Checking further on the authorship of the Sankey diagram I presented in this post, I came to the LCMP website at the University of Cambridge. LCMP? Yes … Low Carbon and Metals Processing. The engineering research group around Julian Allwood and Jonathan Cullen have three large research themes: WellFormed, WellMet2050, and WellMade.

The below Sankey diagrams are from the report ‘Going on a metal diet’ by Allwood, Cullen published within the WellMet2050 research theme.

The first Sankey diagram shows the global steel flows in 2007

the other the global aluminium flows in 2007:

One page 7 of the report the authors explain

“In our maps, the width of each line is proportional to the mass flow of metal. Values for the major flows are given in Mt (million tonnes). Steel flows less than 1 Mt and aluminium flows less than 0.05 Mt are not shown. Each major process step is shown as a vertical black line, with three possible outputs: useful metal (colored), process scrap (grey) and metal losses (black). Useful metal continues to flow to the next process step, while scrap loops back to the appropriate melting stage where it is recycled. Internal recycling loops, for example from the continuous casting processes for steel are shown small oval loops. (…)The working papers … give more detail about creating the Sankey diagrams

Unfortunately these two mentioned working papers are not (yet?) available on the website. These really fantastic Sankey diagrams have been compiled from different data sources. I thought I’d share them with you. Please visit the LCMP website and read about their other exciting projects.

This 1996 article on Energy Conservation Management (source: Charles M. Gottschalk: Industrial Energy Conservation, UNESCO Energy Engineering Series, John Wiley & Sons Ltd., Chichester, West Sussex, UK, 1996) has a Sankey diagram of the energy flow in a boiler system (will post that one separately one day, maybe). It also features the following steel factory material flow chart (flows are in MT per month, although the text says it is in tons/month):

I took this and converted it to a Sankey diagram, to better comprehend where the real material flows are, and where the biggest losses occur. Losses are distinguished from the other material flows with a
darker blue color. I actually produced two different versions: the first one sticks very closely to the original layout of the processes, the second one has more of a top-down flow direction.

I was unable to hook the flows to the corner of a process node as it is done in the original diagram (e.g. flow from cold mill to slitter line), Also, the little “bridges” where the arrow from circling to annealer crosses the three other arrows, cannot be reproduced in the same way with the tool I am using.

In the second version I need more space, but I think it is much more comprehensible as it sticks to a top-down flow direction. Losses branch off sideways. I added arrow heads for very thin arrows, otherwise they would sometimes be hardly visible. I refrained from putting the units behind each value to keep it like in the original.

Seems as if I have to much spare time, but this ain’t true…