Validation Processes for Subsurface Fracture Models

Abstract

Full characterization of fracture networks in the subsurface requires prediction and calibration as subsurface sampling of fractures is typically sparse and along one-dimensional drill tracks or well-bores. This data must be extrapolated to characterize the fractures in three-dimensions. A statistical approach to this extrapolation can carry high risk, particularly in heterogeneously fractured areas, with limited direct observations. As a result, proxies are typically defined to capture the three-dimensional variability and used as a guide for the accurate prediction of fracture orientation and distribution. Proxies for three-dimensional variability may include, for example, the use of outcrop analogues (i.e. Shackleton et al 2001), seismic attributes (coherency data, seismic amplitude etc.) (i.e. Dubrule et al. 1998], or may be related to the geological history and structural framework, critical in the generation of fracture systems in the real world (for example, Shackleton et al. 2001, Bond 2010, Kloppenburg 2011 and Mannino 2012). The definition of these proxies as fracture attribute data can be a complex process (Aguilera 1980, Berkowitz, 2002) and should always be based on an understanding of the fracture set being modelled. Validation provides a way to define the uncertainty associated with each assumption made during the modelling process and determine if the model is fit for purpose. The validation methods described here are split into validation of the geological model and validation of the created fractures. Validation of subsurface fracture models using structural modelling workflows is carried out using MoveTM software. Without first validating the geological model fracture modelling may be invalid. Risk associated with fracture prediction can be minimised by using validation as a way to define and reduce uncertainty with the modelling process. Fracture systems have a critical effect on the production and recovery of reserves in extractive industries and an equally critical impact on geological storage operations. Therefore for successful reservoir management Reservoir Engineers require accurate fracture predictions as input to reservoir models. Similarly, mining engineers and geologists require predictive models for fractures in geotechnical planning and exploration, which frequently depends on the prediction of mineralised zones controlled by fracturing.

AAPG Datapages/Search and Discovery Article #90216 ©2015 AAPG Annual Convention and Exhibition, Denver, CO., May 31 - June 3, 2015