Gypseous soils are highly susceptible to wetting induced collapse because dissolution of gypsum bonds destabilizes their metastable granular fabric. This study evaluates the effectiveness of enzyme induced carbonate precipitation (EICP) in mitigating the collapse behavior of a soil containing 53% gypsum by mass. Nine EICP formulations (A-H and J) were prepared using urea, urease, and, in selected mixtures, CaCl2·2H2O as an external calcium source, whereas the remaining formulations relied on calcium released through gypsum dissolution within the soil matrix. Collapse behavior was assessed using a single oedometer test at σv = 200 kPa under 24 h of soaking, supported by UU triaxial tests conducted under total stress conditions (σ₃ = 200–400 kPa).The collapse potential of untreated soil was 2.211% for 24 h. The collapse potential of untreated soil was reduced by 44.2–78.1% when treated with EICP. The most effective formulation was mix J, followed by mix H and mix B, whereas mixes E and C were the least effective, but still showed some improvement. The time dependent decomposition of the collapse strain showed that the untreated soil had almost equal immediate and progressive components, while the treated mixtures were dominated by the progressive (1–24 h) component, which represented about 72–90% of the total collapse strain. Thus, EICP substantially reduced the magnitude of collapse, but did not eliminate the time dependent wetting response.Calcimeter measurements indicated that the net CaCO3 content ranged from 1.38% to 3.40%. However, collapse improvement did not vary monotonically with carbonate content, indicating that precipitation topology and contact level cementation were more influential than bulk carbonate content alone. The UU triaxial results, support this interpretation: EICP introduced a substantial apparent total stress cohesion component (cUU, app = 3.5–53.2 kPa) while only slightly increasing the apparent total stress friction angle (φUU, app = 40.5°− 42.6°) relative to the untreated soil (cUU, app = 0.0 kPa; φUU, app = 40.4°). Strong positive correlations were observed between measured CaCO3 and both parameters, although the variation in φUU, app remained narrow.Interestingly, mixtures devoid of CaCl2 (E, F, and G) also showed considerable increases in strength and collapse resistance, implying that the Ca2+ released from gypsum dissolution can partially support EICP reactions in soils with high gypsum content. Collapse reduction was generally governed by a combined mechanism involving interparticle carbonate bonding, partial constriction of pore throats, and changes in hydraulic connectivity during wetting.

Enzyme-Induced Carbonate Precipitation for Stabilizing High-Gypsum Soil

期刊论文
发表日期：

2026年7月15日

查看全文

Samer J. Abdullah,
Mohammed Y. Fattah,
Aqeel S. G. Al-Adili

摘要

Gypseous soils are highly susceptible to wetting induced collapse because dissolution of gypsum bonds destabilizes their metastable granular fabric. This study evaluates the effectiveness of enzyme induced carbonate precipitation (EICP) in mitigating the collapse behavior of a soil containing 53% gypsum by mass. Nine EICP formulations (A-H and J) were prepared using urea, urease, and, in selected mixtures, CaCl₂·2H₂O as an external calcium source, whereas the remaining formulations relied on calcium released through gypsum dissolution within the soil matrix. Collapse behavior was assessed using a single oedometer test at σ_v = 200 kPa under 24 h of soaking, supported by UU triaxial tests conducted under total stress conditions (σ₃ = 200–400 kPa).

The collapse potential of untreated soil was 2.211% for 24 h. The collapse potential of untreated soil was reduced by 44.2–78.1% when treated with EICP. The most effective formulation was mix J, followed by mix H and mix B, whereas mixes E and C were the least effective, but still showed some improvement. The time dependent decomposition of the collapse strain showed that the untreated soil had almost equal immediate and progressive components, while the treated mixtures were dominated by the progressive (1–24 h) component, which represented about 72–90% of the total collapse strain. Thus, EICP substantially reduced the magnitude of collapse, but did not eliminate the time dependent wetting response.

Calcimeter measurements indicated that the net CaCO₃ content ranged from 1.38% to 3.40%. However, collapse improvement did not vary monotonically with carbonate content, indicating that precipitation topology and contact level cementation were more influential than bulk carbonate content alone. The UU triaxial results, support this interpretation: EICP introduced a substantial apparent total stress cohesion component (cUU, app = 3.5–53.2 kPa) while only slightly increasing the apparent total stress friction angle (φUU, app = 40.5°− 42.6°) relative to the untreated soil (cUU, app = 0.0 kPa; φUU, app = 40.4°). Strong positive correlations were observed between measured CaCO₃ and both parameters, although the variation in φUU, app remained narrow.

Interestingly, mixtures devoid of CaCl₂ (E, F, and G) also showed considerable increases in strength and collapse resistance, implying that the Ca²⁺ released from gypsum dissolution can partially support EICP reactions in soils with high gypsum content. Collapse reduction was generally governed by a combined mechanism involving interparticle carbonate bonding, partial constriction of pore throats, and changes in hydraulic connectivity during wetting.