Computing and Innovations in Steelmaking

Computing comes to south Wales: Steelmaking in Port Talbot

Ray Neate and John Dacey’s pioneering computer simulation work at the Abbey Steel Works, Port Talbot.

John V Tucker

The development of a simulation of steelmaking operations in the Steel Company of Wales that Ray Neate and John Dacey reported in 1958 was the result of two man years of work in designing and coding a simulation model in Port Talbot to run on a Ferranti Pegasus in London. The work was published in final form in Volume 2 of the British Computer Society’s new Computer Journal in 1959. It attracted attention and afterwards the Steel Company of Wales purchased a Pegasus, which arrived in Port Talbot in 1960.

Background

The Abbey Steelworks at Port Talbot was built by the Steel Company of Wales, with a budget of £60 million. This project was planned by the British Iron and Steel Federation in 1945 and was a component of the UK Government’s post-war reconstruction. Existing steel making at Margam was to be transformed with a new strip mill, melting shop, and cold mill next to Guest, Keen, and Baldwins. In addition, the blast furnaces were to be rebuilt and the docks for receiving raw materials improved. This new, vast steel works established the state of the art in Europe steelmaking.

An excellent recent introduction to the works is Miskell (2017).

Operations Research at Port Talbot

In the design and operation of the Abbey Steelworks an important step was the creation of a strong operational research unit at Port Talbot. Ray Neate’s and John Dacey’s computer simulation is one of several pioneering operational research projects that arose in the early years of the Abbey.

Operations research was hardly known and practiced in industry of the early 1950s. So when operations research began in the Steel Company of Wales in 1952 it was pioneering. The unit was multidisciplinary created in part from existing staff by H G Jones. Its focus was on problems that arose in production planning and control. Internationally interesting research soon emerged, not least by Jones and his colleagues. One example is Jones and Lee (1955) — surely one of the first industrial applications of Monte Carlo methods. The Monte Carlo algorithmic method was conceived by Stanislaw Ulam in 1946 and used for essential and secret calculations for the design of the hydrogen bomb: the method became ubiquitous in science and engineering. H G Jones later became Professor of Operations Research at Strathclyde University from which he continued to publish interesting and novel contributions, e.g., integrating people’s behaviour into operations research on production processes.

Jones reflected on his early years at Port Talbot in Jones (1994). An invaluable account of the first 10 years of operational research at Port Talbot is Thomas (1963). G W Thomas became manager of operations research in 1957 and did much to develop computing in the unit.

Computing

Data is was recognised as essential to the work of the unit, as was computation, which used Facit and similar electro-mechanical calculating machines. In the mid 1950s the unit became interested in computers. Following Ray Neate’s and John Dacey’s work, the Steel Company of Wales purchased a Ferranti Pegasus Mark 1. It cost £49,450 in July 1959 and was delivered in 1960. Continuing to develop the unit’s extensive publication record, the move to computers was discussed in Cartwright and Thomas (1961) and Thomas (1961).

Thus, computing takes root in south Wales through the work of operational researchers at Port Talbot, charged with designing the Abbey Works.

Photograph of John Dacey and Raymond Neate

References

W. F. Cartwright and G. W. Thomas, The integration of production planning, electronic data processing, and process control. J. Iron Steel Institute (London), 198 (1961), 250.

H. G. Jones, Early OR in the Steel Company of Wales, The Journal of the Operational Research Society, 43 (6) (1992), 563-567. https://www.jstor.org/stable/2583012

H. G. Jones and A. M. Lee, Monte Carlo methods in heavy industry. Journal of the Operational Research Society, 6 (3) (1955) 108−116. https://doi.org/10.1057/jors.1955.12

H. G. Jones and G. H. Jowett, The handling of uncontrolled data, Operational Research Quarterly, 6 (2) (1955), 49-61.

L. Miskell, Doing It for Themselves: The Steel Company of Wales and the Study of American Industrial Productivity, 1945–1955, Enterprise & Society, 18 (1) (2017), 184-213. https://doi.org/10.1017/eso.2016.55

R. Neate and W. J. Dacey, A simulation of melting shop operations. The Computer Journal, 2 (2) (1959), 59–67. https://doi.org/10.1093/comjnl/2.2.59

D. T. Steer and A. C. C. Page, Feasibility and financial studies of a port installation. Operational Research Quarterly, 12 (3) (1961), 145–160. https://doi.org/10.2307/3006772

G. W. Thomas, The Ferranti Pegasus computer installation of the Steel Company of Wales Limited. Elektronischedatenverarbeitung, 5 (1961) 13.

G. W. Thomas, Operational research in the Steel Company of Wales Limited. Operational Research Quarterly, 14 (3) (1963), 247–262.

History of Computing Collection