--> Abstract: Advances and Future Trends in Geomodeling Techniques to Populate Facies and Petrophysical Properties, by Claude Scheepens; #90205 (2014)

Datapages, Inc.Print this page

Advances and Future Trends in Geomodeling Techniques to Populate Facies and Petrophysical Properties

Claude Scheepens
ConocoPhillips

Abstract

In the last two decades several interesting advances in the world of geomodeling have taken place, several new methods were developed to populate facies and petrophysical properties. In this abstract two techniques will be discussed that have been developed and implemented within ConocoPhillips to better represent and more accurate model facies and petrophysical properties.

The first technique discussed is the Bayesian updating technique which consists of combining multiple attributes at various scales. There are many attributes that correlate parameters, typically only the attribute with the highest absolute value correlation is chosen to be carried forward to influence prediction. Bayesian Updating is a statistical theory relating conditional probabilities through multivariate correlations. At ConocoPhillips a Bayesian updating Petrel plug-in is developed to quantify the relationships and update the models.

The second technique discussed is a multipoint simulation method with an existing reservoir model as training image. The multipoint simulation (MPS) method has been increasingly used to describe complex geological features of petroleum reservoirs. This method is based on multipoint statistics from training images that represent geological patterns of the reservoir heterogeneity. However, the traditional MPS algorithm requires the training images to be stationary in space, although the spatial distribution of geological patterns/features is usually non-stationary.

We focus on a case where the training images comprise patterns that are non-stationary, in the sense that they are location dependent. These training images can be built by process-based, object-based or any other type of reservoir modeling approach. In ConocoPhillips we have incorporated a new MPS algorithm as a Petrel plug-in that can use an existing model as a training image and condition it to well data.

We will present results for both techniques.

AAPG Search and Discovery Article #90205 © AAPG Geoscience Technology Workshop, Permian and Midland Basin New Technologies, September 4-5, 2014, Houston, Texas