AN APPROACH TO IDENTIFICATION OF SOURCE ROCKS

Abstract

Hydrocarbon distributions were studied in nine crude oils from the Surat Basin, Queensland, six being from the Precipice Sandstone, and three from reservoirs in the Evergreen Shale. In addition, distributions were determined in solvent extracts from eight shale cores taken from strata adjacent to three of the oil reservoirs.

The distribution curves provide data on relative concentrations of n-alkanes above about C9 (nonane) and also of "branched-cyclic" compounds. These latter are all saturated substances, comprising branched alkanes together with cyclic compounds, branched and unbranched. Aromatic and asphaltic compounds are not considered.

The n-alkane distributions of the oils all have their maxima below C20, whereas nearly all the shales maximise around C27. The distributions in the shales is confirmed as typical by results from a continuously cored stratigraphic well drilled through the Evergreen Shale in the north-eastern part of the basin, near Taroom. (This work is to be reported elsewhere.) Further, the distribution is the expected one in rocks in which most of the organic matter has come from land plants.

Since the concentrations of extractable hydrocarbons in the non-marine Evergreen Shale are as high as those found in marine shales, there is no apparent reason why it could not have acted as the source of the oils. The n-alkine distributions suggest, however, that it did not. While this might be acceptable for the oils from the Precipice Sandstone, which underlies the Evergreen Shale, and unconformably covers older rocks containing Permian marine sediments, it is probable that at least the Evergreen oils had their source in the Evergreen Shale, and that the distribution differences in the n-alkanes are caused by differential adsorption of these compounds on silicates (e.g. clays) during migration.

If the migration of oil is to be understood, then this differential "filtration" effect requires study.

Branched-cyclic distributions are very similar in the oils, but cannot be correlated well with the shales, which is to be expected, even if the shales are the source rocks of the oils, owing among other factors to the rigour of laboratory extraction procedures compared with what goes on in nature.

It is probable that the oils of the upper and lower sands in the Moonie field had the same source, and accordingly the branched-cyclic patterns are identical, although the n-alkane distributions are slightly different. But the branched-cyclic patterns of all the Surat Basin oils are very close, except in the case of the Conloi oil. However, the branched-cyclic pattern of an oil from another Australian basin, far distant from the Surat Basin, resembles very nearly that of the majority of Surat Basin crudes. Thus at the level of resolution so far employed by us, branched-cyclic distributions do not generally vary much from one oil to another. High resolution of branched-cyclics would almost certainly allow "finger-printing" of crude oils, but the possibility of using these compounds to match oils with their source rocks has yet to be thoroughly explored.