Abstract: Methods for Testing the Validity of Surface Geochemical Anomalies

James M. Fausnaugh

A combination of statistics and compositional analysis is currently being used to test the validity of surface hydrocarbon anomalies. Once the hydrocarbon anomaly is detected a rigorous compositional analysis is undertaken to ascertain its source. Some of the techniques under investigation are Factor Analysis, Discriminant Analysis, diagnostic hydrocarbon ratios, and Pixler Plots.

Factor Analysis and Discriminant Analysis are used to detect any underlying structural relationships between variables. These statistical methods simplify the interpretation by reducing the number of variables to be used in the interpretation. The resulting factors are reclassified as a function of hydrocarbon composition. The extraction of various compositional structures can show the difference between reservoir and non-reservoir hydrocarbons. Compositional grouping based on molecular weight forms the basis of the interpretation. If thermogenic hydrocarbons are present, methane, ethane, and propane form one group while the heavier hydrocarbons form a separate distinct group. Non-reservoir hydrocarbons can also show specific relationships in composition, but usually exhibit no partic lar orientation or trend.

Hydrocarbon ratios have been used for many years to characterize reservoir attributes. Statistically processing the raw data prior to determining the ratios can yield information not readily apparent when using simple ratios.

Pixler plots show the relationship of methane to ethane, propane, butane, and pentane respectively. When plotted in graph form, these relationships, for any particular sample, may be representative of a producing or non-producing area or an oil-bearing or gas-bearing area. A more generalized compositional analysis is obtained by determining the slope and intercept of the line drawn through the coordinates of these ratios for each sample. Once determined, the slope and intercept values can then be plotted in map view.

AAPG Search and Discovery Article #90950©1996 AAPG GCAGS 46th Annual Meeting, San Antonio, Texas