Predicting Hydraulic Flow Units for Enhanced Permeability Modelling Berkine Basin, Algeria

Kevin Corrigan and Chris Howells
Anadarko Algeria Company LLC

Abstract:

The Berkine Basin represents one of the significant success stories of Algeria with the discovery of several billion barrels of hydrocarbons. One key factor in the success of the Sonatrach-Anadarko Association was the initial value of the conventional core data. To date, in excess of 8km of core have been acquired from many different fields over a geological area extending several hundred kilometres and which, in many cases, is continuous across the reservoir interval. This extensive data acquisition and analysis program has resulted in a significant increase in geological understanding of the reservoir interval and work is currently directed towards identifying geological controls on subsurface flow of hydrocarbons and the need to better describe the permeability distribution within the reservoir.

The presentation focuses on a study of a Berkine Basin field, the results of which have subsequently been applied to nearby satellite fields. The ultimate objective of the study is to better describe the 3D subsurface flow in the Triassic sandstone (TAGI) reservoirs in the Berkine Basin by improving the calculation of permeability. To this end the applicability of using Hydraulic Flow Units, as predicted by the use of an artificial neural network, is tested.

The approach utilizes a program called Spotfire to identify the controlling factors on permeability and to maximize the benefit of this extensive dataset. It can be shown that a single porosity-dependent permeability predictor is insufficient to describe permeability in every well, even after extensive subdivision of the TAGI sandstone layers. It has been recognized that application of a Timur-type equation leads to a significant improvement but only in zones of irreducible water saturation above each OWC. The prediction of Hydraulic Flow Units, using the method of Abaszadeh, Fujii and Fujimoto, reduces the uncertainty in the calculated permeability, once sufficient training of the artificial neural network has taken place, and gives confidence to permeability estimation where core is not present.

The authors would like to thank Anadarko Algeria Company LLC and its partners Eni-Agip, Maersk Olie Algeriet AS and Sonatrach for permission to give this presentation.

Biographical Summary:

Kevin Corrigan joined Anadarko Algeria Company LLC in 1996 where he is currently a Senior Petrophysical Advisor working in the North Africa and North Atlantic region. He has over 28 years of experience in the industry, is a Chartered Engineer and holds a BSc. degree in Physics from the University of Leicester. He started work in Schlumberger in their Log Interpretation Centre in Paris, and then as a Field Engineer in Libya and the Middle East. This was followed by 5 years in BP in their International Exploration Group in London as a petrophysicist and later in Aberdeen as a Senior Petroleum Engineer. Prior to joining Anadarko, Kevin was a consultant for 11 years working on a number of integrated, international projects out of the UK.