Effect of Ureolytic Bacteria on Compressibility of the Soils with Variable Gradation

To conduct the preliminary research on the influence of bacteria on the compressibility parameter of the soil, three types of soils were used: non-cohesive soil, cohesive soil, and anthropogenic soil which was Class F [ASTM C618-19] fly ash [40], the coal combustion by-product. The chemical and phase composition of tested fly ash is presented in Table 1. In Fig. 1 the grain size distribution curves of tested materials are presented, according to the EN 933-1 standard. The physical parameters of presented materials, like the density of solid particles (ρ_s) and median particle diameter (D₅₀), are given in Table 2.

Figure 1.

Based on Fig. 1, in accordance with EN ISO 14688-1 standard, the tested non-cohesive soil is fine sand (FSa), and the tested cohesive soil is silt (Si). Class F fly ash (FA) has a grain size corresponding to that of sandy silt (saSi), and in the literature, fly ash is called transitional soil [41].

In this paper, strain Sporosarcina pasteurii (DSM 33) was selected as a commonly used ureolytic bacteria in the MICP process [18, 20, 23, 42]. Sporosarcina pasteurii is a soil bacterium known as a strain which produces high levels of urease and tolerates extreme conditions [10]. The bacteria from the freeze-dried form was rehydrated and moved to solid medium nutrient agar with urea. After incubation at 25°C for 48 h, the bacterial colonies were formed. Activated bacterial colonies were moved to the flask which contained 50 mL of urea medium. The urea medium for bacteria, urea medium and cementation solution were prepared according to the procedure described in [33]. Finally, two flasks were prepared, one containing urea medium with bacteria, and the second one, cementation treatment consisting of urea medium and CaCl₂ stock solution (with a concentration of 0.5 M of Ca). Urea [CO(NH₂)₂] was used as ammonium and an energy source for the hydrolysis process [43]. The bacterial suspension was analysed for optical density measured at a wavelength of 600 nm (OD₆₀₀) using an ultraviolet-visible spectrophotometer. OD₆₀₀ corresponds to bacterial concentration, which might be used to describe the activity of bacteria. The bacterial concentration was calculated according to [14]: (3) Bacterial density (cells/ml)=OD600⋅4⋅108 Bacterial\,density\,\left( {cells/ml} \right) = O{D_{600}} \cdot 4 \cdot {10^8}

The bacteria concentration used in the presented research was OD₆₀₀ = 0.3, which corresponds to the bacterial density of 1.2 × 10⁸ cells/mL. In Fig. 2, the bacteria-urea solution is presented, the same bacteria-urea solution directly after the cementation solution is added and after a few days of storing at room temperature (21°C), where the precipitated particles are visible.

Figure 2.

As reported in the literature [44], when the value of OD₆₀₀ is higher than 0.15, the bacterial suspension is suitable for the biocementation process. In the research of Wang et al. [14], the bacterial solution had an optical density from 0.2 to 3.0. Other factors of the experiment were constant. The higher bacterial concentrations lead to higher activity, so faster and more crystals should precipitate (depending on the cementation solution concentration). As [14] observed in the microscale, OD₆₀₀ had an impact on the sizes of the precipitated crystals. In the case of the lowest OD₆₀₀ value, crystals formed slower and there was a lower amount of them, but the largest among the other cases. When the bacterial density was high, CaCO₃ crystals were small and unstable, but over time might transform into larger and more stable forms. The worst results were achieved for the very high bacterial density (OD₆₀₀=3.0), where the rate of CaCO₃ precipitation increased, but formed large amounts of unstable forms of CaCO₃. It might cause problems in the case of soils with small pores, where unstable forms might be trapped in pores and cause clogging and inhomogeneity of improved soil.

Premixing was chosen as a method of introducing the solutions to the soil because the cohesive soil (Si) was used and the fly ash had the granulation of saSi. As the literature shows [5], percolation is a mainly used injection method. However, there might be a problem with the homogeneity of the biocementation process through the sample. Premixing resulted in a homogeneous sample [45]. Although the homogeneity of the samples was achieved, the results of the tests of unconfined compressive strength (UCS) revealed lower values of UCS that the one for the biocemented samples where the injection method was used. The uniformity of biocemented samples is important, because of the repeatability of the test results, and as Konstantinou et al. [46] claimed, the literature shows contradictory results.

In the literature [12, 13, 19, 46, 47], in most cases, sandy soils are used to conduct the MICP process, where the bacterial and cementation solution with the use of a peristaltic pump or other equipment, can easily flow through the soil particles. Mitchell & Santamarina [41] prepared a graph showing the bacteria size and types of soils, where the process of biomineralization may be a misguided soil improvement method. However, since then, few studies [48, 49], have tested the effect of MICP on fine-grained materials. In this study, sandy soil, silt, and fly ash were tested.

Before sample preparation, each soil was put into a laboratory dryer at a temperature of 100°C to get dry material and to be sure, that no other microorganisms are present. In order to prepare samples, the urea-bacterial solution was mixed with urea-cementation solution in the proportion 1:1 and directly added to the soil, followed by thorough mixing. The volume of the mixture that was added to the material was calculated according to the initial void ratio. The samples were to be placed in a mould (oedometer ring), so the mass of the soil was calculated and weighted before moulding. A different amount of mixture was added to each material as different moisture contents were desired.

Samples were prepared in a rigid ring for an oedometric test of 65 mm in diameter and 19.5 mm in height. On the bottom and top of the samples filter papers and porous stones were placed. Then, the samples were left at room temperature (protected from drying) to enable the biocementation process, which is connected to the retention of ureolytic bacteria in the soil. The simultaneous introduction of bacteria and calcium into the soil was the procedure followed by [35]. It was a good solution in a method where no peristaltic pump was used.

In the initial process of biocementation, microbes are attaching to the soil particles. Samples were kept in conditions of constant temperature (21°C) and protected from drying. Curing time, from the preparation of the sample to the tests lasted a total of 14 days, after which the compressibility tests were conducted in an oedometer. The curing time was based on the literature [50], where the researchers found that bacterial activity decreased after 16 days of treatment.

The loading path of the oedometer test was 8, 15, 30, 60, 120, 240, 540 and 725 kPa. The unloading was 240, 90 and 30 kPa. Each load increment was applied to the sample for 24 hours. Before placing samples in the oedometer, the weight of each was checked. The samples were flooded with water. The vertical deformations during the test were collected by the computer program.