Challenges in modelling of hydraulic fracturing in low-permeability coal seams with complex cleat networks and stress regimes

Abstract

Improved hydraulic fracturing models can better inform operational decisions regarding production from low-permeability coals and ultimately convert currently classified contingent resources to reserves. Improving current modelling approaches requires identification and investigation of the challenges involved in modelling hydraulic fracture stimulation in complex eastern Australian cases where permeability systems and stress regimes can vary significantly. This study investigated differences among existing and emerging advanced hydraulic fracture models and codes including numerical methods used to model fluid and rock behaviours during treatments; the ability to contextualise structure, behaviour and interaction of natural fractures with the propagating hydraulic fracture (e.g. cleat or natural fracture fabric, discrete fracture networks and pressure-dependent leak-off); and their capabilities in handling simultaneously growing or complex fracture development. One finding is that the new generation of models or codes that fully or partially use particle-based numerical methods are more capable in handling complexities associated with hydraulic stimulation of naturally fractured reservoirs. However, the computational cost and time for these models may cause concerns, particularly when modelling large reservoirs and treatments. Based on these limitations, many of the advanced, industry preferred, commercial hydraulic fracture simulators still choose to incorporate limited complexities with regard to natural fractures or represent them mathematically or implicitly. This investigation also indicates that most emerging models provide better representation of natural fractures, visualisation and integration into workflows for completion or stimulation design.