Sulfide soils are silty soils, often found in saturated conditions, under the groundwater level. Characteristics of these soils, including particle size distribution and consistency limits along with chemical composition and environmental properties, cause excavation to be necessary for construction purposes. The excavated sulfide soil usually is transported and deposited in landfills. These soils are either deposited in saturated conditions or chemical buffers are added to the soil to prevent acidification. Special conditions of these landfills complicate the disposal procedure and the landfill maintenance which makes those financially expensive. Reusing sulfide soil in construction is a solution to reduce the expenses related to the management of sulfide soils. Since the mechanical properties of these soils are not suitable for construction purposes, the first step is to improve soil characteristics to the level that fulfills the needs of construction applications. One solution to improve the mechanical properties of the soil is adding a binder to the soil.

The main focus of the research was to improve the mechanical properties of soil. The research activities were divided into two parts. The first part was conducted in a laboratory environment to develop mixtures, while the second focused on transferring the results to field conditions. The laboratory tests included mixing soil and binder i.e., cement was added to the soil at different percentages to evaluate the soil improvement. An unconfined compressive strength (UCS) test was conducted on the stabilized sample to evaluate the efficiency of the stabilization. The resultsof UCS for the stabilized samples were compared. Since the soil contains a high amount of water, the traditional sample preparation was not suitable. Therefore, an alternative method was developed and evaluated. Moreover, the effect of curing time on the strength and consistency limit of stabilized samples was evaluated. At last, the effect of different variables, including porosity, binder content and initial water content, on the UCS of soil was investigated to identify potential correlation between UCS and different soil variables.

The results of the tests showed that adding a binder, regardless of the type of sulfide soil, positively affects the UCS of prepared samples and increasing the curing time increased the UCS of the samples. At higher cement content, the effect of curing time was more significant. Also, it was shown that at higher water content, the effect of binder is lower in comparison with the same soil at lower water content. By lowering the water content, the strength of stabilized soil reaches a maximum and drying further the soil, below the optimum water content, led to strength reduction. A correlation between UCS of sample and porosity/binder ratio was employed to predict the strength behavior of stabilized soil based on variables such as porosity, initial water content and binder dosage.

In order to evaluate if laboratory results can be applied to geotechnical applications, the second part of this research included a field mixing experiment for a large-scale mixture of soil and cement. The effect of the mixing procedure with common equipment on the homogeneity of industrial-size mixture was investigated. A sampling strategy for collecting representative samples of mixture was selected and assessed. the number of mixing steps and the effect of binder dosage on the uniformity of samples were studied. Results of UCS of samples prepared from field and laboratory mixture were compared and evaluated. A field evaluation was conducted to determine the quality of the mixture and how many mixing steps are required to reduce variability between samples. Two different percentages of binder were added to the 5 Tons of soil. The UCS test samples were prepared from the soil-cement mixture in the same way as they were prepared in the laboratory and cured for a specific time. The UCS test was conducted on cured samples. The test results were compared to evaluate the mixture homogeneity in the field.

The results showed that homogeneous mixtures can be obtained in the field with the available equipment. Assessing the sampling strategy showed that increasing the sampling sections from 5 to 12 and preparing single UCS sample from the collected soil provides representative samples from the soil mixture pile. Additionally, it was shown that by increasing mixing steps from 2 to 3, it was possible to eliminate samples with notable lower strength than average UCS. A greater number of mixing steps improves homogeneity while reducing the average UCS. It was found that mixing soil and binder in the laboratory improves strength better than mixing them in the field. When applying laboratory results to field design, this point must be taken into account.