Data-Driven Analysis of Complex Hydrocarbon Accumulations in Fractured-Vuggy Reservoirs: Making Decision Based on Facts Rather Than Previous Experiences

Abstract

Chinese deep oil and gas reservoirs are mainly in Paleozoic carbonate strata, which are characterized by ancient carbonate rock, strong heterogenous, poor petrophysical properties. Combined with multi-stages of hydrocarbon accumulations caused by multi-stage tectonic movements, the hydrocarbon distributions in these reservoirs are very complicated. Shallow exploration experiences cannot be directly applied to deep explorations. At present, even the favorable traps have been identified, there are still huge uncertainty to determine whether oil/ gas/water was exit.

Facing with the problems of strong heterogeneity and multi-stages of hydrocarbon accumulations and adjustments in deep carbonate reservoirs, this paper proposes a new workflow, using data-driven technology to synthesize geological, geophysical and production data, based on facts rather than prior experiences to make exploration decisions. 1) Cleaning data from many disparate sources, screening the data to eliminate the abandoned wells due to engineering problems; 2) Classifying the scheme of well types based on production capacity, fluid properties, petroleum accumulations, using data methods (pattern recognition, machine learning and artificial intelligence) to classify each well; 3) Geological interpretations, based on classification results of drilled wells, combined with geological and geophysical interpretations, discovering hidden relationships of different types of drilled wells from data, and pushing back hydrocarbon accumulation processes; 4) Decision-making for hydrocarbon exploration: Combine the static distributions of traps with the dynamic hydrocarbon accumulation processes to carry out the decisions of well drillings.

The multi-staged petroleum accumulation process analysis of deep reservoirs is the fusion of domain expertise (years of geology, reservoir, and production engineering knowledge) with data driven analytics. The “CCGD” workflow was applied in Tarim Basin, and got the data-driven deep hydrocarbon accumulation laws, with three ultra-deep condensate gas wells successfully drilled.

AAPG Datapages/Search and Discovery Article #90350 © 2019 AAPG Annual Convention and Exhibition, San Antonio, Texas, May 19-22, 2019