1999 Conference Abstracts

back to list...

1999 Abstract: Lucas and Bartek

Prediction of the Nature of Reservoir Heterogeneity of an Incised-Valley Fill and the Minimum Data Density Required to Characterize it: An Example from the Northeastern Gulf of Mexico

Jennifer L. Lucas
Department of Geology
University of Alabama
Box 870338
Tuscaloosa, AL 35487-0338

Louis R. Bartek III
Department of Geological Sciences
University of North Carolina–Chapel Hill
Chapel Hill, NC


This study involves characterization of the minimum amount of data required to accurately determine the three- dimensional (3-D) variability of the distribution of heterogeneity within reservoir facies of an incised valley fill. The focus of this investigation is on strata deposited in a continental-margin environment with low sediment supply, low velocity tidal currents, and major storm influence (frequency: once every 6 years). Minimum data density (MINDDA) analysis was conducted on this system, which is one end member of a spectrum of systems that are currently being studied to determine the minimum amount of data required to accurately recognize depositional systems and make predictions of lithofacies distributions in a whole spectrum of depositional conditions. A 1.2-km interval between data points was initially used to map the area, which was then undersampled at data-point intervals of 2.3 km, 4.7 km, and 9.3 km. Comparisons were made to observe and note contrasts between maps with decreasing data density. In order to ascertain the optimum data density for this type of depositional system, increasing percentage of difference between maps was plotted on the y-axis versus an increasing sample interval on the x-axis to determine the point of diminishing returns (PDR). PDR is the point where the difference between maps, sampled at different intervals, is very small. Data falling beyond this point fail to proportionally increase the overall understanding of the nature of heterogeneity in facies distribution and are therefore insignificant. Accurate determination of the PDR saves time and money for exploration and production by diminishing the risk of an incorrect prediction of homogeneity, which could result from working with a sample density that is insufficient to establish the existence of sedimentary facies with limited distributions. It also assists the interpreter of 3-D seismic data with determining the maximum interval between profiles that can be used characterize the system without encountering problems with spatial aliasing of the distribution of reservoir attributes.

The most prominent geological systems in this study area are two incised valleys (eastern and western), the incised valley fills, and interfluves associated with the incised valleys. Results from initial analyses reveal that the incised valley fill associated with the depositional- boundary conditions of the system investigated in this study require a sample interval of 2.3 km between data points, whereas the interfluves require a sample interval of 4.7 km to accurately map variations in facies distribution. MINDDA has already been applied to a high-energy, high- sediment-supply end member, and upon completion of this study, a comparison between the two end members will also be conducted.