The Story of the GWR 28XX 2-8-0s

by Roger Molesworth

(From "2807 News" - June 1998)


The GWR was first in the UK with a 2-8-0 wheel arrangement, when George Jackson Churchward rolled out his prototype heavy freight engine, number 97, in June of 1903. This was one of several patterns for standard loco designs, most of which were a resounding success (however, there was one, an 0-4-4T number 115, which proved to be unsatisfactory and never led to a class of engines). No. 97 had the same boiler as no. 98, which was a 4-6-0 and was the prototype of the Saints. Boiler pressure on 97 was 200 lb/sq.in, non-superheated; cylinders were two 18" x 30", and no. 97 emerged with a 4,000 gallon tender. It was initially painted black with red lining, but it was not long before it fell into line with other GWR engines, and was re-painted Brunswick Green with orange lining!

'... just in case the 2-8-0 had difficulty in getting round tight curves!'

No. 97 proved to be very successful and in 1905 the design was put into production, continuing to be produced in batches until 1942. Although this was a 'standard' design in name, ask anyone who has worked on them (especially our own 2807) and they will tell you that 'there is no such thing as standard!'

Numbers 2801 - 2820 were produced in the first batch. The most obvious change from 97 was the height of the boiler. This 'standard' boiler was of two-part construction, having a conical section at the firebox end, leading to a parallel section through to the smokebox. On no. 97 the boiler centre-line was 7'51/2" above rail height, whereas 2801-20 had theirs pitched higher, at 8'2", daylight being clearly visible between the bottom of the boiler and the frames. These production engines also came with slightly smaller 3,500 gallon tenders, which later became 'standard'. However, some differences were not obvious to the naked eye, such as the increase in piston valve diameter from 81/2" to 10". At the same time, the boiler pressure was raised to 225 lb/sq.in, which also added a little weight!

If you could have compared 2807 as she was rolled out in 1906, with how she is now (well, how she will be when she is finally assembled!) you might notice the outside steam pipes were absent from the smokebox sides; the frames were square - no gentle curve down from the running boards over the cylinders to that in front of the smokebox; and the chimney and safety valve bonnet were taller then. If you were very observant, you might also notice that Churchward has some reservations about the long wheelbase, and hence the rod connecting the front pair of driving wheels had a swivel-joint, to allow a small amount of sideways movement - just in case the 2-8-0 had difficulty in getting round tight curves!

In 1907 the next batch was 2821-30, and true to 'standard', these had increased diameter cylinders of 183/8"x 30". Furthermore, the chimney and valve bonnet were made lower, and the chimney wider!

In 1908 no. 2808 was the first to have superheating added - oh, and it's cylinders were increased to 183/8"x30" at the same time!

1911 saw the construction of 2831-35, and these introduced the familiar curved drop ends to the running plate (in front of the cylinders).

For the batch 2836-45 of 1912, top-feed boiler water supply was introduced, which then continued with engines to 2855. In December 1912, the prototype no. 97 was re-numbered as 2800. At some point, the smokebox was extended forwards by about 22"! No. 2846 was modified in 1917 by applying castings onto the frames to improve the weight distribution. This practice was continued thereafter.

In 1918 the batch 2856-83 came out with stays at the front (either side of the smokebox, down to the front plate), giving added strength to the buffer beam. This batch was completed in 1919, and things went relatively quiet for a few years. Scottish trainspotters were in for a treat in 1921, when 2804 went to the North British Railway for trials.

From 1934, locos were progressively modified with outside steam pipes, and the vacuum brake 'swan-neck', which had always been prominently high on the front buffer beam, was lowered.

'Standard they may have been, successful they certainly were ...'

More engines began to roll off the production line again, in 1938. These were modified with larger, side-window cabs, retaining outside steam pipes, and short safety valve bonnets. On the left side, along the running board in front of the cab was added a long box intended for the stowage of fire irons. However, in practice firemen found it easier to swing them onto the tender than line them up and push them into this box! A shield was positioned behind the whistle, and ATC equipment was fitted (most visible being the battery box beneath the right hand side of the cab). These visibly modified engines were known as the 2884 class.

The number range was now full, so 1939 - 1942 saw the production of nos. 3800-3866. The GWR had high hopes of more being produced for the war effort, but these plans were thwarted when the Austerity 2-8-0s were ordered, derived from William Stanier's LMS 8F. Stanier was, of course, an ex-Swindon man and built his designs from a GWR background! But this spelled the end of the line for 28xx derivatives.

Well, there was a brief flirtation with oil, when in 1945 twelve 28xx and eight 2884 locos were converted to oil burners, and given numbers in the range 4800-11 and 4850-7. Actually, this caused a problem and the little 0-4-2Ts had to be renumbered from 48xx to l4xx to accommodate this experiment! It was, of course, the tender oil tank which was the give-away for these locos. By January 1950 all had reverted to burning coal. As it happens, they needn't have bothered, because oil became hard to come by, and the experiment was abandoned in September of 1948. But while they were in Swindon under conversion, these engines acquired sliding shutters to their cab windows.

The final excitement was participating in the loco exchanges of 1948, when 3803 was used to evaluate the efficiency of this design. It turned out to come second in every category, though no other engine came consistently first! It certainly did GWR proud, with the exception of the time it ran into a snow storm and lost traction. The GWR repeated some of the tests later, using Welsh Steam Coal (as opposed to Yorkshire hard coal which was standard for the trials) and these results put 3803 first in every category!

So, that was about it, really. Apart from the 2-8-0T engines that were a development of the 28xx, which is another story (but maybe not as long!). Standard they may have been, successful they certainly were - the first locos to demonstrate a pulling power of 100 loaded coal wagons - but when it comes to fitting the 2807 jigsaw puzzle back together again, it is proving quite an eye-opener on Swindon's standardisation!