Volcanic rock hydrocarbon reservoirs might be underestimated in exploration and exploitation due to threedimensionally heterogeneous distribution of volcanic lithofacies. Volcanic-related reservoirs are not among the primary producers of hydrocarbons in a global sense. However, further exploration and exploitation of volcanic reservoirs with an appropriate geological model may offer considerable potential for the discovery of significant reserves.
An understanding of the architecture of volcanic reservoirs is critical for their exploration and exploitation. We applied multiple-point geostatistics to the modeling of volcanic rock reservoirs in the Yurihara oil and gas field, in northeast Japan. The volcanic rock reservoirs in the Yurihara field consist of basaltic rocks erupted underwater during the Miocene period: sheet flows, pillow lavas, breccias, and hyaloclastites. This volcanic activity was related to the opening of the Japan Sea, a backarc basin behind the Japanese islands.
Volcanic morphology in slow spreading mid-ocean ridges is used as a modern analogue in developing a training image in the multiple-point geostatistical approach. Two conspicuous volcanic morphologies can be found in slow-spreading ridges:
flat-topped seamounts (dominant sheet flows and surrounding pillow lavas) and
hummocky terrains (conglomerations of pillow lavas). Identical morphologies and lithologies are suggested in 3D-seismic profiles in the basaltic reservoirs of the Yurihara field. The formation of flat-topped seamounts along en echelon-aligned fissures and surrounding hummocky terrains was modeled as a training image for the Yurihara basaltic reservoirs. Realizations were developed using borehole imaging, logging, and core data as hard data and were successfully used for flow simulations.
The application of multiple-point geostatistics to volcanic reservoirs has been demonstrated to be proficient in predicting reservoir performance through simulations using geostatistical models.
