Datapages, Inc.Print this page

Better Prediction of Fluid-flow Modeling -Multi-Component Modeling of Petroleum Migration and Accumulation

Armin Kauerauf1,Thomas Hantschel1, and Robert Tscherny2
1IES GmbH, Ritterstr. 23, 52072 Aachen, Germany
2Chevron Energy Technology Company, 1500 Louisiana, Houston, TX 77002

Fluid-flow modeling is in many cases the most challenging and time consuming task in an integrated basin modeling approach. Because of this the chemistry and physics are often simplified in standard fluid flow modeling workflows. Compositional changes and variation of dependent physical properties, such as viscosities and densities, are not often taken into account. The standard approach for fluid flow modeling often includes bulk petroleum or one oil and one gas component and using methods like the black-oil models for chemo-physical description. The black-oil model is based on two pseudo components describing pre-defined properties. These pre-defined properties reduce the predictive ability of fluid flow modeling significantly, as it forces the model to a decisive (and pre-defined) outcome. However, it is well known that accurate modeling of the reservoir fluids densities is not only necessary for API gravity prediction but also for break-through analysis. Thus compositional effects cannot be neglected in general in migration modeling. One way to increase the predictability in fluid-flow modeling is to use multi-component description together with flash calculations to describe the fluid during all stages of migration. This analysis must include all stages of migration from expulsion, secondary migration, entrapment, breakthrough to dismigration.

This talk will show and compare an implementation of multi-component methodologies into fluid flow algorithms: The modeling methods used are Darcy flow modeling, ray-tracing based flow modeling, a combination of Darcy flow and ray-tracing (hybrid) and invasive percolation. Focus is put on multi-component implementations of these methods. The same PVT-analysis algorithm is applied in all models. This enables better comparison of the fluid flow methods itself. A result from a case study clearly shows the necessity of applying multi-component fluid flow modeling with advanced PVT-property prediction as a ’standard’ method. This example shows the advantages and disadvantages of the individual methods. Although, statements concerning the superiority of one method compared to another cannot be made because each method has its own advantages and disadvantages.


AAPG Search and Discover Article #90066©2007 AAPG Hedberg Conference, The Hague, The Netherlands