One of GE’s largest export customers in the “Universal” era was South African Railways (SAR).
SAR was Cape gauge, but with both an unusually generous loading gauge for that track gauge, and generous axle loading allowances on most of the main lines, although some were quite restricted. SAR had one the one hand pursued an aggressive electrification policy, and on the other had continued to develop steam traction into the 1950s, given that it had abundant coal reserves. Thus it was not an early starter in respect of mainline dieselization, although it was not as late as some African countries.
It started its mainline diesel fleet with 45 of the GE U12B model, as captured in this GE advertisement:
I have heard it said that at the time, SAR preferred four-stroke diesel engines over the two-stroke type. If that were in fact the case, it might have been a factor in the choice of GE as supplier. One supposes that EMD would have bid its G12, in B-B form, for this business. Alco I think would have had a problem in that it did not have an export model in the same power class.
Whatever the reasons for SAR’s choice of GE, it was the start of something big. (To borrow from the title of the book on GE commerical aircaft engines - "Starting Something Big".)
Having sampled diesel power, SAR’s next move was to dieselize its South West Africa (now Namibia) lines. These ran through desert country, and were far removed from the coal-fields. Thus both water and coal supplies were big problems for steam traction. Also these lines had a quite restrictive axle loading limit of 28 000 lb. That compares with for example around 33 000 lb for a baseline GE U18C of the time. Nevertheless, SAR wanted a U18C-sized locomotive and had asked bidders to propose an eight axle, 1-C-C-1 design. That wheel arrangement was a logical way to add carrying axles to a C-C design, in order to lighten the driving axle loads. SAR would have been able to observe such locomotives at work in neighbouring Rhodesia, which had deployed an English Electric 1-C-C-1 fleet starting in 1955. These had running gear arranged to minimize lateral railhead forces, minimize weight transfer and (by appropriate axle spacing) to provide rail bending moment relief. SAR also had a fleet of heavy 1-C+C-1 electric locomotives in service, but with their articulated trucks, their dynamic behaviour was quite different, and they were reputedly hard on the track and on themselves. (That was not so surprising. The Japanese National Railways (JNR) experience was that for acceptable tracking with articulated trucks on the Cape gauge at up to around 60 mile/h, 2-C+C-2 was better than 1-C+C-1. Thus it had used the former for electric passenger locomotives and the latter for electric freight locomotives.)
Apparently the 1-C-C-1 requirement did limit the number of bids, although I have never seen complete information as to who did actually bid. An EMD bid was unlikely, as it was at the time still pursuing strict standardization at the time. EMD licensee Henschel might have bid an EMD-powered special design. Henschel had been an important supplier of steam locomotives to SAR, and had also supplied a trial fleet of (essentially unsuccessful) EMD-powered diesel-hydraulic locomotives. English Electric probably did bid, basis its experience with this wheel arrangement, and Belgian builder Cockerill, a Baldwin licensee, was known to have bid. (Cockerill did though gain business in other African countries with C-C designs that could be traced back to the Baldwin AS616E export model built for RFFSA Brasil.)
Significantly Alco did not bid against the 1-C-C-1 requirement, even though it had access to a GSC 1-C truck that was probably designed against the SAR and possible other like requirements. It has been written that this apparent intransigence by Alco lost it what would otherwise have been certain business, and so handed a 115-locomotive order to GE and in turn rescued its locomotive programme. Whilst that might have been the case, the circumstantial evidence at least points in the other direction. SAR was already in the GE camp with its U12B fleet, so unless GE and/or its locomotives had performed poorly, one imagines that GE had the starting advantage of fleet commonality. GE’s Universals seemed to have been selling well enough since their release; one doesn’t get the impression that its new locomotive programme was in any danger of being axed. And at the time, Alco did not have an existing C-C locomotive that could have been simply adapted to the SAR requirement. The DL-540 – which had but one customer - was too big and heavy, and the DL-541, which somewhat addresses the DL-540 issues, was still in the future, although no doubt it could have been produced earlier than 1960. And had SAR really wanted an Alco bid on the 1-C-C-1, surely it could have made that very clear.
Anyway, GE got the business with its U18C1 model, and advertised it accordingly:
Apparently SAR had included a less preferred option for 230 “half-power” C-C locomotives, possibly as a backup in case no-one offered a satisfactory 1-C-C-1 design. There Alco would have been in a strong position with its DL-531. GE’s U9C would have required a little trimming to meet the 28 000lb axle loading limit.