Abstract This paper presents a data science focused methodology for calculating carbon footprints for products and services in the upstream oil and gas service industry. Complementing traditional approaches, the method leverages existing corporate data sources, machine learning, and other data science techniques to navigate the complexities and challenges associated with conducting life cycle assessments (LCAs) across the extensive range of products and services in this industry, enabling the acceleration of decarbonization actions. Our approach leverages a comprehensive corporate greenhouse gas inventory (including scopes 1, 2 and 3 both upstream and downstream), along with revenue and associated product and service parameters recorded during delivery to customers. These datasets are integrated, and data-science methods are employed to explore and refine relationships among these parameters. Using data science methodologies, we derive over 561,000 emissions factors for various combinations of countries, products, service types, and timelines. These emissions factors are then further refined using additional techniques to estimate the carbon footprint of individual products and services provided to exploration and production (E&P) customers. The methodology has been successfully applied to estimate greenhouse gas (GHG) emissions associated with product and service delivery in the E&P service sector, and its results have been validated against cases where domain-specific LCA calculation methodologies and corresponding activity data are available. Although the intention is to complement an LCA approach rather than replace it, our novel method enabled estimations of product and service carbon footprints in a far shorter timeframe and at greater scale. This leads to rapid insights such as emissions trends, hotspots and reduction opportunities both at large scale, such as at country or business unit level, and at small scale, such as for individual operations or product deliveries. It enables targeted insights into what services and products warranted a more traditional LCA approach to characterize their footprint. By reducing the time needed for emissions calculations and linking emissions sources with known decarbonization technologies, this method can accelerate decarbonization actions. Although this paper focuses on a use case specific to the oil and gas service industry, this methodology also has the potential to benefit the broader E&P sector as well as other industries in moving from emissions calculations to decarbonization actions.
Edmundson et al. (Mon,) studied this question.