Logging natural fractures

Following on from the previous blog on logging natural fractures (not including induced fractures), this blog looks at constructing a fracture intensity log. There’s a lot of confusing terminology our there - some people use intensity, others density and there a whole host of “P” numbers that are quoted and used.

Clearing up the confusion

The idea is to provide some clarity here. Software such as Petrel will create a P32 log of fracture intensity from the imported (likely image log) interpretation of dips and azimuths, but the units are not the numbers of fractures per metre. Instead, the P32 captures the area of fractures per unit volume of rock mass (Fig. 1). Although these logs may be required for modelling purposes, a better starting point is to bin the fractures into depth increments such as 0.25 metres or 1 metre. This intensity log is equivalent to the P10 of Dershowitz & Herda (Fig. 1).

Create a P10 log from image log data - simply fractures per metre

I recommend creating a P10 in a spreadsheet from the fracture image log interpretation. The P10 can be brought into modelling software and should have a similar shaped log to the P32 if the same window is used for binning. The value of the P10 log is that the fracture density per metre can be communicated to others in the project and this type of log is likely more suited to operational decisions such as where to re-perf the formation. A lithology log should be displayed alongside the fracture intensity log to check if there is a mechanical stratigraphy i.e., more fractures in certain rock types.

Integrating data from rock core

Often core picks up the smaller fractures than cannot be detected on image logs. These may be important for fluid flow. Therefore, P10 logs created from core data should be used to calibrate the image log data. This should be carried out for each fracture sets which are defined on the basis of orientation.

Fig. 1. Different fracture abundance measures (Dershowitz & Herda, 1992)

An example P10 log is shown in Fig. 2. Here, natural fractures are counted ( every 0.25 metres and plotted over a short interval of chalk core (from the Valhall Field, N Sea). This is the red curve. However, there was some uncertainty as to whether certain fractures were induced by coring/handling etc. Therefore, I included these to create an upside log (green curve). By doing this, we are handling uncertainty.