The engines that could have been, and those that never were
PHIL ATKINS explores the tantalising subject of engines that nearly were: designs that failed to get beyond the drawing board, and new members of existing classes that were never built.
Nearly 23,000 steam locomotives entered service on Britain’s railways between 1901 and 1960 ‑ but hundreds more were also authorised at one time or another, and so temporarily ‘existed’ on paper at least, only to be cancelled for a variety of reasons. Some of these were repeat orders for existing classes, allocated higher running numbers that would never be reached. In other cases, entirely new designs were involved, unfortunately destined never to see the light of day, although the active construction of at least one of these actually commenced. During the 1920s, and on two different railways, the unusual situation arose of scheduled large passenger tank locomotives being changed during construction to become tender engines instead.
Chop and change
Britain’s larger railways predominantly built their own locomotives, and before 1923 locomotive cancellations were uncommon. Nevertheless, late in 1903, at the end of his long and distinguished 31‑year tenure at Derby, S.W. Johnson of the Midland Railway initiated an outside‑cylinder 0‑8‑0 heavy mineral engine, and a 4‑4‑4 passenger tank engine. Ten of each were authorised for construction in Derby Works, but both were cancelled almost immediately in early 1904 by his successor, Richard Deeley, who himself also put forward 0‑8‑0 proposals without result. (These were followed by several Fowler 2‑8‑0 schemes, of which one at least resulted in the Somerset & Dorset 2‑8‑0s). The proposed 4‑4‑4T eventually appeared, re‑vamped as the dimensionally similar but distinctly disappointing ‘Flatiron’ 0‑6‑4T, of which 40 were built at Derby in 1907. Later that same year, the North Eastern Railway ordered ten more of the two extremely elegant four‑cylinder compound 4‑4‑2s, Nos. 730/1, that it had built the previous year. Although seemingly never cancelled, for some reason these engines simply did not materialise, reputedly because of the problematic question of patent royalty payments demanded in connection with certain unusual design features. It is possible that the very impressive but distinctly unsuccessful NER Class R1 (LNER D21) 4‑4‑0s were built as substitutes for these. At this time on the Great Western Railway, G.J. Churchward’s locomotive standardisation programme was getting into its stride, particularly personified by his taper boiler 4‑6‑0s, the two‑cylinder ‘Saints’ and four‑ cylinder ‘Stars’. Direct developments of the latter as the ‘Castles’ were built until as late as 1950, but the last ‘Saint’, No. 2955
Totworth Court, emerged from Swindon Works as early as April 1913, although five more (GWR Nos. 2956-60) had originally been authorised. Possibly on account of reduced railway revenue resulting from the recent miner’s strike, these had been cancelled in November 1912. Already responsible for four distinct classes of four-cylinder 4-6-0s - which between them had large and small diameter boilers, and three different coupled wheel sizes - during 1912 on the London & South Western Railway, Dugald Drummond initiated a fifth such design with the intermediate 6ft coupled wheels and reverting to the original large boiler diameter, together with a corresponding 0-8-0 heavy mineral engine. What need there was for a heavy freight locomotive on the LSWR, which had received no 0-6-0s since 1897, is unclear - unless it was to move Admiralty coal to Portsmouth and Devonport in time of war. The designs were respectively designated ‘K15’ and ‘H15’, but on the LSWR alpha-numerical locomotive classifications were, in reality, works order numbers. It is therefore likely that five engines of each type had already been approved before Drummond’s sudden death in November 1912, after which design work for them abruptly ceased. His successor, Robert Urie, quickly initiated his own ‘H15’ 4-6-0 with 6ft coupled wheels, and the ten engines thus built in 1914 at Eastleigh Works almost certainly made use of the large diameter boiler shells already fabricated for the projected Drummond engines. These massive Urie twocylinder engines would last until the 1950s. Between 1911 and 1917, the London & North Western Railway at Crewe Works built 30 non-superheated 0-8-2 heavy shunting tank engines, based on its most recent 0-8-0s, and ordered 40 more during 1920-21. These could not be progressed immediately, and were later cancelled and replaced by 30 ponderous, superheated 0-8-4 tanks, ordered in September 1922. These emerged in the very early days of the LMS during 1923, and some were even employed on passenger duties. In January 1922, the LNWR had amalgamated with the Lancashire & Yorkshire Railway, whose chief mechanical engineer, George Hughes, now headed the new combined organisation. Hughes soon envisaged a 4-6-4 tank version of his new superheated four-cylinder 4-6-0 express engines, and, in addition to 20 4-6-4Ts ordered from Horwich, another 40 were ordered from Crewe Works in March 1923. The first, No. 11110, emerged from Horwich just a year later, by which time the decision had been taken to build only the initial ten. Instead, 20 4-6-0s, LMS Nos. 10455-74, were then built at Horwich, using frames already profiled (surprisingly at St Rollox Works) for more 4-6-4Ts but cut back, while the remaining 30 tank engines were cancelled altogether. Oddly enough, most of these 4-6-0s lasted barely ten years, being outlived for the most part by the 4-6-4Ts, yet such was the limited work that could be found for such large tank engines that there had been an abortive proposal in 1931 to rebuild these as 4-6-0s as well.
Unlike the LNER, the LMS inherited from its antecedents neither a powerful, up-to-date, eight-coupled heavy mineral engine, nor a modern 4-6-2 express passenger locomotive. In the former category, the LMS directors swiftly authorised the construction of 100 2-8-0 locomotives in the ‘8F’ power category. No details about these have survived, and in the event a similar number of ‘4F’ 0-6-0s were built instead. Cancellation was probably on account of weight constraints, particularly regarding the Midland Division, where they were most urgently required. For this, a 2-8-2 in various forms was also considered, latterly in four-cylinder compound form, in parallel with a corresponding passenger 4-6-2 for the West Coast Main Line. It would appear that five 4-6-2s and five 2-8-2s were actually ordered from Crewe Works in March 1926. A number of drawings for these survive at the NRM, and the late E.S. Cox recalled that “a foundation ring and some flanging blocks were made and, although I never actually saw them, there seems little doubt that some cylinders were actually cast as well”. More recently, documentary evidence has come to light that suggests two boilers were actually built, which were put to stationary use after the 4-6-2 project was abruptly abandoned in late 1926. This followed tests with a GWR ‘Castle’ 4-6-0 on the LMS, which demonstrated that a large, well-proportioned 4-6-0 (i.e. the ‘Royal Scot’) would suffice instead. Unlike its three contemporaries, the LMS suffered no outright locomotive cancellations amid the national economic depression that followed, instead persisting with lavish new build programmes. There were, however, a few minor changes of plan following the accession of William Stanier as CME. For example, the last five scheduled Fowler ‘Patriot’ 4-6-0s, Nos. 5552-6, emerged instead as the first Stanier ‘Jubilees’, and the last five Fowler 2-6-4Ts, authorised as Nos. 2425-9, actually appeared instead as the first five Stanier three-cylinder 2-6-4Ts, Nos. 2500-4. Forty-five of the latter had originally been approved, but these actually terminated with No. 2536, while Nos. 2537-44 emerged instead in 1935 as the first Stanier two-cylinder 2-6-4Ts. A final 45 of these ordered in 1939 were not built on account of the war, but in effect were later re-ordered in 1944 in modified, steel-saving form to emerge as the first Fairburn 2-6-4Ts during 1945-46. The necessarily parsimonious LNER had to endure a number of cancellations in the interests of economy. These began with a single ‘N7’ 0-6-2T, which was deleted from a total of 33 such engines ordered from Doncaster Works in 1926 and built during 1927-28, simply to save about £4,000 (nearly £250,000 today). This was nothing compared to the 20 ‘V1’ 2-6-2Ts cancelled in 1931, followed by five (unidentified) ‘A3’ 4-6-2s and seven ‘O2’ 2-8-0s (Nos. 2962-8) in 1932. The ‘A3s’ had been ordered in April 1932, probably to have been numbered 2500-4, but were called off a year later - yet a final nine examples, Nos. 2500-8, were then ordered only eight months subsequently in November 1933. Ten more ‘V1s’ were cancelled in 1936, and overall no fewer than 55 Gresley 2-6-2Ts were ordered only to be cancelled, which also included 25 of the later higher pressure ‘V3s’ during wartime in 1941-43. In addition, a total of 53 Gresley GNRtype ‘J50’ 0-6-0Ts were ordered from Gorton Works between 1930 and 1939, which were all later cancelled. After the war, this shortfall in shunting tanks was solved by the purchase
Another 20 Maunsell Class ‘Qs’ were ordered in July 1937, but placed in abeyance by Maunsell’s successor, Oliver Bulleid
A total of 53 Gresley GNR-type ‘J50’ 0-6-0Ts were ordered between 1930 and 1939. All were cancelled
of 75 ex-Ministry of Supply ‘Austerity’ 0-6-0STs, which became Class J94. Of greater interest was the proposed LNER three-cylinder 2-8-2T, of which 12 were originally ordered in 1929 for the 1930 building programme. They would have been numbered 2875-86, but were cancelled. Nevertheless, detailed design work proceeded and ten were re-ordered in April 1932, only to be cancelled for the final time two months later. This Gresley ‘P10’ would have appeared before the only British 2-8-2Ts actually built - the GWR ‘72XX’ class, of which the first 40 were created in 1934 by rebuilding new 2-8-0Ts Nos. 5275-94. Ordered in August 1929, a few weeks before the Wall Street Crash, these were completed in 1930, but never entered traffic as a result of the ensuing depression in the South Wales coal industry. (A further ten 2-8-0Ts on order, Nos. 5295-9, and 6200-4, were never built.) Following the outbreak of war in 1939, another 20 2-8-0Ts, Nos. 5255-74, were ordered, of which only the first ten were built. The conflict also prompted the ordering of 60 Collett ‘2884’ 2-8-0s for overseas service, but this was later cancelled, and LMS Stanier ‘8F’ 2-8-0s, and subsequently MoS ‘Austerity’ 2-8-0s, were chosen instead.
On the Southern Railway, in addition to 20 Maunsell Class ‘K’ 2-6-4Ts (the infamous ‘Rivers’ Nos. A790-A809, built between 1917 and 1926) a further 20 engines, Nos. A610-A629, were ordered in June 1926. However, there were already misgivings about their limited usefulness, and a scheme for a 2-6-0 tender version, likewise having 6ft coupled wheels, had already been prepared in 1925. Then, the following year, on August 24 1927, No. A800 River Cray derailed at Sevenoaks, resulting in 13 fatalities. This put the reputation of the class, dubbed ‘Rolling Rivers’ by the press, under a cloud. As a result, the existing 2-6-4Ts were all rebuilt as nameless 2-6-0 tender engines (Class ‘U’), and the further 20 on order were turned out new in this form during 1928-29, to be followed by a final ten new 2-6-0s in 1931. On its formation in January 1923, the Southern Railway had inherited a recent order by the LSWR for four more Urie ‘G16’ 4-8-0 heavy shunting tanks, but with larger bunkers and the superheaters omitted, for service at Feltham yard. This was finally abandoned in 1926 following the decision to build the lighter and quieter three-cylinder Class ‘Z’ 0-8-0T, of which eight (SR Nos. B950-7) were later built at Brighton in 1929. Although a further ten ‘Zs’ were ordered from Eastleigh Works in April 1930, these were axed on economy grounds just a year later. By 1936, even the newest 0-6-0s on the Southern were approaching 30 years old, and the 20 Maunsell Class ‘Qs’, Nos. 530-549, were ordered as routine replacements, eventually being built at Eastleigh during 1938-39. It is not generally known that another 20 were also ordered in July 1937, but placed in abeyance a year later by Maunsell’s successor, Oliver Bulleid. However, these were not finally cancelled until July 1941, just as Bulleid ordered 40 of his own unmistakable ‘Q1s’, Nos. C1-40, which were all rapidly built during the following year. A further 20 ‘Q1s’ were ordered in June 1943, but cancelled in 1948. Thus the Southern Region in the 1950s would not be graced by 0-6-0s numbered 30550-69 or 33041-60, not to mention some 0-8-0Ts sporting Nos. 30958-67. Further north there was no sign either of 20 Peppercorn ‘A2’ class ‘Pacifics’ numbered 60540-59, which had been authorised in the place of 28 Thompson ‘A2/3s’, but which themselves had been cancelled in May 1948. (At one point there had been negotiations concerning the possibility of some ‘A2s’ being built by Beyer, Peacock & Co. in Manchester). Similarly absent from the 1950s East Coast Main Line scene were Gresley ‘V2’ 2-6-2s Nos. 60984-7, which had been completed at Darlington Works in 1944 as the unsightly Thompson ‘A2/1’ 4-6-2s, by now BR Nos. 60507-10.
Finally, turning to the BR Standards, of which there were a total of 999, it is well known that others were also ordered but never built. Most notable were a further 15 ‘Clan’ 4‑6‑2s on the 1954 building programme, Nos. 72010‑24, whose prospective names were even officially announced. The first five, bearing the names of Saxon warlords, were destined for the Southern Region ‑ despite the fact that it was already well‑stocked with such ‘Light Pacifics’ ‑ while the next ten were for the class’ spiritual home of Scotland. In March 1955, under a sub‑heading of ‘Crewe Works’, The Railway Observer reported that: ‘it is estimated that the first of the new order of ‘Clans’ will be ready about May’, but then went completely silent on the matter. Also on the 1954 programme were ten Class 4 4‑6‑0s (Nos. 75080‑9), five Class 3 2‑6‑0s (Nos. 77020‑4), and 18 Class 3 2‑6‑2Ts (Nos. 82045‑62). All 48 locomotives were formally cancelled in August 1956 ‑ but this is not quite the end of the story, for there could have been even more Standards than these. In E.S. Cox’s authoritative study of these engines, British Railways Standard Steam Locomotives (Ian Allan, 1966), he commented on No. 71000 Duke of Gloucester (page 111): “the possibility of building more of these engines was raised in 1956,” when steam’s future was still unclear. He may actually have meant ‘on the 1956 programme’, which had initially been debated in early 1955, notwithstanding the announcement that January of the very hastily drafted British Railways Modernisation Plan. This had proclaimed the eventual total replacement of steam traction, but at that time it was not foreseen that this might happen in little more than just ten years. Initially it had been intended to thoroughly test a relatively small number of pilot diesels over a few years before placing bulk orders for diesels c1960. In the meantime, it would still have been necessary to routinely build new steam locomotives simply to maintain the numerical strength of the operating locomotive stock. Indeed, a further 36 ‘Britannia’ 4‑6‑2s (for the London Midland, Eastern and Western Regions) were, provisionally at first, put on the 1956 programme, together with a further 20 ‘Clans’ specifically intended for the North Eastern Region. It would be interesting to speculate as to which duties and routes they might have been deployed. There were also to have been additional Class 4 and Class 5 4‑6‑0s, Class 3 2‑6‑0s and 2‑6‑2Ts, Class 2 2‑6‑2Ts, and 30 more Class 4 2‑6‑4Ts, 20 of which were earmarked for the Western Region. Respective BR ‘top’ running numbers would therefore have been Nos. 70090, 72044, 75119, 73175, 77037, 82072, 84039 and 80185. Also listed at this stage were five Class 8 2‑8‑0s to a new design for an indifferent Western Region, which had not initially appreciated the new ‘9F’ 2‑10‑0s; admittedly, there had been somewhat alarming teething troubles with steam brakes and jamming regulators concerning the initial batch on the WR during 1954. The proposed Standard ‘8F’ 2‑8‑0 had alternatively been outlined in simple diagram form in November 1953, both at Derby, as a truncated ‘9F’ with wide firebox, and at BR HQ as simply being derived from the Standard Class 5 4‑6‑0. In the event, neither scheme was developed, and additional ‘9Fs’ were built instead, which the WR had grown to love by the late 1950s. But how quickly policy changed ‑ by late 1956 in fact ‑ as needless to say, none of these ‘extra’ Standards actually received final authority. Even the last ten 2‑6‑4Ts built at Brighton Works, concluding with No. 80154, only made it past the assembly stage because the parts had already been made. The last steam locomotives to be approved by BR were ‘9Fs’ Nos. 92203‑50 for the Western Region, but only ‘with reluctance’ as there were no suitable diesels available in 1956 to replace life‑expired Churchward ‘28XX’ 2‑8‑0s, the oldest of which were already past their theoretical 45‑year working life. (The WR had sought to build more of these as recently as 1953, a full 50 years after the appearance of the prototype, prompting the BR Standard 2‑8‑0 counter‑ proposal). It may or may not have been a coincidence that this total of 48 ‘9Fs’ corresponded precisely with that of the other BR Standard engines simultaneously cancelled from the 1954 programme, thereby resulting in no net increase. Yet 60 years later, the number of BR Standards built will indeed increase past the 1,000 mark, with new‑builds of No. 72010 Hengist and No. 82045 under construction, ‘2MT’ No. 84030 being created from 2‑6‑0 No. 78059, and even a proposal for a ‘3MT’ 2‑6‑0, No. 77021. With the burgeoning popularity of new‑build steam, could any more of the ‘cancelled’ locomotives that we have discussed ‑ or even the ‘might‑have‑beens’ such as the BR Standard 2‑8‑0 ‑ one day become a reality after all?
If all had gone according to plan, the ‘Q1’ 0‑6‑0s could have been numbered up to 33060 in the BR sequence. The doyen of the class, preserved as part of the National Collection, heads a three‑coach Bulleid set past Holywell waterworks at the Bluebell Railway, on February 11 2000.
It looks like a ‘cut and shut’ locomotive because that’s practically what it is… No. 60508 Duke of Rothesay was one of four Thompson ‘A2/1’ 4‑6‑2s originally ordered as ‘V2’ 2‑6‑2s, and built using materials from the cancelled Gresley machines. Note the V‑fronted cab and banjo dome. Chunky Urie ‘H15’ 4‑6‑0 No. 30482 ‑ the ultimate result of Dugald Drummond’s attempts to develop 4‑6‑0s for the LSWR ‑ at Waterloo on March 5 1955. Maunsell ‘U’ 2‑6‑0 No. 1618 was to have been a ‘K’ 2‑6‑4T, until the fatal Sevenoaks derailment. It is preserved on the Bluebell Railway, where it was photographed at New Coombe Bridge on June 9 1992.
Youngest-surviving BR ‘4MT’ 2-6-4T No. 80151 is doubly fortunate not only to have reached preservation, but to have been built in the first place. Towards the end of its last stint at the Bluebell Railway, it brings up the rear of a train leaving Horsted Keynes behind ‘P’ 0-6-0T No. 178 on May 16 2012.
Diagrams of proposed BR Standard 2-8-0s for the Western Region, 1953.
The LMS number 5554 was originally allocated to one of a final batch of ‘Patriots’, but ended up on ‘Jubilee’ Ontario, which on August 22 1964 was preparing to head north from Carlisle. One of the 999 BR Standards that reached the rails, ‘4MT’ 2‑6‑0 No. 76085, shunts on the right. The would‑be prototype of a fleet of 60 4‑6‑4Ts, LMS No. 11110, near the end of its short life, at Manchester Victoria in April 1939. The ten‑strong class was withdrawn from service between 1938 and 1942, after a maximum service life of only 18 years.