Geotechnical Enhancement and Carbon Sequestration in Expansive Soils using Ground Granulated Blast Slag
Year: 2025
Type: article
Abstract: Expansive soils, characterised by high plasticity and susceptibility to moisture-induced volume changes, pose major challenges to infrastructure performance. This study investigates the use of Ground Granulated Blast furnace Slag (GGBS) as a soil stabiliser and a medium for CO2 sequestration under accelerated carbonation conditions. Laboratory experiments with GGBS additions of 10%, 20%, and 30% showed marked improvements in soil behaviour. Unconfined Compressive Strength (UCS) increased notably, with 20% GGBS-treated soils initially reaching 2.32 kg/cm², though exposure to CO2 over 24 h led to a gradual strength reduction to 0.69 kg/cm². Plasticity Index (PI) values decreased across all treatments, with the untreated soil dropping from 32 to 24.2, while the 30% GGBS mix showed a sharper initial decline from 22 to 16.29, stabilising at 21.1 after 24 h. CO2 adsorption capacity was highest in the 30% GGBS samples, peaking at 36 mg/g before reducing to 25 mg/g, reflecting substantial carbonation potential. Machine learning models, including Random Forest and XGBoost, were employed to predict UCS, PI, and CO2 uptake. Among them, XGBoost consistently achieved superior predictive accuracy across all variables and exposure conditions. These findings highlight the promising role of GGBS in sustainable soil stabilisation, offering both mechanical improvement and environmental benefits through carbon sequestration. Major Findings: The addition of GGBS significantly improves the geotechnical properties of expansive soils while enabling short-term CO₂ sequestration through accelerated carbonation. A 20% GGBS mix yields the highest initial strength, while 30% GGBS shows maximum CO₂ uptake (36 mg/g) and plasticity reduction. Machine learning models, especially XGBoost, effectively predict UCS, PI, and CO₂ adsorption with high accuracy, reinforcing the experimental trends.
Authors: H. N. Raghavendra, H. Likhitha, N. Muralidhar, T. P. Sanjeev, Arunkumar Yadav
Language: English
Subfield: Civil and Structural Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4415104709