Mitigation of liquefaction of sand using microbial methods.
Date of Issue2013
School of Civil and Environmental Engineering
Many existing liquefaction mitigation methods such as compaction and cement mixing are usually too expensive to be applied to a large area. An attempt to develop a more cost-effective method for mitigation of soil liquefaction, the so-called biogas desaturation method, was made in this study. The liquefaction potential of fully saturated sand can be greatly reduced when the sand is made slightly unsaturated by the inclusion of gas. The microbial denitrification process, which is able to produce nitrogen (N2) gas, was selected for the desaturation of liquefiable sand. N2 gas is suitable for this purpose because it is inert and its solubility in water is low. A preliminary study on the biogas desaturation process showed that the degree of saturation of initially fully saturated sand could be effectively reduced to as low as around 80% in 4 days. Triaxial undrained compression and extension tests were carried out on loose and medium dense sands. For loose sand, the undrained shear strength showed a significant improvement when Skempton’s pore water pressure coefficient B-value reduced from around 0.95 (fully saturated) to around 0.3 and other conditions were the same. There was also a change in the stress-strain behavior from complete strain softening to strain hardening with the decrease in the degree of saturation. For medium dense sand, reduction in the degree of saturation led to an increase in the yielding stress and a decrease in the maximum pore water pressure. A fully instrumented laminar box model with a shaking table was adopted to conduct model tests to verify the effectiveness of the biogas method under cyclic load. Liquefaction occurred for saturated samples at loose states under amax=0.5 m/s2 and at medium dense states under amax =1.5 m/s2. Liquefaction did not occur for biogas desaturated samples with the degree of saturation of 80% to 95% at the same conditions. Both a photographic system and a computer tomography (CT) system were used for the observation of gas distributions in sand. It was observed that gas bubbles were either in the single bubble form or in the form of gas pocket with small gas bubbles aggregating together. A sand column test was conducted to study the stability of gas bubbles in sand under both hydrostatic and flow conditions. Under the hydrostatic condition, there was almost no change in the degree of saturation. When water flows under a hydraulic gradient of 0.1 through sand, it took 3 to 5 days for desaturated sands with the degree of saturation of 89% to gradually become fully saturated. This implies that the stability of the gas bubbles is affected by the groundwater flow.