WATCH VIDEOHistorically, the oil and gas industry had rarely been focused on quick wins. In the US onshore world, a company can spend 1 to 2 years evaluating a small area of land for purchase and then another 2-3 years developing it before seeing any oil. In the recent years, with the use of faster hardware and more advanced algorithms, there has been rapid changes in how new areas are evaluated. Data science has been instrumental in significantly reducing the amount of time that geoscientists, petrophysicists and engineers spend evaluating potential land leases. At Anadarko, with the use of machine learning, deep learning, GPUs and distributed computing, we have been able to reduce the time it takes to evaluate a US onshore basin from 1 year to 1 month, calling it Rapid Basin Evaluation. This process is centered on analyzing data of the wells that have already been drilled around the area of interest. The original process required several manual and labor intensive steps from a team of geologists, petrophysicists and engineers. In the new approach this effort is reduced significantly, by requiring experts to analyze only a handful of key wells and leave the rest to be analyzed by an appropriate data science solution using cloud infrastructure.
