Green Cement: AI Finds a New Recipe

Cement production is a highly energy intensive process that generates significant planet-heating emissions. Researchers in Switzerland developed an artificial intelligence (AI) model capable of finding new cement recipes that maximize building material quality whilst minimizing associated emissions.

—

Artificial intelligence (AI) researchers have discovered a new cement formula linked to significantly less carbon emissions compared to traditional cement.

Current cement production is linked to some 8% of global carbon dioxide emissions, over three times the amount of the global aviation sector.With demand for cement set to increase by 4.75% between now and 2030 and by another 48% by 2050, the industry has committed to sector-wide annua l4% emissions reduction targets through to 2030 in alignment with net-zero targets by 2050. Finding low-carbon cement alternatives will be crucial to meet these targets.

Cement Production

Cement is mixed with water to form concrete – a building block of infrastructure from subway stations and buildings to sewage systems.

Producing cement involves several stages, including mining, raw material grinding, and heating to extremely high temperatures in a kiln, followed by clinker cooling and grinding. These steps require significant amounts of both thermal and electrical energy. 

New Recipes

One promising strategy to reduce carbon emissions is to change the recipe. A paper published in Springer Nature in June details a new approach to finding new cement recipes, involving an advanced modelling approach based on machine learning. 

Researchers at the Paul Scherrer Institute in Switzerland have created a model driven by AI that speeds up the exploration of novel cement compositions capable of maintaining high material standards while improving their environmental impact by reducing carbon emissions.

Extending the AI model to other factors could provide substantial benefits. For example, AI could help identify a formula that improves the durability and strength of cement, something of increasing importance in places vulnerable to natural disasters.

“Instead of testing thousands of variations in the lab, we can use our model to generate practical recipe suggestions within seconds – it’s like having a digital cookbook for climate-friendly cement,” said mathematician Romana Boiger, lead author of the study. The AI model is 500 times quicker than a traditional physics model, a testament to the increased role AI technology plays in the green transition. 

Other Solutions

Low-carbon alternatives to cement already exist. Limestone calcined clay cement, for example, has an emissions reduction potential of up to 40%.

Algae-grown limestone, specifically the use of microalgae in cement production, offers a promising path towards carbon-negative construction. Microalgae can capture and store CO2 in mineral form, potentially offsetting a significant amount of carbon emissions and even leading to carbon storage within building materials. A building using this technology could potentially sequester up to 250 tons of carbon. Cement can also be recycled by heating it in an electric furnace, and the resulting material can be used as a partial substitute in steel recycling.