twice the steel it needs

Problem statement

Design codes set a floor on how strong a structural member must be but no ceiling, and in practice engineers routinely provide far more resistance than the code requires: analysis of steel-framed buildings designed by leading UK firms found beam utilisation ratios (design effect of actions divided by design resistance) averaging below 50%, and "average utilisation ratios in buildings are often below 0.80" across materials. Every fraction of unused capacity is steel, concrete or timber — and embodied carbon — that was mined, made, transported and erected for nothing. The MEICON survey of practising structural engineers, run in 2017 (129 responses) and repeated with identical questions in 2022–23 (108 responses), shows why the waste persists even as attitudes change: material efficiency has jumped from a priority for 21% of respondents to 38% (tying with cost as the top design criterion), yet when asked to size a beam for a 200 kNm moment, 59% still chose a resistance above the requirement, and asked how often a utilisation of 1.00 appears in their final designs, 66% gave low-frequency answers ("never" or nearly so). The report's verdict: "despite these advances, overdesign still prevails in practice."

Why this matters

The survey report puts the construction and buildings sectors at "approximately 37% of global CO₂ emissions and ... 34% of global energy consumption" (citing UNEP's Global Status Report) and construction-related spending at 13% of global GDP; structural frames are typically the largest share of a building's embodied carbon. Halving the effect–resistance gap in steel frames alone would cut structural steel demand by a large fraction at zero performance cost — the report describes it as "significant potential for saving material and accordingly reducing emissions and energy in construction." Unlike new low-carbon materials, this saving needs no new product, no new supply chain and no code change to be legal: it needs the design and delivery process to stop rewarding spare capacity.

What’s been tried and why it hasn’t worked

Guidance exists: IStructE's climate-emergency series (2020) urged engineers to "maximise utilisation," and firms such as Buro Happold issued internal minimum-utilisation targets that tighten as a project matures, backed by colour-coded utilisation in BIM models and asks that bids allow time for optimisation. The 2022–23 survey shows the culture has shifted (19% more respondents report clients or design teams requiring embodied-energy minimisation; 20% more accept revising member sizes after concept design; construction-error worries fell 17 points) — but the anticipated utilisation ratio "is only slightly higher, without suggesting any major changes compared to the design practice in 2017." The reasons respondents give locate the bottleneck in process economics, not knowledge: 33% overdesign to absorb "possible changes in the span, loading, or layout before construction," 16% "like to build in a bit of spare capacity just in case," 11% cite possible change of use, 9% cite standard section sizes, 7% simply feel uncomfortable at Ed = Rd; the report links this to a "linear" design process in which late changes are expensive, "coupled with low fees" that make an adaptable, over-strong design the prudent choice. Rationalisation compounds it: asked how many beam calculations they would run for a repetitive floor plate, the median answer was 10, but the median number of section depths they expect in the as-built structure was 6 — optimisation "may be lost in the building process, where constructability limitations and contractor issues become relevant." A code-level fix has been floated — an upper limit on design resistance (Ed ≤ Rd ≤ βEd) — and respondents split almost evenly (45% low scores, 49% high scores), with 11% warning that a cap would simply become the new target and others fearing unforeseen failures; asked what minimum average utilisation a code should require, the median answer was 0.80.

What would unlock progress

Progress needs the spare capacity to become visible and priced. Three levers emerge from the survey: (1) automation that makes late redesign cheap, so engineers no longer need to pre-buy flexibility with tonnage (the report's IQ12–IQ13: "How might automated procedures be developed to update designs to account for changes in span, loading, and layout before construction?"); (2) contractual and fee structures that compensate optimisation time and hold utilisation targets through fabrication, so the design intent survives rationalisation; and (3) reporting — 77% of respondents disagreed that material utilisation is normally presented to clients (Q9), so a standard "utilisation statement" alongside the embodied-carbon figure would let clients see what they are paying for. Adjacent precedents: aerospace and automotive structural optimisation, where mass is priced explicitly and design-fabrication loops are automated; and energy-performance disclosure in buildings, which moved behaviour once a number had to be shown.

Entry points for student teams

A team could build a utilisation-audit tool that reads a structural BIM or analysis model and reports the distribution of utilisation ratios, the material attributable to rationalisation versus loading uncertainty, and the embodied carbon of the gap — then test it on open or partner models. A design-optimisation team could prototype the automated re-sizing loop the survey asks for: given a late change in span or load, regenerate member sizes and fabrication data within a rationalised catalogue and measure how close to Ed = Rd it gets and how many section depths it needs. A policy/economics team could model the β-cap versus minimum-average-utilisation proposals on a set of real frames and estimate material savings and gaming risks. Relevant skills: structural engineering, computational design/BIM scripting, optimisation, construction economics.