Modelling the Environmental Fate of Petroleum Hydrocarbons During Bioremediation

Abstract

This chapter provides the key steps and parameters required for three different numerical modelling approaches to predict the environmental fate of petroleum hydrocarbons in contaminated soils during bioremediation. The first approach is the molecular dynamic simulation which is used to characterise the molecular-scale adsorption, the diffusion and the distribution of the saturate, aromatic, resin and asphaltene (SARA) fractions of oil. Such approach provides insights into the microscopic aggregation, the sequestration and the collision mechanisms which are essential for a better understanding of hydrocarbon bioavailability and biodegradation. The second approach is the use of fugacity modelling to compute the equilibrium distribution of the aliphatic and aromatic hydrocarbons in an environmental matrix composed of four compartments: soil, water, air and nonaqueous phase liquid (NAPL). Further to this, the contribution of the biotic and abiotic processes to the loss of petroleum hydrocarbons including (1) biodegradation in soil and NAPL, (2) advection in air, (3) leaching from soil and (4) diffusion at the soil–air, soil–water and soil–air boundaries can be estimated during biopiling experiments. The third approach is the use of machine learning (ML), an assumption-free data mining method, to predict the changes in the bioavailability of polycyclic aromatic hydrocarbons (PAHs) in contaminated soils. The main advantage of ML models is that they are data-based technique allowing computers to learn and recognise the patterns of the empirical data and work well with highly non-linear systems without relying on prior knowledge on bioremediation processes which make their prediction more realistic than conventional statistical methods. ML outputs can be integrated into microbial degradation models to support decision making for the assessment of bioremediation end points.