Utilization of structural steel in buildings

Over one-quarter of steel produced annually is used in the construction of buildings. Making this steel causes carbon dioxide emissions, which climate change experts recommend be reduced by half in the next 37 years. One option to achieve this is to design and build more efficiently, still delivering the same service from buildings but using less steel to do so. To estimate how much steel could be saved from this option, 23 steel-framed building designs are studied, sourced from leading UK engineering firms. The utilization of each beam is found and buildings are analysed to find patterns. The results for over 10 000 beams show that average utilization is below 50% of their capacity. The primary reason for this low value is ‘rationalization’—providing extra material to reduce labour costs. By designing for minimum material rather than minimum cost, steel use in buildings could be drastically reduced, leading to an equivalent reduction in ‘embodied’ carbon emissions.


SECTION 1: BUILDING DATA
This section details the results for each of the 23 buildings analysed. As agreed with the providers of the raw data, each building is identified only by a number, with the following information provided:  Building type;  Number of beam data obtained and number analysed;  Table with summary of results by floor and overall;  Graph of frequency of occurrence against utilisation ratio for each floor and overall;  Plot of beam layout on each floor analysed showing utilisation ratio of each beam;  Graph of frequency of occurrence against utilisation ratio for the columns in the building. For all buildings it was possible to provide the first four items. However limitations in the data resulted in three categories of building for the remaining two items:  For 17 buildings over 70% of the beams on each floor could be plotted, and once this level was reached the floor was deemed finished, as patterns were clear. Where necessary to complete the floor geometry, and so aid comprehension of the data, omitted beams were added in manually (coloured grey). Column locations were also added manually for this reason.
 For 6 buildings (#s 8, 9, 11, 16, 17, 21) there was insufficient information on beam layout to produce plots;  For 1 building (# 10) there was insufficient information to produce a graph of column data.
For graphs, utilisation ratios are groups into bands of 10% to aid clarity; these bands are inclusive of the identifying upper bound, for example the data point at 0.2 includes U/Rs from 0.11 to 0.20.
For all plots of beam utilisation ratio per floor the legend below is used:

Engineer's comments
A proportion of beams governed by construction loading scenario, otherwise a standard building. A spot-check of beam sizes did not reveal any further rationalisation by fabricator. Robustness was not a governing criterion.

Engineer's comments
The applied loads were reduced late in the project programme -too late to redesign, which resulted in spare capacity in places. Deflection governed most elements' design.

Engineer's comments
Computer model used mainly for stability and column design purposes -may explain why so many beams omitted from analysis. Design around edges governed either by vibration or by minimum sizes for façade supporting steelwork (to facilitate faster construction).

Building #7
Type: school 766 of 908 beams analysed (84%)  There was not sufficient information to plot the Top Roof level.

Engineer's comments
Vibration governed in some places but mainly stress and deflection governed.

Building #9
Type: office 512 of 606 beams analysed (84%) Figure 34: graph of frequency of occurrence against utilisation ratio for beams by floor and overall for building #9 No data on beam layout was available for this building; therefore floor plots could not be created.

Engineer's comments
Vibration was a governing criterion in a small area. Many of the beams not analysed were speciallyfabricated beams, expected to have high U/R.

Engineer's comments:
Deflections governed design. Not surprised that had high U/R as had time to design thoroughly and no late changes were made.
Insufficient information was available about the columns in this building to allow analysis.

Engineer's comments
Steelwork was rationalised to enable cheaper procurement -fabricator further rationalised the design also.

Engineer's comments
Steelwork was rationalised to enable cheaper procurement -fabricator further rationalised the design also. Regular column grid prevented by client desire to provide minimum required area (lower heating costs) and to minimise cladding cost.  This building was composed of many different levels with less than 20 beams on each, which did not merit plotting individually, hence only the first floor is examined in detail.

Engineer's comments
Steelwork was rationalised to enable cheaper procurement -fabricator further rationalised the design also.

Building #14
Type: school 751 of 760 beams analysed (99%)  Roof column layout drawings were not available so Figure 53 was plotted using engineering intuition based on the other two floors.

Engineer's comments
Many beams governed by loading during construction. Increasing mass to take this load was deemed the cheapest option, as other solutions required more labour on site. A small number of beams were governed by vibration concerns.

Engineer's comments
Complex procurement involved fabricator twice 'transposing' sections between UK, Russian and Chinese steel catalogues (for cost reasons), choosing heavier section each time 'to be conservative'. Design was originally stress-governed however.

Building #16
Type: mixed use residential 364 of 536 beams analysed (68%)  Figure 62: graph of frequency of occurrence against utilisation ratio for beams by floor and overall for building #15 Insufficient data on beam layout were available for this building to create layout plots.

Engineer's comments
Complex procurement involved fabricator twice 'transposing' sections between UK, Russian and Chinese steel catalogues, adding weight each time 'to be conservative'.

Building #17
Type: mixed-use 631 of 947 beams analysed (67%)  Figure 64: graph of frequency of occurrence against utilisation ratio for beams by floor and overall for building #17 No data on beam layout was available for this building; therefore floor plots could not be created.

Engineer's comments
Vibration governed much of design; this combined with desire to minimise structural depth (to reduce cladding costs) lead to a heavy solution. Data was only obtained for floors 2 and 10 of this 11-storey building.

Building #19
Type: school 499 of 527 beams analysed (95%)  There was insufficient information to construct a plot for the roof level.

Engineer's comments
Vibration considerations governed design of many areas.

Building #21
Type: residential 71 of 73 beams analysed (97%) Only data for one floor was available for this building, and no information about beam layout.

Engineer's comments
Smaller beams oversized to allow faster assembly. Repetition in section sizes encouraged to facilitate faster construction.

Building #23
Type: school 528 of 558 beams analysed (95%)  Figure 85: graph of frequency of occurrence against utilisation ratio for beams by floor and overall for building #23 The presence of gridlines with identical names but non-identical coordinates required that the figures below were assembled manually in places, using engineering intuition to assess where beams were located.