Fracture mapping and analysis

This tutorial walks through a fracture study on a virtual outcrop from beginning to end: getting the fractures off the model, turning them into orientations, splitting those into sets, measuring how intense the fracturing is, and — if you have the licence for it — building a Discrete Fracture Network that reproduces what you mapped.

Each step has a reference guide behind it that covers every parameter. This tutorial is the route through them: what to do, in what order, and what to decide at each fork.

The workflow

   Outcrop model
        │
   1.   ├─► Map fracture traces ──────────── by hand on the model
        │                              └──── or from photographs (AI / ACO)
        │
   2.   ├─► Orientations ─────────────────── measured, or estimated from traces
        │
   3.   ├─► Fracture sets ────────────────── stereonet, or auto-cluster
        │
   4.   ├─► Intensity ────────────────────── scan lines (P10)
        │                              └──── intensity maps (P21 / P32 / P33)
        │
   5.   └─► Discrete Fracture Network ────── generate, condition, validate

The first four steps are the study. The fifth turns it into a model you can propagate away from the exposure — it needs the DFN licence feature, and everything before it stands on its own without it.

What you need before you start

A 3D outcrop model	A triangulated mesh is the most useful — the intensity maps and the DFN comparison both need mesh geometry. A point cloud works for capturing orientations but not for the mesh-based steps. See Importing 3D outcrop data.
A coordinate system on the project	Intensity is measured per unit area and per unit volume, so the numbers are only meaningful if the model is in real units. Set it before you start: Setting the coordinate system.
Calibrated photographs (optional)	Only for the photo-based capture route in step 2 — an SfM image with a recovered camera, or a photo registered to the model.
The DFN licence feature (optional)	Only for step 5. If the Interpretation tree has no Discrete Fracture Network node at all, you do not have it — see Feature availability.

Work through the basics first

This tutorial assumes you can already create a project, load a model and digitise a polyline. If any of that is unfamiliar, start with the basic tutorial.

A note on what you are measuring

Two things are worth deciding before you click anything, because they shape every number that comes out the far end.

What counts as a fracture. Set the rule before you map, not while you are mapping: a minimum trace length, whether you include bedding-parallel partings, whether you separate joints from faults. VRGS lets you record the distinction (orientation Classification, and the facies/horizon Active Descriptor), but only if you decide it up front.

What the exposure is not showing you. An outcrop samples fractures with a bias: sets running parallel to the face are under-represented, and fractures longer than the exposure are cut off at its edges. VRGS corrects for the first where it can — the Terzaghi correction on a scan line — and the DFN validation report handles the second as censoring. Neither correction invents data that the outcrop never held, so it pays to know which sets your exposure is blind to.

The steps

1. Mapping fracture traces — digitising traces on the model by hand.
2. Fractures from photographs — the AI and ACO detectors, and projecting the result onto the model.
3. Orientations and fracture sets — getting orientations, and splitting them into sets on the stereonet.
4. Measuring fracture intensity — scan lines, and P21/P32/P33 intensity maps.
5. Building a DFN — generating a network from what you mapped, and checking it against the outcrop.