The International Steam Pages

Modifications to improve the reliability and operability 
of steam locomotives on the Kirklees Light Railway.

As originally constructed the locomotive fleet of the Kirklees Light Railway, when pressed to maintain the level of service required to operate the railway, particularly after the opening of the extension to Shelley in 1997, revealed a range of mechanical shortcomings as well as acute limitations in the available drawbar output inherent in the original designs. In the light of the problems mentioned above, revealed by the intensive operation required of the locomotives, it was considered that a number of different aspects required to be addressed:

  • Firstly, the mechanical reliability of the locos needed to be significantly improved as no matter how efficient a locomotive may be, if it is not mechanically reliable it is of little use as an operating tool. Also a quite impressive amount of time was being wasted on so called ‘routine maintenance’.

  • Secondly, the locomotives had to be consistent performers as to steaming ability and quality of work output with a minimum level of driver/fireman input, this ultimately being of greater importance than achieving the absolute maximum drawbar efficiency.

  • Thirdly, an improvement in the overall efficiency of the locomotives was to be aimed for, as much to improve the output at the drawbar as to reduce the overall fuel and water consumption. Also of almost equal importance for our application of steam traction on tourist trains was the reduction of smoke and particle emissions whilst the locomotives were being worked hard up 1 in 72 gradients for lengthy periods at a time.

With this in mind and the acceptance that ultimately, to achieve truly significant results, quite significant rebuilding will be necessary including such major items as new cylinders, axlebox designs with associated frame alterations and major re-design of the whole steam circuit, a series of modifications were undertaken. Also the fairly average performance of the original designs meant that quite significant improvements could be made with relatively simple modifications, of course, the closer the locomotives approach the ideal the harder it becomes to make significant improvements in the performance.
Taking the work done by Ing. L. D. Porta as inspiration and spurred on by the achievements of David Wardale, so fascinatingly written about in his book ‘The Red Devil’, using information and guidance courtesy of Shaun McMahon a modest programme of improvements was started, and is still continuing to this day. Our position of being able to manufacture virtually all that was required ourselves, and the relatively manageable size and cost of the work involved has helped considerably in the scope of what has been achieved, and what, hopefully, will be achieved in the future.

Mechanical reliability.

The one significant advantage possessed by the locomotives of the KLR was the fact that as originally built, appreciable use had been made of roller bearings, this has more than proved to be worthwhile and is being applied more extensively whenever the opportunity presents itself. Main axleboxes, coupling and connecting rods were so equipped on three locomotives from new and, with the exception of occasional damage caused during dismantling, are still using the original bearings after up to fifteen years service (Fox and Badger currently average 3,000 to 4,000 miles annually with 11inch diameter drivers). Due to the limited life of the plain bearings (aggravated by inadequate size) more extensive use has been made of roller bearings which are now showing a real reduction in time spent on ‘routine’ maintenance, particularly as the bearing sizes we use show very little cost difference compared to machining and fitting a Bronze bearing. Ultimately the aim is to fully equip all the motion work, including all the valve gear suspension links with roller bearings, attention also being paid to better bearing sealing against ingress of contaminants.
Unavoidable wear, such as tyre wear, is progressively being reduced by the use of more appropriate grades of materials, better detailed design and measures such as driving wheelsets being machined larger than driven wheelsets leading to the removal of less material when re-profiling becomes necessary.
Another area of reliability and service life which can be one of the most expensive, insidious and time consuming is that of the locomotive boilers, our boilers are of all welded steel construction with no formed plates other than the barrel itself. Although the angular form of our fireboxes, particularly the outer wrapper is far from ideal, this is not an area that has given any trouble, so far. 
Operated from new with the then company policy of not using any form of water treatment it was inevitable that sooner, rather than latter, repairs would become necessary. The first retubing became necessary on Fox’s boiler (originally the most heavily used) after ten years in service, at the same time a section of the barrel just behind the front tube plate also required replacement due to corrosion. Only four years later the whole firebox required re-staying, this unfortunately, falling at the same time as Badger’s boiler became due for retubing and a similar barrel repair, culminating in two locos being out of service in quick succession, a considerable amount of work in only a few months and a not inconsiderable bill, even though most of the work was done ‘in house’.
As a result of this, and having no particular desire to do it all again with Hawk, which is now approaching its first ten years of service, it was decided to investigate the possibility of using some form of water treatment. It was at this time that Shaun McMahon supplied me with a copy of Porta’s paper on boiler water treatment and after some studying it was decided to source the appropriate chemicals and proceed with a trial application on newly re-stayed Fox. The treatment was also intended to eliminate scaling on the internal plates of the boiler, a problem that had not reached such serious proportions with ourselves as on some other railways, but something still worth preventing, the treatment was also intended to progressively remove old scale deposits. This last was something that would require close monitoring as it had been noted in other applications as causing considerable build up of loosened scale around the foundation ring and associated problems requiring, initially, frequent washing out.
Despite a quite extensive list of potential problems highlighted by Porta, causing some considerable amount of water analysis and monitoring in the early stages of application, the system worked remarkably simply and effectively from the start. The only minor problem was caused by the recommended antifoam being unavailable, and the substitute not being quite as effective as the manufacturer claimed, the dose included in the mixture had to be doubled from that originally advised after a brief bout of priming. The routine application consists of the drivers simply adding a pre-measured amount of treatment every time water is taken and that is it.
The end result after two years of use is regular blow downs are eliminated, only being required after significant periods when the boiler water chemistry becomes concentrated enough to override the effects of the antifoam (usually around thirty to forty days in steam), washout periods are extended to basically the annual inspection (around one hundred and twenty days in steam) and the boilers are now internally very clean, as commented on by our boiler inspector.

Consistency of performance.

Upon completion of the railways extension to the ultimate terminus of Shelley, making the total length of run 3 ½ miles with a gain in height of some 220 feet, the reality of operating an hourly service over the full length of the line became suddenly very harsh. The time, effort and fuel required to clean the hard pressed fires every second trip, with clinker covering the entire grate, pieces too big to fit through the firedoor, and barely five minutes lay over at either end it was quickly realised that something had to be done. The rapidity with which things could go ‘pear shaped’, particularly with drivers of varying ability, under such conditions had to be experienced to be believed!
The overall improvement in output consistency sought was obviously going to require a wide ranging scheme of work, simply attacking the firing problem was not going to achieve the necessary results, particularly as the locos were being pushed harder than ought really be required, the limitations of the whole steam circuit being particularly acute. Coincidentally it was around this time that David Wardale published his seminal book on the work he carried out in South Africa and this provided the inspiration and much needed technical information for the modest level of progress made so far with our locomotives.
The first step was to incorporate a basic form of Porta’s Gas Producer Combustion System (GPCS) where steam, mostly from the cylinder exhaust with the remainder made up from the air pump exhaust and blower, is fed in under the fire and is drawn through the firebed along with the primary air, cooling the fire to below the temperature at which clinker forms, in an endothermic reaction, which produces combustible Carbon monoxide and Hydrogen gas (producer gas). This gas is thoroughly mixed with more, secondary air, above the fire, where it is burnt cleanly with no smoke and a minimum of excess air, this fact of admitting the majority of the air required for combustion above the fire also means there is significantly less gas rising through the firebed to entrain and carry away small coal particles. This last is what gives the GPCS such an increase in efficiency over conventional combustion at high rates of steaming, but at the rates we normally operate at the overall improvement in efficiency is quite marginal, the more significant advantage in our application, as well as the almost total lack of clinker making the daily operation of the locos much simpler and reducing losses due to fire cleaning, is the very real reduction in char throwing and smoke, particularly when operating open carriages on a continuous gradient.
The physical alterations required are remarkably simple, although there is no denying that hollow stays in the firebox sides would significantly improve the turbulent secondary air mixing over the fire, our present system of directing the secondary air through a group of tubes on an air box welded to the inside of the firedoor works remarkably well, helped considerably by the relatively large door area in relation to the grate area and the deep fireboxes. The steam supplied to the ashpan consists of approximately 8% of the main cylinder exhaust led via a pipe direct from the valve outlets through a changeable restrictor to a perforated pipe alongside the primary air opening where it is mixed evenly with the incoming air to the underside of the fire. This is supplemented by the air pump exhaust being fed into the supply pipe upstream of the restrictor, part thus providing a fairly constant background level of steam to the fire whilst the remainder is exhausted through the main blast pipe, this supply also reduces the amount drawn from the main blast while running without significantly increasing the steam/air ratio. Also a small supply, about 20%, is led from the blower valve to ensure that whenever the blower is used there is always an adequate ashpan steam supply to avoid clinkering.
Simultaneously incorporated with the GPCS were alterations to the draughting of the locomotives, partly to compensate for the loss of available steam for drawing the fire because of the supply taken to the ashpan, and also as a means to reduce the exhaust back pressure at the cylinders to aid in increasing the cylinder efficiency and thereby reducing the steam demand, and hence the load on the boiler and fire, for the same work output.
Unfortunately due to the then stipulated constraints of working within the existing chimney profile a less than satisfactory arrangement of Lempor type exhaust was arrived at. This included a multiple jet blast nozzle, converging/diverging nozzles were settled on after some basic tests with a 1.1 to 1 exit area to throat ratio in comparison with simple converging nozzles. A large radius entry section to the chimney was incorporated and a tapered diffuser section taken to the absolute limit of the chimney wall thickness, resulting in an increase of exit area of some 56% on Fox and Badger and 80% on Hawk. The end result was an increase of blast nozzle area of around 25% whilst still maintaining more than adequate steaming despite a loss of 8% of the exhaust to the ashpan and the necessary increase in firebox turbulence requiring a greater smokebox vacuum.
Latter work involved some improvements to the steam circuit and alterations to the valves and cylinders where possible within the limitations of the existing components with the aim of further reducing steam consumption and hence the load on the boiler, also simultaneously increasing the available output from the locomotives. Latter still the cylinder insulation on Fox was significantly improved, an easy matter when it was totally lacking to start with! The reason behind this was to reduce condensation losses in the cylinders as much as possible, tests on stationary engines in warm still air showing up to fifty percent of steam admitted being lost to condensation, this last work alone making a noticeable difference and similar work is planned for Badger as soon as practical.
A problem that has handicapped the operation of our locomotives on the relatively steeply graded line has been the question of adhesion, or lack of it! In poor conditions, which are not infrequent when operating a year round service, conditions could become severe enough to stall Hawk, a total adhesion loco of some considerable weight, prior to the fitting of effective sanding equipment. Riding on the front of a locomotive, holding on to the lamp bracket with one hand and sprinkling sand on the rails with the other, Darjeeling style, has to be experienced!
Initial attempts with traditional gravity sanding proved very unreliable, a combination of fifty percent reverse running throwing water up the sand pipes, and severe cross winds on the exposed embankments carrying the sand away before ever reaching the rail head, made the whole exercise pretty futile. A more concerted effort to solve the very real problem resulted in our present system of steam sanding, which, despite it’s remarkably simple design, has proved to be very effective and robust, requiring zero maintenance and working every time without fail. The system was first fitted to Fox around twelve years ago and is still as installed, only ever being removed and inspected if required to gain access for other work such as wheelset removal. An extra benefit is the very small amount of sand actually used as it is placed directly between the wheel tread and the rail at the point of contact and so very little is actually wasted.
More work on the adhesion problem is still pending as soon as time permits, such as rail de-sanding jets behind the driving wheels to clear the drag inducing layer of sand off the rail head before the trailing load passes over it, the last thing you really needed in conditions of poor adhesion is more resistance behind the drawbar, this is one of Porta’s ideas also used by David Wardale in South Africa. Another idea of Porta’s is to use steam jets to ‘wash’ the rail head clean of contaminants ahead of the driving wheels, adhesion on a CLEAN wet rail being almost as good as a dry rail, this idea is attractive for many reasons, not least of which is you are not going to run out of working medium and it also eliminates the abrasive sand particles getting carried into bearings and wearing surfaces.
Closely related to the adhesion problem as well as time spent out of traffic for ‘routine’ maintenance is the question of tyre material and profile, time between turnings can be reduced not only by the obvious solution of optimum material but also by such devices as Porta’s ‘high adhesion’ profile keeping the tread clear of the area of the rail head likely to be most contaminated and also preventing contaminants from the locomotive itself finding their way onto the tread, reduction in wheel slip, even if undetectable, all contributes to reduction in wear. Even the most obvious, but often overlooked problems, such as directing excess axlebox lubrication away from the wheels can have a very real effect, along with ensuring drains and similar sources of contaminants are directed well away from the rail head. Other ideas, such as machining the driving wheelset slightly larger than the driven wheelsets to offset the increased rate of wear on this axle and combat the ‘micro’ slipping that occurs every wheel revolution, help maintain a more even grip throughout the period between turnings, instead of the usual decline, often largely unnoticed, connected with traditional maintenance methods. All the above mentioned ideas are gradually being incorporated into all of the railways locomotives as the opportunity presents itself although it is expected to be some considerable period of time before all the predicted benefits become fully apparent.

Thermal efficiency.

Although an improvement in thermal efficiency was not the primary aim of the development work on our locomotives it is an integral part of the ongoing progress, as any improvement in efficiency, as well as being seen as a means of reducing fuel and water input for a fixed output, also increases the locomotive output for a similar energy input.
Many aspects of the work so far achieved, and still to be achieved, have a significant impact on the overall efficiency and output of the locomotives, the individual steps achieved so far, even the smallest, cannot be seen in isolation as each interacts with all the others creating an overall impact on the locomotive performance. Even areas that previously have been completely ignored, such as complete cylinder insulation and far thicker insulation on the whole of the boiler all have their part to play alongside more obvious areas such as improved cylinder and exhaust design, all combining to increase the overall thermal efficiency, resulting in more work at the drawbar with less coal input to the fire.
Even though the process of development is ongoing and will probably never reach an ‘ultimate’ point, the progress so far achieved has made a positive impact on the operation of the railway with a significant reduction in the amount of time spent on maintenance, more consistent locomotive performance with appreciably heavier loads and a cleaner, more ‘open coach’ friendly level of emissions. Whilst the existing designs will continue to limit how much can ultimately be achieved there are still significant improvements to be made, and given a blank sheet of paper a completely new design could show even bigger advances beyond what has already been achieved. At this point in time our locos are far from being the ultimate expression of a ‘modern’ steam locomotive but are starting to show promise, and hopefully in the not too distant future will be able to demonstrate what can truly be achieved with modern steam traction in the traditional format.

Ian Screeton,
1 September 2006

Click here to return to the modern steam locomotive developments page.

Rob Dickinson