Subseismic-scale geological information from reservoir analogs, when integrated with reservoir seismic data, substantially improves reservoir modelling. Wüstefeld et al (2018) developed a new workflow for 1) automated detection of subseismic-scale fracture surfaces exposed in reservoir analogs using terrestrial light detection and ranging (t-LIDAR), and 2) integration of the analog-fracture data in the standard industrial reservoir modelling routines (e.g., in Petrel software). In this workflow, the fracture surfaces detected along horizontal scan lines are used to derive one dimensional fracture density (P10) that is further used as an input for discrete fracture network modelling for the reservoir. Apparent P10 values along scanlines need to be corrected to get actual fracture densities (Terzaghi 1965).
We developed a script in MATLAB that uses the fracture surfaces data (detected through standard workflows in 3D point cloud data) to obtain Terzaghi-corrected P10 values for each fracture orientation. Based on the user-defined condition for subparallelness (e.g., angle between fractures < A°), the script uses normal vectors of the detected fracture surfaces to classify them into clusters of subparallel fractures. It then obtains the mean orientations of different subparallel-fractures-clusters. Finally, the normal vector corresponding to the mean orientation of each cluster and spatial positions of the detected fracture surfaces are used to calculate perpendicular distances between the subparallel fractures (i.e., Terzaghi-corrected P10 values). The corrected P10 values may then be used for further reservoir modelling approaches or distances between neighboring subparallel fractures can be used to assess clustering based on the normalized correlation count approach.