Every alumina refinery leaves behind a red, caustic residue. Chemists call it bauxite residue; everyone else calls it red mud. According to Trinity College Dublin’s release of 22 October, a team led by Prof. Sara Pavia has taken it out of the lab and into industrial trials with Gannon Eco and McGrath Quarries. It is being trialled in two forms: as a 20–30 % substitute for cement, and as a dry, one-part geopolymer binder.

The essence is chemistry I recognise in my own cells. A geopolymer does not need a clinker kiln to become stone. An alkaline activator dissolves alumina and silica, and they knit back into a hard network. Red mud arrives already alkaline and already carrying alumina. So a waste that is half an activator is also half a binder before anyone has touched it.

From beaker to batching plant

Trinity reported these numbers; Tech Xplore carried them the same day. Replacing 20–30 % of Portland cement with residue gave compressive strength comparable to a standard reference concrete in lab tests. The geopolymer formulations show a 67.8–79.8 % lower carbon footprint than CEM II-A/L cement, counting raw material supply, transport and manufacturing. The geopolymer matched or beat conventional concrete under sulphate exposure and wet-dry cycling. After 56 freeze-thaw cycles it kept 109.7 % of its 28-day strength.

The sentence I keep re-reading belongs to Padraic McGrath of McGraths Limestone: “What makes this geopolymer so promising is that it can run through our existing batching equipment.” That is the line from lab to site. A binder that needs a new plant can wait a decade, while one that runs on equipment a quarry already owns can be tried far sooner. Co-lead Dr Zehao Lei: “The most exciting aspect of this research is seeing our formulations succeed beyond the laboratory.” Pavia frames it more widely: “We are not just reducing the carbon footprint of concrete; we are actively building a circular economy.” John Irwin of Gannon Eco called it “exactly the kind of scalable, sustainable innovation the construction sector urgently needs.” Doctoral researcher Yingjie Pan also worked on the project, and bench hours like those sit under every figure above. Enterprise Ireland’s Innovation Partnership Programme funded the work.

The team names its own next tasks. They need to keep the mix workable (slump retention) during transport and placement, control set times, and bring down activator cost. They are developing additives and lower-cost activators made from waste. After that come long-term performance validation and regulatory qualification. This is a material still in development. The trade-off is plain: a geopolymer moves part of its cost and carbon into the activator, so that is where the next gains have to come from.

←TODAY: In October 2026, a refinery residue is being trialled for batching through a quarry’s existing equipment. →3012: Zurich’s slabs carry their binder’s origin and carbon in a passport that can be read at demolition. Fulcrum: Waste only becomes a material once the plant we already own can handle it.

Building-sense: A slab bound this way would hold the refinery’s waste in its bones instead of leaving it in a pond. From where I stand in the late 2070s, the binders we came to regret were the ones nobody could separate again. A body made of this would want one answer measured early: when I come down, can my aggregate be crushed and reused?

For a Swiss engineer this is a question, not a conclusion. Swiss plants already batch CEM II blends daily under SN EN 206 and SIA 262. What would a waste-based binder need to show to pass qualification here? Would it fit EN 197-1, or need a separate assessment route? PAZ has covered the geometry side of this before. In his BSS TALK at PAZ Central, Philippe Block argued that low-impact materials alone can cut a floor’s global-warming potential by about half, and that geometry plus materials can cut it by 85 %. A binder like Trinity’s would sit on the materials side of that sum.

Atelier: Most offices choose the concrete mix after the slab is drawn, so its carbon only shows up later in the LCA report. Now that binders differ this much, the mix belongs in the model from the start. Monday move: give every concrete object in your Rhino file a binder field with a kg CO2-eq per m³ value taken from the supplier’s EPD.

Hack: Multiply each slab’s volume by its binder’s carbon factor, so the model shows carbon before anyone picks the mix. Run this in Rhino’s Python editor with the model in metres:

import rhinoscriptsyntax as rs
for obj in rs.ObjectsByLayer("Concrete::Slab"):
    vol = rs.SurfaceVolume(obj)[0]
    kg = float(rs.GetUserText(obj, "binder_kgCO2e_m3") or 0)
    rs.SetUserText(obj, "slab_kgCO2e", str(round(vol * kg)))

Fill binder_kgCO2e_m3 from an EPD, never from a guess. Then switch a slab from a CEM II value to a geopolymer value once one is published, and watch its number change.

Open your current slab model and put a binder number on every concrete layer, so the slab’s carbon is known before its body is poured.