

A similar comparison for the Big Four gives the unsurprising following comparison:

Of course this is skewed by the large number of Bulleid pacifics, for example, and of course the geographical location of a certain Welsh scrapyard has much to do with this...
But does suggest that the relative focus on LNER new-builds has some justification (A1, P2, V4 perhaps, B17 just possibly), and also that on pre-Stanier LMS types (Patriot, Fowler 4P tank). I will not say much about GWR new-build projects - my own preference is for Crewe and Horwich and might colour any opinion expressed on the various, no doubt entirely distict and disssimilar, GWR 4-6-0s being recreated...

 

13 locos for the LNER? Is that it? Are you sure? 6 A4s, 1 A3, 1 A2, 2 B1s, 1 B12, 1 N2, 1 N7, 1 E4, 2 Atlantics, 1 O4...

Is that just locomotives built by the LNER or total LNER preserved? Which of course which change the numbers quite a bit...

 

Platform 5's "Preserved Locomotives of BR" lists 8 MR locos, two of which were used on the Somerset & Dorset Joint line, and one of which was on the LT&S Railway, taken over by the MR in 1912. So I have some doubt as to the source of your figures. I appreciate that in terms of a percentage of the total fleet at Grouping a few more is a small change, but in terms of comparing what is preserved at present, it makes a significant difference.

In addition I am uncertain if the number of locos of a particular company is a good representation of how that company is represented in preservation - there is a lot more in preservation than just locos. Perhaps route-miles of preserved lines based on particular companies needs to be added in?

 

I'm a bit mystified by the tables. The first one includes Caley but not Furness, second one vice versa. And neither mentions the GN and NE. Did some rows get missed when you made the images to upload?

 

The thing is where to draw the line under as it the engines built in that grouping ie LNER engines built (completed) 01/01/1923 to 31/12/1947 or engines built to a new design which could include been completed on or after 01/01/1948 and not engines to a pre grouping design. In the first case the LNER A1 (A3) would be included and not the A2 which was completed after 01/01/1948 and the latter case the other way around as the A1 (A3) was to a GN design and the A2 to a LNER design. The NER J72 is an extreme case as it was built to a pre grouping design but was built after BR was formed.

In the first case it will be A1 (A3) 1 A4 6 B1 1 B12 1 D49 1 J27 1 K4 1 N7 1 V2 1 Y1 1 TOTAL 15

In the second caes it will be A2 1 A4 6 B1 2 D49 1 K1 1 K4 1 V2 1 Y1 1 TOTAL 14

.

 

I think there is a bigger error in the figures than that. The figure of 86 for the Southern includes all the preserved pre-grouping locos from the LBSCR, LSWR, SECR. (And, for info, it also includes the Bulleid Pacifics built by BR). So to be comparable, the LNER figure ought to include the equivalent pre-grouping locos, which brings in the GNR Stirling Single and J52; various NER locos including the Tennant 2-4-0; "Aerolite" etc; a GER 2-4-0 and 0-6-0 etc - so the number is considerably larger than 13.

Tom

 

Another issue of course is how representative of the companies stock are the preserved locos?

Given - for example that there are no Granges or Clans - at the moment, their absence is rather mitigated by the other GWR 4-6-0's & BR Pacifics that are preserved.

 

Is the next thread going to be manufacturers? That would be interesting as some of the manufacturers do not have any railway steam locomotives preserved.

 

As I said in my original post, there are many ways to slice this orange, and I chose to list LT&S and S&D separately, so they didn't get included in the MR total. I should, of course, have checked whether the locos at grouping number did the same...!
But as I also said, either way, the Midland was a very big company, and surprisingly few of its locos are preserved. Thanks goodness we have a compound. It is such a shame that the Johnson 0-4-4T and Kirtley 0-6-0 laid aside by the LMS on the 1830s were subsequently scrapped (along with a Kirtley 2-4-0, but thankfully another one was saved later).

 

I've ordered each table from worst to best represented, and then shown only the "top ten" (bottom ten, I suppose). The Furness has 3 surviving locos, so doesn't appear in the first table, but as all are extremely early, it jumps up the rankings when these are excluded, meaning the Caley drops out of the top ten.

 

I was trying, for comparison only, to include only locos built by the LNER. I may well have made mistakes.
Of course this is arbitrary and possibly not a good way of checking representation as a company's fleet included the absorbed locos. For the LNER and Southern in particular they were always a major part.
I did do many more tables than the ones shown, including one for the Big Four including absorbed locos. Will post it when I am at the computer, i.e. not middle of the night... Also with corrections from posts here.

 

Woops, my mistake. I did tables with and without the pre-grouping absorbed locos and seem to have got my SR numbers twisted. Will sort and repost.
It is somewhat arbitrary whether or not to include locos to previous designs built later. E.g. S15s built by the Southern or Bulleid's built by BR. I have included them (or tried to) in the totals for the railway where the design originated as we would normally talk of them as 'Southern' designs.
The purpose is comparison only, and do long as like is compared with like it should be fair. So if Bulleid pacifics built by BR are Southern, I have included Castles and Manors built by BR, Ivatt's built by BR, A1 built by A1ST, etc.

 

If anyone wants to hazard better estimates for any of the numbers, I am happy to correct and repost.

 

If you are going to include the later Bulleid Pacifics as an SR design (but built by BR) then to be consistent surely you have to consider the preserved N class as an SECR design? The S15 is another case in point: the later 8xx series engines by convention are considered an SR Maunsell design, rather than as a continuation of the LSWR design. Yet while there clearly are variations between the LSWR and SR S15s, one wonders whether they are any more significant than between, say, a 1923-built Castle and one built in 1950? If a 1950 Castle is considered a Collett design, featuring the inevitable design tweaks you'd expect from a long-lived class (and that is the usual convention), then why isn't a 1936-built S15 similarly considered a Urie LSWR design?

You also have to think of the denominator in your fraction: 1837 locos, as the number absorbed by BR from the SR, contained a majority of pre- grouping locos. So at the very least if you are seeing who did well or badly in preservation, you really need to count the pre-group locos in your total; or else adjust the denominator to account for only those locos actually built between 1923 and 1947.

It's a non-trivial task even agreeing terms!

Tom

 

Indeed chaps, and I did try to hint at all these pitfalls out in my original disclaimer.
There are undoubtedly errors in my counting, which was quick and dirty. Happy to have these corrected.
As to what you compare, then any like-with-like comparison is valid. As to which is most significant, that is a matter of opinion.
There will always be some places where a line has to be drawn. E.g. the Southern S15s were an evolution from the LSWR S15s - are they a separate class or not? How about the similar evolution of GWR Star to Castle? With some locos it is night debatable when they were actually built, due to heavy rebuilds which probably left little of the original.
Another discussion could be which early locos to exclude. We would probably all not count Lion, but I have excluded the 1860s Furness no 20, as archaic and not representative of the pre-grouping company in its heyday, while I have included the similar age Terriers as I decided they were representative of the LBSCR. Others might have different choices.
However I doubt very much whether any fiddling with where the Maunsell N is counted or whether Furness no 20 is included will make a difference to the fundamental fact that we are blessed with lots of GWR and SECR locos, very few Midland or LNWR, only one GSWR, and no Cambrian.

 

I think you also have to consider the spread of locomotive types, not just absolute numbers. As an example, consider:

LBSCR - 13 preserved locos
LSWR - 12 preserved locos
LNWR - 9 preserved locos

On the surface, the LBSCR and LSWR figures look broadly equivalent - if anything, the LBSCR got the better of the deal, especially as they were a smaller railway.

But the LBSCR locos include 10 of one type, and twelve from one Locomotive Superintendent, so it is a somewhat unbalanced representation of LBSCR motive power, even if you just consider the "mature" period of the pre-grouping. Of 13 locos, you've got 11 0-6-0T; one mixed-traffic 0-6-2T and one early express passenger 0-4-2 tender engine. There is nothing preserved by DE Marsh; nothing by L Billinton; no mainline goods engine; no later express passenger engine etc.

By contrast, of the twelve LSWR locos, you have representatives from four CMEs (Beattie, Adams, Drummond; Urie) covering the whole mature period of the company; you have eight different designs of loco (0298; 0415; O2; T3; M7; T9; B4; S15); four generations of passenger tank engine (0298; 0415; O2; M7); two generations of front-line passenger loco (T3; T9); a shunting loco (the B4) and one representative of the state of the art as the company transitioned into the big 4 (the S15).

Coming to the LNWR, of the eight preserved locos, one is narrow gauge ("Wren"); one is exceptionally early (Liverpool and Manchester "Lion"); two are significantly early and much rebuilt ("Cornwall" and "Columbine"). In effect, there are only four that could be taken as representative of the main motive power on the railway in its later period: "Hardwicke"; an 0-4-0ST shunting engine; the G2 freight engine and the Coal Tank - and that from what was more or less the biggest of the pre-grouping companies.

One would always want more, but you could probably argue that the LSWR is, by the standards of most pre-grouping companies, exceptionally well represented; and for such a significant company, the LNWR is rather poorly represented.

Tom

 

Quite so. I was going to move on to that level of sophistication later, but total number is still a good starting point.
The LBSCR number is skewed by all those marvellous Terriors. Why is why I am delighted you are building a new Atlantic, and why I think it is a shame itbis not in LBSCR condition. And why I think that another LNSCBR tank would be a cmgood candidate for future project.
My data table has many columns, mostly still to be filled in. I will share when I get a chance.

 

Regarding different types, we have had a previous discussion on te. topic: http://www.national-preservation.com/posts/1408727/
The outcome of that was the following: