The effect of present day in situ stresses and paleo-stresses on locating sweet spots in unconventional reservoirs, a case study from Moomba-Big Lake fields, Cooper Basin, South Australia

摘要

The effect of stresses on permeability is a combination of external stress and pore pressure. We are examining if and how present-day in situ stresses and the spatial distribution of permeable domains in the Moomba-Big Lake fields in the Cooper Basin are correlated. We analysed image logs, well logs, and formation tests and calculated the orientation and magnitudes of the three principal stresses. A 3-dimensional model was constructed and the calculated stress magnitudes and orientations were applied to the model. The resulting stress distribution under the current day stress state showed a highly permeable domain indicating a sweet spot in the Big Lake field. This is currently the location of a gas pool that forms, with the Moomba field, one-third of the gas reserve in SA. No potential sweet spots are located in the Moomba area according to the stress model. We also used the finite element method (FEM) and the boundary element method (BEM) for modelling the behaviour of folds, fractures, and faults that formed during the tectonic history of the basin. We used geomechanical restoration techniques for locating sweet spots in the Moomba-Big Lake fields. The methodology attempts to reconstruct the current day structural and geometrical placement and predicts fractures generated due to stresses released during past tectonic events. Orientation of predicted fractures using FEM-based geomechanical restoration correlated well with the orientation of the image log fractures. The spatial distribution of paleo-stresses applied on the predicted fractures showed a potentially stressed fracture set in the location of the currently producing Big Lake sweet spot. However, orientation of predicted fractures using BEM-based geomechanical restoration correlated well next to the Big Lake fault but did not show any correlation away from the major fault. This is due to the fact that BEM restoration takes in consideration fault dislocation as the only driver of fracture generation and ignores the other factors. However, paleo-stress distribution using BEM restoration predicted the same producing area but with less accuracy due to the fundamentals of the BEM. No fracture density information can be extracted from any of the methods as the methodologies generate fractures with density that depends on the initial project mesh size. Accordingly, these methodologies can be used for locating the current-day and paleo-stresses, as well as fracture orientation but not density. Also, reservoir permeability is proved in this study to be controlled by a combination of current day and stored paleo-stresses.