ICE 2022


This work aims to validate a novel approach for secondary hydrocarbon (HC) migration by applying agent-based modelling (ABM) to a data rich petroleum system. The method allows modelling HC flow from source rock to reservoir dynamically. Agglomerations of HC molecules are modelled by independent but interacting agents (particles) whose movement is subject to simple rules that reflect fluid behaviour within a geological environment. These rules are based on the main drivers of migration: buoyancy, porous media permeability and present structural and stratigraphic features. Due to the fast computation this method is suitable for uncertainty quantification studies and allows to quickly screen through various geological scenarios. For instance, it can help identify and verify uncertain scenarios of kitchen location and migration pathways. This study will validate the method applied to the Wessex basin case study with its well-studied petroleum system. Firstly, regional open-source data is gathered and interpreted. Secondly, broad 2D seismic interpretations are utilised to construct a 3D structural model covering the entire petroleum system. Facies distributions are modelled with universal kriging and consider interpreted core data, wireline logs, and outcrop analogues. Uncertainties of each interpretation and modelling workflow step (e.g., fault transmissibility, fault offset, range of permeability) are identified and integrated into a Monte Carlo simulation to generate multiple scenarios. Secondary migration is then simulated through ABM for each scenario. Tectonostratigraphic events within the area of interest, resulting in a diverse spectrum of possible migration scenarios, are accounted for through adjustment of generated and expelled HC volumes as well as through stochastic geological model restorations. Lastly, accumulations of agents from the simulated approach are compared to proven oil reserves in place to elucidate potential play scenarios and validate the method. Through several postprocessing steps the migration pathways are analysed and assessed on their geological realism. Results indicate that this method is suitable to quickly mitigate the risks of potential prospects by validating and/or discarding various migration scenarios. Applying this original approach and complementing it with 3D geological models can be utilised at underexplored frontier basins to enhance the quality of quantitative judgment on future development decisions.