Propablistic Method of Petrophysical Analysis

Dear All
Can every body Help us to contribute in Probabilistic Methods for Petrophysical Analysis just like Methodology in Elan in geoframe package or Mineral Solver in Interactive Petrophysics or Quanti-Min in Techlog Softwares.
I want to understand the Theoretical basic theory how to applicative this procedures.
Thanks alot .
wael

Well, the way I perceive "Probabilistic analysis" is that petrophysicists set up or choose a certain mineral model that applies to the formation. They choose which log types should be calculated using the models. The computer algorithms than use this model to calculate the logs that are then compared to the original logs.

I'm not totally familiar with the methodology though..

Is this what you are referring to?

In this method you are supposed to tell the software what minerals (rocks) and fluid phases are present in the zone of interest. You also need to define which are the logs that are the most representative to define the zone. According to this input software calculates the most probable distribution of minerals and fluids. It also quantifies error for every model it calculates.

You can control the outcome (diminish the error) by changing some of the parameters, usualy clay and clay minerals, or by choosing different log curves as input and giving them different weighing factor (trust). The model you are building needs to be determined in such way:
number of minerals+number of fluid phases<=number of input curves+1.

This method of interpretation is not totaly controlable like the deterministic approach but gives excellent results and needs to be used only when you have extensive knowledge about the reservoir lithology (lots of core data, local knowledge...)