US Class 59s
In the 1980s, the Class 59s became the first US-built and the first privately owned diesel locomotive to operate regularly on the British main line network. DAVID CLOUGH examines the origins and early trials of the ‘59s’ and their British counterpart - th
A look back at the Class 59s - the first US-built diesel locomotive to operate regularly on the British main line network.
This year marks 30 years since the arrival of the Class 59s, which revolutionised thinking about locomotive performance in Britain. It also marks 25 years since British Rail’s Railfreight Sector staged trials involving the type and its British counterpart (Class 60s), to assess the capabilities of the designs.
By the start of the 1980s, orders were in place for two Type 5 (3,000hp-3,500hp) freight diesels - Classes 56 and ‘58’. The former had been a hastily produced design that was proving to be an engineering challenge. In contrast, Class 58 was still on the drawing board.
Thoughts now turned to a medium-power machine, later referred to as Class 38 and sketched out as being between 1,800hp and 2,200hp, and which would be required by the early 1990s as a replacement for Classes 20, ‘31’, ‘33’ and ‘37’. A starting tractive effort matching a Class 56 was sought, but with power sufficient to operate to schedules set for a pair of Class 20s. By 1983, consideration was being given to buying an overseas design for trials.
One issue was the low utilisation of freight motive power, which caused poor returns on capital. Additionally, one design might have sub-classes for deployment by different sectors within BR. This was to happen during the Class 37 overhaul programme - some examples received electric train supply facilities, while others were to be increased in weight to offer a higher starting tractive effort.
Evidence of the state of flux of future locomotive procurement was referred to by the Transport Minister, in his letter of June 14 1984 to the BR Chairman. In sanctioning the small second tranche of 15 Class 58s, the Minister concurred that this was appropriate because of a lack of clarity about future needs.
Regarding new-build locomotives, three types were in the frame a year later: a four-axle Type 2, a Type 3 and a Type 5. But years then passed, as did the proposals for new locomotives. By 1988 one idea under consideration was to build a three-axle Type 1 (designated Class 18), of 800hp and passed for 50mph. This appears to be little more than a modernised version of the ill-starred Class 14 of the 1960s!
Events outside BR were now to have a significant impact on freight traction. Foster Yeoman Ltd (FY) operated a quarry at Merehead in the Mendips. The majority of the stone was shipped by rail to depots mainly across southeast England.
Over the years, a range of traction had been used on these trains, including Class 52s and pairs of Class 37s. As Class 56 construction progressed the Western Region ( WR) received an allocation, and eventually these began to be diagrammed for this traffic, with pairs used on the super-size 4,300-ton workings to Acton.
As noted above, the Class 56 had already been recognised as a problem child. Failures were common across the entire fleet, but especially with the WR's allocation, which were worked at full power for longer durations than those used on coal flows on the Eastern and London Midland Regions. During one week in 1983, Foster Yeoman counted 34 different ‘56s’ at Merehead.
Failures and the non-availability of traction disrupted supplies to FY’s customers. It was not uncommon for fleets of road wagons to have to be hired in at short notice, and at considerable inconvenience. Even when trains did run, 40% arrived late, affecting terminal capacity, wagon utilization and customer satisfaction.
Contemporary BR practice across its diesel locomotive fleet was to roster for typically 75% availability for traffic, with the remainder sidelined for maintenance (either planned or unplanned). No commercial haulage business would be viable if only three-quarters of its capital assets were in use at any one time, yet BR was charging Foster Yeoman for such a level of availability for its traction. In fact, WR Class 56 availability sometimes dropped to as low as 60%.
FY calculated that amending the movements from its quarry would enable just six locomotives to be required, with maintenance carried out overnight and at weekends when Torr Works was closed. Having a dedicated pool was accepted as offering cost savings.
Neither Class 56s nor ‘58s’ were judged likely to fulfil the requirement for an availability of at least 90%. And so, with BR’s consent, FY looked overseas. Railfreight was already considering importing at least one locomotive for evaluation, and FY offered the perfect opportunity. The specification required:
A minimum consistent availability of at least 95%.
A level of reliability that would enable six locomotives to haul four million tons of aggregates annually, to schedules and timings then in place for Class 56s.
Straightforward maintenance within limited servicing facilities at Torr Works, to be capable of completion at weekends when the quarry was closed (FY already had facilities for the fleet of wagons that it leased and maintained by Procor).
Proof that the above was already being achieved consistently by similar locomotives.
Three overseas manufacturers were invited to submit proposals. Henschel offered its DE2500 design, three prototypes of which were operating in Germany and had an AC
No commercial haulage business would be viable if only three-quarters of its capital assets were in use at any one time, yet BR was charging Foster Yeoman for such a level of availability for its traction. In fact, WR Class 56 availability sometimes dropped to 60%.
Except for the observance of signal and other checks, and when descending Beattock and Shap, full power was used almost throughout the runs, and the power controller was held fully open for part of the descent of both these banks in order to reach 75mph.
traction package. General Electric of the USA put forward its -7 series that also used a three-phase drive. Finally, General Motors (GM) Electro Motive Division proposed its JT26CWSS.
While the JT26CWSS was based on the proven SD40-2, GE’s -7 series was still experimental, and FY did not wish to use its trains as a proving ground for a design that turned out to be a poor performer in North America. FY already had experience with GM products, by virtue of its shunting locomotive at Merehead.
The addition of GM’s patented Super Series wheelcreep traction control for high adhesion enabled FY to cut the order from six to four examples, because the new design would be able to handle a 4,300-ton train singleheaded. A ‘59’ has a maximum tractive effort of 114,000lbf, compared with 61,600lbf for a Class 56. Subsequently, increasing traffic led to the acquisition of a fifth example.
BR’s Department of Mechanical and Electrical Engineering liaised with GM over the design of what became Class 59s. For a start, the SD40-2 was outside the BR loading gauge and had to be slimmed down. Prevailing limits on maximum axleload and bogie rotational and torsional stiffness could not be exceeded.
For European operation, a large silencer was needed. This meant that roof-mounted dynamic braking equipment could not be installed.
BR also required a cab based on that of a Class 58, which had been accepted by the train drivers’ union and so circumvented any difficulty from that quarter. Each wheel was to have its own brake cylinder, rather than the standard GM arrangement. Finally, BR stipulated the fitting of safety equipment such as automatic warning system, driver’s safety device and headlamps.
GM did refuse to accommodate any changes proposed to its basic package, however. This included conversion to a control system of 110V.
Agreement was reached between FY, GM and BR in November 1984, and included a 12-month delivery date. In fact, the quartet was shipped at the start of 1986 and offloaded at Southampton between January 21 and January 23. All went to the Railway Technical Centre at Derby on January 24 for inspection, including weighing.
Two locomotives were returned to Merehead to facilitate crew training at Westbury and Acton, while the other two were put through tests based on Derby. Toton depot’s load bank was used to ensure the specified power was being developed. Despite it being used for the 3,300hp BR Type 5s, the similarly rated GM diesel overwhelmed the equipment and burned it out.
Other commissioning work encompassed noise levels and ride quality. The ride coefficient was found to be outside BR’s parameters, and work was needed on this. Haulage trials on the Western Region during early February established the actual performance to be in line with specification.
New rosters were introduced from May 11 that enabled 17,470 tons of stone to be shipped daily. Following a trial, the working to Acton was loaded to 46 PTA hoppers instead of 43.
BR was contracted to maintain the little fleet at Torr Works, and initially 12 staff at Bristol Bath Road were trained for the purpose. Just as with its wagon fleet, FY did not want to become involved in this area.
During the first four years of operation, the class achieved 97.7% availability against a target of 95% (availability here was measured by dividing the number of actual trains worked by the number booked to be worked, and multiplying by 100). Reliability in terms of technical failures was far superior to any British diesel, and some of the faults were attributable to the equipment specified by BR, rather than GM kit.
Good news travels fast, and the availability, reliability and haulage potential of the ‘59s’ initiated a rethink about future freight traction. Suddenly a world of possibilities appeared before the eyes of the business managers, and a new strategy was drawn up in July 1986.
Interest was heightened because GM had already made an attractive offer to supply 30 locomotives for £ 30 million. Some senior professional railwaymen who should have known better thought BR should just take up this offer, but were slapped down from on high
because of the requirement to follow proper procurement procedures for a nationalised industry.
Another issue was that British manufacturers were behind America and Germany (in particular) in technology advancement, and the question was whether BR should allow them time to catch up. Brush Traction was developing a high adhesion system and using Class 58, 58050, as a testbed, but was struggling to get the technology to work.
Government spending limits was inevitably the elephant in the room when capital expenditure was being considered. Prior to the July 1986 rethink, the freight traction renewal policy had been to acquire several batches each of five locomotives from different manufacturers for evaluation, as a basis for Class 38s. This would have to be dropped if the intention was now to pursue a Class 59 procurement process.
A debate ensued over how to proceed, and included the option of hiring (rather than buying) the products. However, it was pointed out that hiring would be seen as a device to circumvent government spending limits, and so could not be adopted.
Encouraging freight customers to follow FY’s approach did prove feasible, however, and FY’s experience led ARC Southern to acquire four ‘59s’ for its stone flows from Whatley quarry (also in the Mendips). Once the electricity industry had been privatised, National Power bought six machines for shipments associated with its power stations.
What eventually emerged from the July 1986 review was a specification for a heavyhaul Type 5 with a requirement for a high availability. Railfreight was no longer happy at having to cope with the cost of one-quarter of its fleet continually under repair, while the Government insisted its operations should not be subsidised. The outcome was the ordering of 100 Class 60s from Brush for the trainload side of Railfreight, after a competitive tendering process.
As the decade drew to a close, with Trainload’s heavy-haul requirements in hand, the question “whither Railfreight Distribution (RfD) diesel traction” now arose. Class 92s were on order for Channel Tunnel traffic, but as the business entered the 1990s it lacked modern diesel motive power.
Two options were considered. One was for a 75mph machine to haul intermodal traffic, and the other was for non-intermodal trains operating at up to 60mph. While the proven performance of the ‘59s’ made them strong candidates, additional information about
performance in the high-speed range was required.
Arrangements were made with ARC to use recently-arrived 59104 on a series of runs (spanning one or two weeks), both to evaluate theoretical train schedules and to assess locomotive ride quality and braking at up to 75mph on simulated RfD services. Class 59s were passed for just 60mph, although since arriving 59104 had been put through tests at up to 80mph.
The initial plan was to use flat-bed Freightliner vehicles, loaded with scrap rails to give the required load. Routes identified for the trials were: Southampton-Birmingham-Coatbridge. Southampton-Leeds. Crewe-Hereford-Cardiff. Melton Mowbray-Cricklewood. Burngullow-Irvine. Dover-Willesden. Willesden-Oxford-Birmingham via the Lickey incline. There was some concern not just about high-speed ride quality, but also about braking performance at such speeds. The issue centred on prolonged braking if an emergency application became necessary, and so descents of Shap and Beattock banks on the West Coast Main Line were selected for this aspect of the trials. Of course, dynamic braking equipment had been omitted from the ‘59s’, to provide space for a suitable engine silencer.
Several pre-test issues were identified. GM had indicated a load limit for the Lickey of 900 tons, and this was to be confirmed. 59104 was to be fitted with yaw dampers to improve its ride, and with some test equipment to monitor this and braking standards. And the tryes were to be checked to see whether re-profiling would be necessary. BR's solicitor was to be asked to indemnify ARC for any loss or damage that might arise during the trials.
Although planned for March 1991, there was considerable difficulty in sourcing the desired trailing load. The months slipped by, and it was late June before two clients were able to offer materials of a combined 1,800 tons (during a two-week window in late July/early August, when their plants would be closed).
In the meantime, the availability of a Class 60, whose first few examples were being released after construction and commissioning, enabled this type to be assessed as well. As with the ‘59s’, the heavyhaul ‘60s’ were limited to 60mph, but clearance for 75mph was sanctioned for the trials. Changes were also made to the routes selected, not least because there would be insufficient time to cover all those proposed originally.
Allerton depot supplied ten wagons to British Alcan at Lynemouth, for 597 tons of aluminium loaded into 20ft and 30ft containers, while 14 wagons for 1,178 tons of steel slabs in flatrack containers from British Steel at Ravenscraig came from Glasgow Freightliner terminal. The wagon types were FFA and FGA.
Neither Class 56s nor ‘58s’ were judged likely to fulfil the requirement for an availability of at least 90%. And so, with BR’s consent, FY looked overseas.
Included in the steel train was a 20ft ‘L’-type container holding a range of spares for 59104, which in the event were not needed. When marshalled together, the trailing load was 1,775 tons.
Depot management at Cardiff, Crewe, Eastfield, Eastleigh, Stratford and Thornaby had to be primed to deal with fuelling and the provision of servicing and stabling facilities. Men from Westbury with Class 59 traction knowledge undertook the driving, accompanied by pilotmen with relevant route knowledge.
59104 had entered service during October 1990, whereas 60006 had completed commissioning in April 1991. Both examples had been engaged on the movement of aggregates - the latter based at Leicester.
By July 28, 59104 had worked light from Whatley Quarry to Mossend Yard, where the wagons were stabled. The combined load of 1,775 tons was used only on Days 2 to 5. On Days 1 and 6 to 9 the train comprised only the steel element of 1,178 tons, while on Day 10 60006 hauled just the 597 tons of aluminium.
Day 1 ( July 29) involved a return run from Mossend to Carnforth to assess the braking and performance of the composite-type of brake block fitted to 59104 while descending Beattock and Shap. Next day these tests were repeated, but with the journey continuing to Crewe for refuelling and then to Temple Mills Yard in east London.
The following three days were used to assess 59104 over typical RfD routes from Temple Mills to Tees Yard, then to Eastleigh via Derby, and finally north to Crewe via Birmingham. Finally for the ‘59’ came a return trip over two days from Crewe to Westbury with just the steel portion.
Days 8 and 9 involved 60006 and the steel load running from Crewe to Mossend and back for brake evaluation. Matters were concluded on August 9 with the haulage of the aluminium from Crewe to Cardiff.
Operationally and technically, all went to plan. The report on the trials was dated December 9 1991, but the results of the braking and riding measurements were not included. Both locomotives were judged to have put up “very similar” performances. Neither experienced any difficulty in starting any of the trains, although only one trip (with 59104) encountered rain.
Projected timings had been prepared, and in the lower speed range these were confirmed. At higher speed, however, the projections proved optimistic and during actual running it took much longer to attain maximum speed.
In fact, except for the observance of signal and other checks, and when descending Beattock and Shap, full power was used almost throughout the runs, and the power controller was held fully open for part of the descent of both these banks in order to reach 75mph. The air resistance of the high-sided containers was said to be a factor here, and 65mph was a normal top speed unless travelling downhill.
Equipment to measure fuel consumption could not be fitted, and so fill-to-fill quantities had to be recorded. Contemporary practice with Class 59 was to leave the engine idling while stabled on depot, but an adjustment was made to the fuel consumed for this.
No journey/load combination was made by both locomotives, and so a direct comparison of fuel consumption under the same conditions could not be made. When hauling the 1,178ton load, 59104 used over 11.5 litres per mile while 60006 used less than ten litres (although the runs were not over the same routes).
These trials had no impact on BR diesel locomotive procurement because no new orders were made prior to privatisation. Class 59s continued to operate at near-100% availability, and the American owners of the privatised freight business bought 250 General Motors 75mph diesels (though not ‘59s’).
By contrast, the Class 60s experienced some early technical issues but then settled down to reliable operation, although the class turned out to be a dead-end in procurement terms.
The assistance of Brian Penney in the supply of information for this feature is kindly acknowledged.
Good news travels fast, and the availability, reliability and haulage potential of the ‘59s’ initiated a rethink about future freight traction. Suddenly a world of possibilities appeared before the eyes of the business managers.