Integration Of Seismic Variance Attribute With Stochastic Error Surfaces To Have A Better Definition Of Structural Uncertainty - A Case Study
D. Ghosh, B. Al-Enezi and D. Al-Matar
Event name: IPTC 2014: International Petroleum Technology Conference
Session: Session 46: E&P GEOSCIENCE - Geological Risk and Uncertainty Management
Publication date: 19 January 2014
Info: Extended abstract, PDF ( 1.51Mb )
Price: € 20
Structural uncertainty is defined by creating stochastic error surfaces built on control points. Uncertainty is zero at the drilled locations and varies smoothly away from the wells. The other factor that enhances uncertainty is fault-zone. This study aimed at generating a composite model integrating these two determinants of structural uncertainty. The study is done on Mauddud surface in part of the Greater Burgan Field, Kuwait. The seismic guided surface was created incorporating tops of 13 drilled wells. Sequential Gaussian Simulation was used to generate stochastic error surfaces having normal distribution using these 13 zero value control points as input. Deviation of the actual Mauddud top from the given seismic surface was calculated to be to the tune of ±60’. The stochastic error surfaces were multiplied with a constant so that the surfaces closely represent the perceived uncertainty captured in these drilled wells. Seismic variance attribute was used to capture the uncertainty in fault zone. Variance was extracted on Mauddud surface from the variance cube generated. This variance surface was normalized with minimum and maximum values 1 and 6 respectively to use it as a multiplier to the stochastic error surfaces. The assumption was that the uncertainty will increase six times where there is maximum variance. The stochastic error surfaces were multiplied by the normalized variance surface to get the composite uncertainty. This uncertainty model was used to predict the uncertainty of Mauddud top in some wells drilled subsequently. The actual tops were found to be within the P10-P90 range except for a graben well where it was beyond the range. This study thus provided a model to quantify the range of uncertainty in predicting tops taking into account both distance from control points and uncertainty associated with fault zones as captured by seismic variance.