Integrating outcrop data at different scales to describe fractured carbonate reservoirs: example of the Maiella carbonates, Italy

Inherent heterogeneity is the main cause of uncertainty in predicting the productivity potential of fractured carbonate reservoirs. Integration of subsurface (well and seismic) and outcrop data helps to reduce such uncertainty. For fractured reservoirs, outcrops are essential because of the difficulty of producing three-dimensional descriptions of fracture systems from subsurface datasets alone. We have studied the carbonates outcropping in the Maiella anticline to verify the consistency between the fracture sets predicted by a strain inversion technique based on the restoration of seismic-scale cross-sections and detailed field measurements. Comparison of the two datasets indicated that: (a) the strain inversion model predicts fracture density correctly when averaged over distances of the order of kilometres, but fails to match observations averaged over hundreds of metres; and (b) fracture connectivity observed in the field is higher than that inferred from the strain model. We compare the limitations of the strain model to the limitations posed by working with seismic data alone, and we warn that outcrop data, collected at the appropriate scale in three dimensions, are necessary information for building plausible geological models consistent with the geological evolution of the reservoir.