There are at present around twenty projects for building new steam locomotives, replicas of designs which did not survive the elimination of steam on Britain's railways in 1968. The creation of replicas, previously thought impossible, has become practicable and relatively affordable due to new techniques and new materials: CADCAM, 3-D printing for patterns, and plasma and laser cutting, SG cast iron, etc.
The A1 Pacific Tornado was the groundbreaker, begun in 1994, it was finished in 2008 and has been used for main line steam specials since then.
The new build proposals are either for locomotives for main line use or for haulage on preserved railways. Most of the choices, however, will be criticised by someone. Every class of locomotive had its strengths and weaknesses and most of them underwent modifications over the course of their time in service with the aim of getting rid of the faults.
Although main line steam is gradually being squeezed off the main lines
as regular traffic continues to grow, on secondary routes it is likely to continue
for the foreseeable future. Trips behind steam locomotives are as popular as ever, and any locomotive entering
service now will have a useful career ahead of it. However, the locomotives have a particular task to do: to haul trains of about 350 tons at speeds of at least 75 mph and preferably 90 mph. In this light, some of the proposals, whilst interesting, are going to result in locomotives which cannot be employed economically, being either too small, or unsuitable for high speed running.
As the generation that remembers steam in service passes into history, there will be a declining interest in particular types familiar from childhood. In twenty years time, most people will be satisfied with a generic steam locomotive, which will not even have to be burning coal.
For my money - and I will stick my neck out here, the most practical generic design remains the GWR Castle class. It is not accidental that construction of these moderate-sized (they weigh only 80 tones and have an axle load of just under 20 tones) but very competent locomotives was spread over 27 years, that the LMS wanted fifty of its own and that eight survive. They seem to have run comfortably at 90 mph, presumably because of the four cylinder configuration. The 4-6-0 wheel arrangement makes them good on hills. With two sets of eccentric-driven inside Walchaerts valve gear and the de Glehn staggered cylinder layout, the Castles are robust and fairly easy to maintain, apart from the large number of oiling points, some not very accessible. Their principal weakness, which showed up in the 1948 trials, and again recent years, is that they do not perform well with poor quality coal. Modifications such as increased superheating and double chimneys were made to remedy those deficiencies, and so the locomotives continued in service until the end of steam.
If we were looking for a standard locomotive for main line steam haulage and seasonal jobs, it would be possible to do a lot worse than build a small batch of Castles with a few tweaks: light oil firing to get over the fussy taste for coal, roller bearings, perhaps, improved materials including the use of stainless steel for some components, better lubrication arrangements and a general review of the details of the design in the light of over 90 years of experience.
To judge from the amount of noise they make, there is probably some energy wasted in the exhaust. Recent experience with the restored King class 6023, restored to its original single chimney form, and a new 4-nozzle exhaust arrangement designed by J J G Koopmans, suggests that there is still room for improvement to the draughting. The last of the unrestored Castles, 7027, never received a double chimney and, with renovation now in progress, would be a good candidate for a similar modification to optimise its performance.
Museum lines are another matter. They have a very specific requirement - they operate trains up to about 250 tones at fairly slow speeds on hilly routes with runs of up to about twenty miles.
This is quite a demanding task and was typically done by locomotives of power class 4, such as the 2-6-4 tanks built by the LMS and British Railways, GWR 2-6-2 types such as the 5100 class, or 2-6-0 tender engines such as the LMS and BR 4MT type and the GWR 4300 class. Many of the surviving locomotives are now getting on for eighty years old, and with few examples of each type, spare parts often have to be specially made.
Museum lines are also finding it increasingly difficult to find sufficient volunteers as the original steam age contingent are well into their sixties and more. So the need is for locomotives which are simple, efficient, reliable and specifically designed for the job that they will be doing. Again, the need is probably for a generic type.
That sounds like a batch of the BR or Stanier or Fowler designed 2-6-4 tanks, or possibly a new design based on the best features of all three, or the GWR 5100 class, though possibly with an option for a tender engine version.
What looks more questionable are the replicas of some of the smaller Victorian types. Charming as they are, they might not have the power needed for the work. It is the same reason why engines such as the Bluebell's delightful Adams 4-4-2 tank is, unfortunately, unlikely to steam again.
One of the problems with using steam locomotives intermittently is the effect of thermal cycling. Boilers would last longer if they were kept hot. One way of keeping them hot would be to use the locomotives for undemanding ordinary work when they are not being used for the special trips. If the locomotives were running on light oil, this would get round the difficulties of having to handle coal. This would have the advantage of generating a bit of extra traffic on lightly loaded trains. Swiss experience is that steam locomotives optimised to present-day levels are, in practice, pretty much the equal of diesels.