top of page
MRPM 2026 - IFSC USP
Vanderley M. John - Confirmed Speaker - MRPM 2026

Porous Media & Sustainability: NMR as a tool to implement IA and reduce CO₂ Emissions in the Cement Value Chain

Cement-based materials are the most widely produced artificial materials, with a global per capita consumption of 4 tonnes per year, accounting for about 30% of all natural resources extracted and 6% of anthropogenic CO₂ emissions. The durability of concrete leads to the accumulation of construction waste, marking the Anthropocene. Concrete is a composite porous rock, initially a concentrated suspension (~20% water), which hardens through hydration, forming new minerals and reducing porosity. As hydration continues, porosity decreases and mechanical strength increases. The rate of strength gain is crucial for construction schedules, while the 28-day strength is vital for structural safety. Typically, cement is part of construction long before 28 days compressive strength results are available. To reduce risks, industry operates with higher safety margins, increased costs, and a larger carbon footprint. AI models have been proposed to improve strength forecasting, but traditional compressive strength testing is slow, labor-intensive, and destructive requiring samples for each age. In consequence, industry keeps testing frequency as low as 1-3 tests a day, which limits AI training data availability, reducing the effectiveness of the models. Industry needs new fast, cheap, digital test methods. Literature shows 1H NMR allows measuring cement paste pore volume with good accuracy in an inexpensive non-destructive test. However, frequently the iron present grey cements – which makes 99% of the market – prevents accurate the results. We succeeded in developing a 1H NMR machine and a new data processing model that can produce reproducible results with a high correlation with compressive strength regardless of the presence of magnetic iron. This solution did work for all (one hundred +) Brazilian commercial cements. We believe this solution have a potential to enable the cement industry to produce at least one order of magnitude larger datasets, enabling generation robust AI models. If industry embraces it, we expect significant CO₂ mitigation at global scale, at no additional cost.

Vanderley M. John

bottom of page