1.Introduction

Gypseous soils can demonstrate a continuous compression and even collapse under external loads. This phenomenon is usually triggered by gypsum softening, dissolution, and leaching due to certain human activities involving water percolation through soil mass and groundwater movement [1]. It has been well established that hydraulic conductivity is a key material property to be controlled in order to reduce the collapse severity of the gypseous soils. A numerous improvement methods have been suggested, by which permeability of the gypseous soil was either directly or indirectly reduced. According to [2], treating a soil containing 40–50 % gypsum content with kerosene can decrease permeability and compressibility by restricting the removal of gypsum from the soil mass. According to [3], adding 4 % automobile oil reduced the permeability coefficient by at least ten times. Al-Sharrad [4] studied the effect of cutback MC-60 on the permeability and compressibility characteristics of Ramadi sandy gypseous soil. Gypseous soils treated with the aforementioned materials have more likely exhibited a reduction in the coefficient of permeability, compressibility, and collapsibility. However, these materials are not considered sustainable due to their harmful effects on the environment and groundwater. In addition, most of these materials bear a density less than that of water and are therefore susceptible to washing by water movement within the soil mass, therefore affecting treatment durability.

The microbial-induced calcite precipitation (MICP) is an up-and-coming technology for soil improvement. This technique involves using urease-producing bacteria mixed with cementation solution, and injected or blended with the soil. In geotechnical engineering, there has been a growing interest in this technology for soil improvement in recent years [5]. The MICP technology utilizes specific bacterial strains found in nature capable of depositing calcium carbonate to repair and fill cracks in concrete and rock materials. Additionally, the technology can help prevent leakage, soil erosion, slope failure, liquefaction of sandy soil [6–9].

The urea can be decomposed into CO32−, HCO3−, and NH4+ by urease secreted from the cell. When Ca2+ binds to CO32−, a large number of calcium carbonate crystals can be formed on the cell surface as given by Eqs. (1) to (3) [10]:

                                                              (1)

                                           (2)

                                                 (3)

The precipitated crystals in bio-cemented soil can have two primary functions; bonding at particle contacts and coating on particle surface [11]. Common microorganisms used in MICP for mineralization include: sulphate-reducing bacteria, denitrifying bacteria, oxidizing bacteria, and urease-producing bacteria [12]. Sulphate-reducing bacteria are primarily anaerobic heterotrophs that do not require oxygen for their metabolic activity [13]. They can reduce sulphates to sulphides while oxidizing organic carbon [14]. This process is represented by Eqs. (4) and (5).

                             (4)

                            (5)

Various bacterial strains have been used successfully for calcite precipitation purposes. Bacillus Pasteurii has been the most used bacterial strain due to its superior characteristics such as its resistance to natural conditions. For instance, they are capable of producing spores, which enable them to survive extreme conditions such as high temperatures, exposure to hydrocarbon pollutants, and extended periods of drought [15]. Urease-producing bacteria are particularly useful in geotechnical engineering due to their low cost and high efficiency in providing cementation. Previous studies on MICP have shown that it has a bio-clogging effect, which reduces the permeability of treated soil samples by causing precipitation of calcium carbonate in the soil pore space [16, 17], and increases the material’s strength and stiffness [18–20]. The current study investigates the role of the MICP treatment in reducing hydraulic conductivity and increasing compressive strength of Ramadi sandy gypseous soil. This soil is typically characterized with profound deformable nature under soaking and water percolation conditions, which has resulted in severe serviceability and stability issues in the superstructures.

2.Materials

2.1.Soil Sample

The soil used in this study was collected at a depth of 0.5 m from a site within the main campus of the University of Anbar. The study area is characterized by high contents of secondary-origin gypsum, ranging from 35 % at the ground surface to about 5 % few meters below the surface [21]. Signs of gypsum dissolution and movement, caused by various hydrological and environmental conditions. The index properties of the soil are given in Table 1.

Table 1. Index properties of the soil.

Soil property

Value

Standard

Gravel (%)

2.3

ASTM D422 [21]

Sand (%)

91.7

Fines (%)

6.0

Coefficient of uniformity (Cu)

5.5

Coefficient of curvature (Cc)

0.77

Liquid limit (%)

Non-plastic

ASTM D4318 [22]

Plastic limit (%)

Non-plastic

Specific gravity

2.4

ASTM D854 [23]

Soil classification (USCS)

Poorly graded sand with silt (SP-SM)

ASTM D2487 [24]

Maximum dry unit weight (kN/m3)

15.3

ASTM D698 [25]

Optimum moisture content (%)

9.5

2.1.1. Bacterial Suspension

The Bacillus Pasteurii PTCC 1645 strain was initially preserved in glycerol at a temperature of 4 °C. To prepare the liquid media was prepared by autoclaving 1 L (pH 9.0) tris buffer, 10 g (NH4)2SO4, 20 g yeast extract at 121 °C, and a 15 Ib/in2 pressure before use. A single colony from the surface of the agar plate with the highest concentration was cultured in 100 ml of this solution. Afterward, the flask was shaken in a shaking incubator at 180 rpm and 30 °C for 48 hours. Then, the bacterial suspension was prepared at a ratio of 1:100.

2.1.2. Cementation Solution

The cementation solution was prepared by mixing equal proportions of 0.25, 0.5, and 1 M of calcium chloride (CaCl2) and urea (CO(NH2)2), as recommended by [27, 28]. A summary of the chemical quantities per liter of cementation solution is given in Table 2.

Table 2. Chemical substances used for cementation solution preparation with a constant urea-Ca2+ molar ratio 1:1.

Chemical materials

Concentration of cementation solution (g/L)

0.25 M

0.5 M

1M

NH4Cl

10

10

10

Nutrient broth

3

3

3

NaHCO3

2.12

2.12

2.12

Urea

15

30

60

CaCl2.2H2O

36.8

73.5

147

2.2.Biochemical Tests

2.2.1. Urease enzyme activity and density of bacterial suspension

To calculate the activity of the urease enzyme, following the same method as [29]. By addition 1 ml of bacterial suspension to 9 ml of 1.11 M urea and measured the electrical conductivity over a 5-minute period at room temperature in terms of mS/min. A bacterial activity of 1 mS/min (electrical conductivity rate) is correlated to the hydrolysis of 11 mM urea during 1 minute [30] .The urease activity was found to be 38 mM urea/min, considered high according to many researchers, e.g. [31]. The activity of the enzyme ranged from 19 to 42.6 mM urea/min under different conditions. The concentration of the bacterial suspension was measured with a spectrophotometer type EMC-11-UV at 600 nm wavelength, OD600. Fresh media was used as a reference. OD600 was 2.55 for bacterial solution.

2.2.2. Calcite carbonate content test

The calcite carbonate content (CCC) of soil samples was obtained using the volumetric calcimeter method outlined in [32]. This method is based on measuring the volume of carbon dioxide gas emitted from the reaction of calcium carbonate with hydrochloric acid (ASTM D 4373 – 02) [33], as given below.

                                                  (6)

A volumetric calcimeter was developed and used in this study to measure the volume of the released CO2, as shown in Fig. 1. The system consisted of a reaction flask (500 ml) with a stopper, burette (50 ml), leveling bulb, and connection nylon tubes. Distilled water was leveled between the burette and the bulb in the apparatus portion. A given mass of CaCO3 produces a certain volume of CO2. Pure CaCO3 powder was used to develop the calibration curve correlating the released CO2 volume to the mass of CaCO3. In more detail, 0.1, 0.2, and 0.3 of calcium carbonate (99 % purity) were placed carefully into a 10 ml beaker. The beaker was lowered carefully with a squeezer inside the reaction flask containing 50 ml of 5 % HCL without letting the two substances mix. Then, the system was closed under atmospheric pressure conditions, and the reaction flask was tilted so that the substances were mixed. Meanwhile, the flask was gently shaken. The excess pressure generated by the released CO2 caused the water in the burette to move towards the bulb. The reading was taken once no water movement was detected inside the burette [34]. To determine the CCC of soil samples, a sufficient mass of each oven-dried specimen was sampled from the top, middle, and bottom. These materials were pulverized and sieved on a No. 40 sieve (0.425 mm), and then 2 g were collected from each sample for the test. The test for the pure calcium carbonate was similarly performed on these samples. The released volume of CO2 during each test was used to determine the equivalent mass of the CaCO3 from the calibration curve. The CCC, in percentage, was calculated as:

                                                              (7)

where  is the equivalent mass of the calcium carbonate, and  is the mass of the soil sample.

a

b

Figure 1. Volumetric calcimeter: a) device image; b) schematic of the device.

2.3.Geotechnical Tests

2.3.1. Permeability-leaching Test

2.3.1.1. Specimen preparation

As a first step, the soil was dried at a temperature of 35 °C to preserve the natural gypsum composition of the soil. Next, the soil was sieved on the No. 4 (4.75 mm) sieve, where only the portion passing that sieve was used for the lab work. A 15 ml preprepared bacterial suspension was mixed with 60 g of the soil sample, making 25 % by weight of the soil sample. This allowed the bacterial suspension to cover soil particles effectively. The mixture was molded gently inside the oedometer plastic ring. The specimen was left on the bench for an hour to ensure the bacteria adhered to the soil particles. Meanwhile, a container with 3 liters of cementation solution was prepared for each of the six specimens. Each container was equipped with an air pump to allow for the circulation of dissolved air and provide an environment favorable for calcite production and precipitation. To prevent fine soil particles from smearing, two layers of fine filter were placed on the top and bottom faces of the specimens. These specimens were finally immersed in the cementation solution for 7, 14, or 21 days. The readings of the pH, EC, T, and OD (oxygen demand) of the cementation solution were recorded daily during the curing stage. Two specimens were prepared for each treatment period; one was for the permeability test, and the other was for the calcite test. The specimens were placed inside an oven for drying at 35 °C for two days. The drying stage was performed to create a reference moisture for all the specimens and to simulate the field conditions in terms of temperature. In addition, drying helps mitigate the bacterial activity, so that the calcite content of the specimens is taken by the end of the curing.

2.3.2. Test Procedure

Permeability-leaching tests were performed using the equipment shown in Fig. 2, which were initially arranged by [35]. The equipment comprised of a conventional oedometer with a modified cell, and a hydraulic system. The cell was provided with inlet and outlet to facilitate the salt leaching process. The function of hydraulic system was to provide distilled water for the leaching stage and to allow for leachate collection and measurement of the coefficient of permeability. This system consisted of controlled-head water storage tanks, PTFE tubes, valves, and a graduated flask. The coefficient of permeability of the material was measured by opening the inlet and outlet valves. The leaching was performed under hydraulic gradient values between 10 and 32. The time required to collect a given volume of outflowing leachate under a constant hydraulic gradient was recorded. The cell was initially mounted in its place and the specimen was incrementally loaded to 100 kPa vertical stress under its after-curing state. Subsequently, the specimen was soaked overnight with distilled water. The variation in specimen height was frequently recorded at various test stages. The permeability-leaching test was commenced through a downward flow by opening a valve installed on the line connecting the cell to the water tank. The time to collect each 500 ml of leachate volume was calculated. The coefficient of permeability,  was obtained under constant head conditions by using Darcy’s law as:

                                                                               (8)

where  is the discharge velocity, and  is the hydraulic gradient.

a

b

Figure 2: a) Schematic plot of the hydraulic system; b) schematic plot of the modified oedometer cell.

2.4.Unconfined Compression Test

2.4.1. Specimen preparation

This test investigated the MICP treatment’s effect on the soil’s stiffness and strength. Test specimens were prepared inside plastic PVC molds with 45 mm in diameter and 90 mm in height. The specimen was prepared by mixing method. The process started by drying a soil sample weighing 180 g at 35 °C. The soil was then placed in a refrigerator to cool it down to 4 °C. Once cooled, the soil was mixed thoroughly with 27 ml (15 %) of a bacterial suspension at a temperature of 4 °C. This was mandatory to prevent premature calcium carbonate production. Afterward, the cementation solution was added at a 1:1 ratio to the bacterial solution; three specimens were prepared with cementation solution concentrations of 0.25, 0.5, and 1 M at a temperature of 4 °C. Finally, the mixture was poured into the PVC molds in three stages and then subjected to gentle vibration to ensure the mold was filled uniformly. The molds containing the specimens were subsequently placed in an oven for curing at 30 °C for seven days before being examined.

2.4.2. Test Procedure

The test was conducted following the guidelines of ASTM D2166 [36]. Before the test, the weight and dimensions of the specimen were measured, and both the top and bottom faces were leveled. The test equipment consisted of a displacement-controlled load frame with a 5 kN load cell and a 0.01 mm precision digital dial gauge to measure axial displacement. The specimen was mounted on the equipment’s base plate and loaded at a 1 mm/min displacement rate until the specimen experienced shear failure.

3.Resultes and Discussion

Chemical Properties

3.1.1. Calcite carbonate content of specimens after treatment

Fig. 3 shows the results of the CCC test on soil specimens treated by Bacillus Pasteurii strain with 0.25, 0.5, or 1 M cementation solution molarities and cured for 7, 14, or 21 days. The untreated specimen yielded about 1 % CCC, which is expected for many gypseous soils. In general, the CCC increased with increasing curing time. After 7 days of curing, the amount of calcite produced by the bacterium was between 3 % (specimen with 1 M) and 7 % (specimen with 0.5 M). After 21 days of curing, the results of the specimens with 0.5 M demonstrated the highest CCC of 15.4 5. This amount is clearly higher than those of the specimens with 0.25 and 1 M cementation solution. This suggests that the 0.5 M cementation solution molarity provides the optimal medium for calcium carbonate precipitation for the gypseous soil under investigation. This agrees with previous studies which were stated that concentrations of cementation solution lower than optimal can lead to poor cementation and insufficient CaCO3 production. On the other hand, higher concentrations than optimal have adverse effects on the cementation quality and quantity. One possible explanation is that high concentrations can inhibit bacterial urease activity [37].

Figure 3. Calcite carbonate content of the treated specimens at various curing ages.

3.1.2. Sulfur content

The energy dispersive X-ray (EDX) analysis was performed on both the untreated and treated samples with different cementation solution concentrations but at the same age, as shown in Fig. 4. It was discovered that the weight ratio of sulfur (S) to the samples with the highest calcium carbonate concentration, i.e., that prepared with 0.5 M cementation solution, was less by 90 % compared to the untreated sample. This reduction was due to the reduction of sulfate through the MICP technique.

(a)

(b)

(c)

(d)

Figure 4. EDX analysis of: a) untreated sample; b) treated sample with 0.25 M; c) treated sample with 0.5 M; d) treated sample with 1 M.

Geotechnical Properties

3.2.1. Hydraulic сonductivity

The permeability of both treated and untreated samples was assessed by measuring the saturated hydraulic conductivity (k) corrected to a standard temperature of 20 °C. The specimens were cured for 7, 14, and 21 days. Fig. 5 showed a typical variation of the permeability coefficient with leachate volume. The coefficient of permeability of the untreated specimen increased almost sevenfold with the progress of leaching. This can be attributed to gypsum leaching, which appeared to cause an enlargement in the voids and water passages through the specimen. The MICP-treated specimens (e.g., with 0.25 M) showed slightly lower  values at the beginning, then marginally increased with the progress of leaching. This performance is associated with the fact that the produced calcium carbonate provides cementation and coating to soil grains and gypsum, so water flow has little effect on salt leaching. For an unbiased comparison, the permeability of the materials was obtained and compared after 1500 ml of leachate volume, as shown in Fig. 6. The results showed that the hydraulic conductivity of the treated samples decreased with increasing curing periods. For the specimens with 1 M, the dissolution of gypsum in the soaking solution may have caused an adverse effect on permeability by increasing the size of the voids. For the specimens with 0.5 M, after 14 and 21 days of treatment, the presence of approximately 15 % CCC caused the voids to become smaller, resulting in a substantial decrease in permeability.

a)

b)

c)

Figure 5. Variation of the coefficient of permeability with leachate volume of samples cured to: a) 7 days, b) 14 days, and c) 21 days.

Figure 6. Values of the coefficient of permeability after 1500 ml of leachate volume.

The effect of CCC on the permeability (measured after 1500 ml of leaching) was inspected in Fig. 7. The results reveal that the higher the CCC, the smaller the permeability. This further reassures the role of the MICP technique in reducing gypseous soil hydraulic conductivity and thereby reducing leaching effects.

Figure 7. Variation of the permeability measured at a leachate volume of 1500 ml with CCC.

3.2.2. Unconfined compression

Fig. 8 shows the effect of soil stabilization with Bacillus Pasteurii and different concentrations of cementation solutions (i.e., 0.25, 0.5, and 1 M) on the unconfined compression stiffness and strength. The elastic modulus was taken as the slope of the steepest portion of the stress-strain curve. A summary of these results is presented in Table 3. The results indicate relatively high unconfined compressive strength values compared to the weakly cemented natural sandy gypseous soil, thanks to the MICP treatment.

Figure 8. Unconfined compression test results of several treated specimens.

Table 3. A summary of the unconfined compression test results.

Test code

CCC (%)

Compressive strength, qu (kPa)

Axial strain at peak stress (%)

Elastic modulus, Es (MPa)

p0.25M

6.6

591

2.5

19.2

p0.5M

7.3

480

1.9

27.9

p1M

5.6

566

2.2

39.5

Microstructural and Mineralogical Analysis

3.3.1. Scanning electron microscopy

The spatial distribution of the calcium carbonate crystals inside three specimens with Bacillus Pasteurii, treated and 0.25, 0.5, and 1 M cementation solution for 21 days, is presented in Fig. 9. In the samples with 0.25 and 1 M, which exhibited 4 and 5 % CCC, sand particles were barely coated with calcium carbonate. On the other hand, in the sample with 0.5 M, which had nearly 15 % CCC, the voids were largely filled with calcium carbonate.

a

b

c

Figure 9. SEM images with calcium carbonate distribution on samples: a) 0.25M; b) 0.5M; c) 1M.

4.Conclusion

The cementation solution concentration plays an essential role in the amount of precipitated calcium carbonate and in the level of improvement. The concentration of 0.5 M seemed to outperform other concentrations in reducing permeability of the gypseous soil. In general, the calcium carbonate content increased with increasing curing time. After 21 days of curing, the results of the specimens with 0.5 M demonstrated the highest calcium carbonate content of 15.4 %. The results showed that the hydraulic conductivity of the treated samples tended to decrease with increasing curing periods. The permeability decreased nearly twice with a curing time increase from 7 to 21 days. The results reveal that the higher the calcium carbonate content, the smaller will be the permeability. The permeability decreased by almost fourfold, with calcium carbonate content increasing by 15 %. The MICP treatment was able to provide sound protection against gypsum dissolution and piping. The results of the leaching stages indicated that the treated specimens tended to maintain the same permeability with leaching, compared to the untreated ones, which exhibited many times increase in permeability due to gypsum dissolution and leaching. With MICP treatment, the permeability became progressively smaller, and water discharge was more and more restricted. Consequently, the treated soil became less susceptible to leaching effects. One of the manifestations of the MICP treatment was that the treated soil demonstrated an appreciable unconfined compression strength of up to 600 kPa.