Datatypes from the sub-surface are often fundamentally limited, either because they are extremely sparse (e.g. borehole image logs, core, well tests), or too low resolution (e.g. seismic), and hence unable to image the vast majority of fractures that constitute most naturally fractured reservoirs. Quantitative analysis of fracture systems in outcrop can provide critical constraint on many aspects of geomechanics, fracture modelling and flow simulation. Key applications include:
(i) to predict the likely range of fracture properties in the subsurface (wells can then be engineered, for example, to intersect or avoid fractures that may be open in the present day stress field);
(ii) to understand the spatial variability of fracture parameters in relation to structures mapped in the subsurface, including folds and faults (development can then be planned accordingly).
We present several case studies that demonstrate the importance of improved geological understanding to the application of fracture characterisation, based on fieldwork and quantitative fracture characterisation of carbonates in the Middle East and shales in the UK, in relation to hydrocarbon extraction.
Analysing mechanical stratigraphy is important to understanding fracture characteristics present in the rock volume and to predict the way in which rock may respond to changing stress. Large areas of contiguous unvegetated outcrop with high topographic relief in the Musandam peninsula in UAE and Oman allow different scales of mechanical behaviour to be interpreted over large areas of rock, and for lateral and vertical variability to be quantified in detail.
Another case study examines homogeneous shale units from the Cleveland Basin, UK. Lateral variability within a mechanical unit can be useful to constrain the range of fracture properties likely to be encountered. Within a mechanical unit potential barriers to fractures (and hence barriers to fluid flow with differing mechanical properties) can be identified. Fracture orientations, size ranges and fracture spacing distributions are compared with predictions from pseudo-wells, and used to assess connected volumes. Contrasting background fracture properties with local variations yields insight into significant local differences in geological histories which would be relevant for a development area.
Further case studies demonstrate how field data that quantifies the spatial variability of fracture parameters in relation to folds (four-way closing anticlines) illustrates the limitations of over-simplistic geomechanical models that implicitly assume a correlation between fold curvature and fracture intensity.