Automating biostratigraphy in oil and gas exploration: Introducing GeoDAISY

Abstract

Biostratigraphy is a key upstream activity. Drilling operations use biostratigraphic data derived from core samples extracted via boreholes (mud logging) in order to guide drilling processes and verify models of basin geology. Current manual data analysis relies on the existence of sufficient numbers of biostratigraphers to be available to fulfil industry needs. However, due to factors such as retirement, and that the number of people being trained in the requisite taxonomic skills has declined sharply over the last couple of decades; the available pool of expertise is dwindling. It is clear that if the high cost of drilling operations is taken into account, the current situation is untenable and in many ways represents a perfect storm.This pessimistic situation can be radically changed by augmenting human expertise using automated species identification tools based on artificial neural network technology. DAISY [Digital Automated Identification System], a proven system of this sort, could revolutionise commercial biostratigraphy operations by enabling microfossil identification to be undertaken by technicians. This would yield immediate benefits for the industry as it would permit routine work to be performed quickly and accurately by less skilful, cheaper and therefore more available staff; freeing biostratigraphers to concentrate on non-routine, more complex tasks. The feasibility study presented here indicates that DAISY can consistently identify microfossils to species, with repeatable, high levels of accuracy. Crucially, it can also act as a permanent repository for taxonomic knowledge, which is currently lost when experienced personnel retire. There might also be additional environmental and social benefits if this technology is widely adopted within the oil and gas sector: as DAISY technology is generic, it can easily be re-targeted to interpret seismic data or even to estimate the impact of upstream exploration activities on abutting ecosystems.