State-of-the-Art Digital Approaches for Energy Efficiency and Carbon Intensity Management, for Upstream facilities

Abstract 1. Objectives/Scope In the challenge of reducing the oil and gas industry's carbon footprint, different dimensions emerge. The first dimension is the carbon-centric optimization of facility design, to minimize emissions while optimizing production and reservoir management through a full lifecycle integration of reservoir, processing facilities and utilities. It can be applied for any development plan or review of operating philosophy. The second dimension, particularly relevant for existing and producing facilities, is the optimization of emissions in day-to-day operations, addressing energy efficiency, and ensuring that no barrels are produced with a high carbon intensity. These two dimensions are intrinsically complementary. 2. Methods, Procedures, Process This paper focuses on the operational dimension, and presents two field-proven digital tools, one focusing on Energy Efficiency, and the other one on Carbon Intensity. The proposed innovative algorithmic strategies are state-of-the-art and go beyond business intelligence or monitoring: they are designed to be prescriptive, providing actionable recommendations. They use data-driven algorithmic methods, where machine learning and physical modelling are combined. The Energy Efficiency ActBoards tool diagnoses inefficiencies and waste across facilities (at equipment, energy system and facility level), and introduces a new powerful, multi-scale and straightforward method to calculate and track energy efficiency in real time. The Carbon Intensity Manager takes on multiple levers (well activation, pressure maintenance, processing, flaring), to analyze the deviations in effective carbon intensity for each individual marginal or incremental barrel. 3. Results, Observations, Conclusions Based on the presented methodologies and their associated case stories, we estimate that emissions can frequently be improved by 10% when focusing on proper energy efficiency levers, and typically by additional 3-5% when optimizing based on carbon intensity (accepting in the latter case some impact on production). The functional complementarity of the tools is discussed, as well as the lessons learned from a few years of implementation. 4. Novel/Additive Information The methods presented in this paper are original both in their design and their implementation. They succeed in analyzing multiple timescales and combine smoothly the whole spectrum of granularity (from the equipment to asset). They are action-oriented and prescriptive, and integrate the tracking of the actions or evolutions in operating philosophy initiated by operational teams. These solutions are field-proven and generalizable to most facilities. Their data requirements are compatible with the generally available data available on 1990s' and more recent fields. Their design proved fairly robust to real life data availability issues (proven in TotalEnergies operated domain). Case studies from the North Sea and West Africa will illustrate the value of moving from monitoring toward detailed fuel gas and energy diagnosis modelling.