Denitrification-based pretreatment for the compaction of silty sands

Abstract

Recent developments in bio-based soil improvement techniques include the use of the bacterially mediated process of denitrification to alter engineering properties of soils. Pham (2017) used the denitrification process to stimulate the production of gas and precipitation of calcium carbonate as a soil improvement method. The ability to desaturate a soil to its optimum saturation by means of gas production inspired the use of this method to improve the engineering properties of soils more efficiently by means of a two-stage process of desaturation and compaction. Andrag (2017) started a research to investigate the potential of this concept and developed a test setup for the assessment of its efficiency. The prospects of this biogenic soil pretreatment method are further investigated in the current study. The objective of the research is to assess how the compactability of an initially saturated soil is influenced by the in-situ formation of gas bubbles and if this gas can be used to reduce the energy requirements for compaction of a soil to a given target density. A review on existing literature is initially carried out to grasp the latest developments on this topic. Information from literature is used to determine the typical characteristics of soils on which compaction activities are often found to be ineffective. Based on this information, a representative silty sand with a high gas storage capacity is created for further experiments. It is found that gradation is one of the main properties controlling the water content-dry density relationship of soils. The water content-dry density relationship of soils composed from industrially produced material is found to be often significantly different from natural soils, which can be explained by differences in gradation. A testing strategy is developed and a setup is established for assessment of the concept of biogenic soil pretreatment for reducing energy requirements for compaction. Denitrifying bacteria are cultivated with a chemostat setup which can supply bacterial inoculum for subsequent experiments. The effect of the biogenic pretreatment is evaluated by comparing the energy requirements for compression of treated and untreated soil samples from initial conditions to a given target density with a constant strain-rate. Soil samples are successfully desaturated to targeted saturations by means of an engineered treatment regime. Concentrations of the substrates are determined based on the stoichiometry of the biochemical reaction, Henry’s law, Boyle’s law and the ideal gas law. Gas is produced at a steady rate inside the soil samples with a very limited amount of gas escaping, confirming the high storage capacity of the assembled sample material. The determined optimum saturation of the silty sand (80%), is therefore achieved without exceeding the gas percolation threshold of the soil. The gas that escapes the sample during the gas production stage and the loading stage, is captured with a gas trap that is included in the setup. The changes in water content and dry density of the soil during the experiment can be accurately predicted by means of a volume balance. The major part of the initially available nitrate in the pore fluid is consumed at the end of the experiments, with a limited amount of accumulated nitrite. Moreover, the denitrification process is successfully buffered by including calcium within the engineered substrate solution.
It is initially presumed that energy requirements for static compaction of a silty sand in the Rowe cell can be reduced by bringing the soil closer to optimum conditions in terms of saturation, before starting the loading stage. Based on an analysis of the work input per volume, it is found that the energy requirements for compaction increase slightly as a result of the biogenic pretreatment. In terms of compactability, no clear benefit is obtained from the biogenic pretreatment according to the energy-based assessment method that is used in this research. The slightly higher energy requirements for compaction of treated samples can possibly be explained by the formation of biofilm or entrapment of gas bubbles in the drainage lines of the setup. Cementation effects as a result of calcium carbonate precipitation are not expected to play a role, since calculations show that the relative amount of calcium carbonate in the soil is very limited at the end of the experiments. It is recommended to further investigate the potential of the denitrification-based pretreatment for the compaction of soils by exploring new methods of assessment. It is suggested to focus on implementation of a dynamic loading component to the test method and to investigate the options for execution of a field experiment. Representation of conditions that apply in practice is thereby essential. The potential of biogenic pretreatment for improvement of the engineering properties of silts and clays can also be investigated as it might be a sustainable alternative for the relatively scarce and costly coarse-grained fill material that is normally used nowadays.