--> Abstract: Introducing Seismic Inversion to Reduce Uncertainties in Static Modeling, by Faical Ben Amor and Ofonmbuk Ekpoudom; #90081 (2008)
[First Hit]

Datapages, Inc.Print this page

Introducing Seismic Inversion to Reduce Uncertainties in Static Modeling

Faical Ben Amor1 and Ofonmbuk Ekpoudom2
1Schlumberger, Lagos, Nigeria
2Schlumberger, Port-Harcourt, Nigeria

Uncertainty quantification is an integral part of a reservoir modeling study and the impact is even more pronounced when there is limited well coverage. The spatial distribution of petrophysical properties within a heterogeneous reservoir is also affected by uncertainties. Porosity modeling is generally considered as a critical step in volumetrics and in the population of other properties such shale volume, water saturation and reservoir permeability. Using seismic inversion to predict porosity has successfully reduced the uncertainty away from the well control.

Given the reservoir extent and lack of well coverage, an acoustic Previous HitimpedanceNext Hit inverted seismic cube was used as 3D soft constrain for porosity population. Previous HitImpedanceNext Hit cube, after calibration with well logs, was critical to estimate horizontal variogram parameters rather than using analog data. For each mega structure compartment, likely to define an independent depositional environment, a variogram map was generated to provide valuable information on anisotropy direction, lateral heterogeneity, chronology of structural events and role of growth faults in controlling reservoir quality and thickness distribution. Vertical variogram was still derived directly from the well information. Previous HitImpedanceNext Hit based co-kriging allowed guiding inter-well porosity distribution. Water saturation model was subsequently derived from plotting porosity vs. Sw above OWC.

Using Monte-Carlo methods, uncertainty parameters (structure, porosity and water saturation) were stochastically combined to assess volumetrics uncertainty. STOIIP from the Previous HitimpedanceNext Hit constrained model has been compared with the one constructed without Previous HitimpedanceNext Hit. The resulting histograms show that Previous HitimpedanceTop based geomodel has less uncertainty on the STOIIP distribution. Using Experimental Design method, a ranking of the uncertainty parameters effect was possible.

Presentation GEO India Expo XXI, Noida, New Delhi, India 2008©AAPG Search and Discovery