Data science is an emergent discipline focused on the extraction of useful information from data. This requires the application of tools and techniques from many fields of study, including:
The upstream oil and gas industry faces both unprecedented opportunity and unprecedented challenge in the years ahead. Technological and process advancements have enabled the extraction of resources previously considered unrecoverable from novel “unconventional” geologies.
However, theoretical understanding of flow and mechanics in these resources has lagged behind technological recoverability. As a consequence, the industry finds itself with great challenges in predictability and optimization as the “factory model” of unconventional resource development appears increasingly untenable.
Additionally, the digital oilfield revolution has brought ever-increasing volumes of raw data into the E&P enterprise. While this represents a huge opportunity, few operators possess the expertise necessary to extract the relevant information from the daunting mountain of real-time measurements, geophysical surveys, production histories, completion results, and more.
terminus data science applies data science techniques to tackle these challenges. In place of a long wait for “first principles” models, data science permits an empirical approach: discover the critical relationships that drive performance directly from the available measurements.
How can data science deliver better decisions for E&P organizations? Some examples:
Organizations often find themselves with a simple question: “what can we do with all this data?”
terminus data science can help identify promising patterns and trends for future investigation as well as spot potential data quality problems and recommend remediative actions.
Through application of data visualization, dimensionality reduction, and pattern recognition techniques—in interaction with technical domain experts such as engineers and geologists—hidden value can be discovered in even large, heterogeneous, and noisy data sets.
Using data from multiple disciplines to build holistic understanding of a reservoir can be particularly challenging, especially when faced with faced with a new play lacking significant producing history, or an unconventional resource play.
Data science techniques can provide a framework for these interdisciplinary studies. Multivariate and non-linear regression models, trained on available data and deployed through interactive visualizations, enable technical professionals to directly discover and inspect relationships and complex interactions between various parameters and outcomes of interest.
terminus data science can deliver these tools into the hands of domain experts and provide training in their use.
In the wake of the unconventional resource tsunami, operators have found themselves grappling with the challenge of optimizing well, completion, and field design as quickly as possible.
Outcomes are controlled by both geologic potential and completion success. Deployment of sophisticated predictive models can help untangle these inputs and enable, for example, geologically-tailored completion optimization.
terminus data science provides technical professionals with the understanding and tools necessary to extract maximum understanding from these data sets in order to make better decisions faster.