4.6 Integrating petrophysical uncertainty

Like all of us, petrophysicists are dealing daily with uncertainty. Well logging is not error free. Interpreting the logs can be challenging. The core measurements might be inconclusive. And above all remains the fact that we work with a very limited dataset to understand a very complex environment – our reservoirs.

It does make sense to include the petrophysical uncertainty in our geomodeling workflows, as we do with geological or geophysical uncertainty. In practice though, few studies seem to consider this uncertainty. If you have to prioritize, we recommend that you first check if your geomodeling workflow properly takes into account the complex relationship between the properties as described in the previous section. Then, once checked, and fixed if needed, look at the petrophysical uncertainty.

As described in this paper, we should ask for three types of input from our petrophysicist: well logs of course, a description of the complex relationships between the different properties (what log was modeled from what log) and the mathematical equations linking some of these properties (if any). Each of these data can carry some uncertainty.

Instead of one version of the log curves, you could ask for several versions them. SW is a good candidate for this request. It is a complex property to model and your petrophysicist might want to see the impact of using slightly different values for the parameters of the Archie’s equation. He could generate for you several logs of SW. The overall relationship between the properties (what log was computed from what log) is the least likely to change in a project. But it could be good to ask about, just in case. At last, your petrophysicist might decide to give several versions of the mathematical equations he is using (different porosity-permeability functions maybe).

Ultimately, these three types of uncertainty can be used, either through simple editing of your current geomodeling workflows, or in extreme cases, in defining several workflows, each one capturing a different petrophysical analysis.

  • Hits: 1972