Go back to article: Visualising electricity demand: use and users of a 3D chart from the 1950s

Use and users revealed

Speculation about the model’s use for engineering or commercial planning directed research in the MSI archives towards the Electricity Council’s Chief Engineers’ Conference and the Chief Commercial Officers’ Conference. In following a lead through the former’s Research and Technical Planning Committee, we discovered that this model was by no means unique. A list of utilisation research papers amongst the standing committee papers contained reference to technical report K/T107, which covered:

‘The utility and preparation of two- and three-dimensional representations of load, more particularly daily load curves…load models, and load maps. An improved design of load model is described, and commercial arrangements have since been made for the supply of suitable cards, etc., for Boards wishing to use this form’ (CEA, 1956, p 3).

Without having worked with the object, and speculated on its use(s), it is unlikely we would have found this documentary source, or realised its significance. Only by using the object and archive together did we make the connection between the two.

Report K/T107 (Schiller, 1945) proved to be an invaluable source in understanding the object’s use and users. The illustrations were of immediate interest – one depicted a load model of identical construction to ours (see Figure 19), and another depicted identical cards (see Figure 20). The first illustration of the report showed load models made between 1916 and 1928, in North and South America and Europe (see Figure 21), indicating there was an existing practice of load representation (and data visualisation) that the models of the 1940s and 1950s had drawn upon.

Figure 19

ERA-type load model, 1945 (Schiller, 1945, p 6)

Figure 20

Cards for ERA-type load model, 1945 (Schiller, 1945, p 8)

Figure 21

Load models from North and South America and Europe, 1916–1928 (Schiller, 1945, p 5)

The report’s purpose was to recommend that electricity suppliers produce load models for ‘comparative load research’, ‘load analysis’ and ‘electricity-supply economics’. This was as we had deduced. Knowledge of the composition and trends in the demand for electricity were fundamental to the engineering and commercial planning of suppliers, and forecasting and load management techniques ensured the financial success of the industry. The author of the report, Paul Schiller, was an engineer with the Electrical Research Association (ERA), an organisation jointly funded by the government’s Department of Scientific and Industrial Research and subscriptions from members. Section K (hence ‘Report K/’) was established in 1941 to conduct cost research, with Schiller as the head, ‘pioneering the scientific analysis of load curves and consumptions’ (Hannah, 1979, p 203). A thorough investigation of the work of Schiller and Section K is beyond the bounds of this article, but he distinguished the purposes of ‘load research’ as ‘the ascertainment of the load characteristics of each of the principal classes of consumer…of interest chiefly in allocating costs’, and ‘the different types of load’ and management techniques in order to encourage ‘desirable loads, appliances and methods by sales propaganda [and] tariff measures’ and to discourage those that appeared ‘undesirable’ (Schiller, 1952).

Report K/T107 communicated that the ERA provided the specifications for the cards and cabinet in ‘standardised form’, and shared the names of printers and cabinet makers with ERA members. The report went on to advise:

‘The best way of building up an annual load model is by having each day’s load curve cut out and put into the receptacle immediately the readings become available. Upon completion of the first year, the load model for the subsequent year is added, and when this is complete, the model for the preceding year is removed and stored’ (Schiller, 1945, p 11).

It was also noted that the transparent cover could be easily removed to enable cards to be ‘picked out and assembled together’ for particular comparisons (Ibid, p 10). Together, these passages explained the need for data at a daily and aggregate level, and the reasons for producing a receptacle to hold the cards for this duration.

The report clearly explained the position of our load model within the history of load representations, the means by which it was constructed, and some of the motivations for its use. But was it used in other ways at the time, or subsequently? Who used this object, and when? These were questions for further research.

Component DOI: http://dx.doi.org/10.15180/180905/005