Digital Twin Applications for Real-Time Optimisation of Oilfield Operations: A Critical Narrative Review of Architectures, Evidence and Operational Readiness
Victor Nnanyelu *
SLB, Lagos, Nigeria.
Babatunde Ojoawo
Ohio University, Ohio, United States.
*Author to whom correspondence should be addressed.
Abstract
Digital twins are increasingly proposed as the operational intelligence layer for upstream oil and gas assets, yet the term is applied to systems that range from static digital models to continuously synchronised, decision-coupled representations. This critical narrative review evaluates the evidence for digital twin applications that can support real-time optimisation of oilfield operations, with emphasis on drilling and well construction, production and flow surveillance, reservoir and integrated-asset management, and reliability and maintenance. Literature was selected through live searches of multidisciplinary and engineering-relevant scholarly sources, supplemented by citation searching and bibliographic verification. The synthesis distinguishes digital models and digital shadows from operational twins and evaluates whether reported systems demonstrate persistent data synchronisation, state estimation, predictive validity, uncertainty handling, decision optimisation and a credible path to physical actuation. The strongest evidence supports component- and process-level twins, hybrid physics-data soft sensing, virtual flow metering, drilling-fluid and wellbore modelling, and condition monitoring in bounded operational contexts. Evidence for fully integrated, autonomous, closed-loop oilfield optimisation remains substantially weaker because validation is often retrospective, experimental or confined to a single asset; uncertainty is inconsistently propagated into decisions; and operational latency, interoperability, cyber-security, model drift and human-oversight requirements are rarely evaluated together. Reservoir-scale twinning also operates on slower update cycles than drilling or surface-process control, making a single universal definition of real time technically misleading. The review argues that operational value depends less on visual fidelity than on decision fidelity: the ability of a synchronised model to estimate state, quantify uncertainty, compare feasible actions and support safe intervention at the timescale relevant to the process. Future progress requires prospective field trials, uncertainty-aware hybrid models, multi-rate cloud-edge architectures, standardised validity and reporting criteria, and explicit governance for human-autonomy boundaries.
Keywords: Digital twin, oilfield optimisation, intelligent drilling, production optimisation, virtual flow metering, reservoir management, uncertainty quantification, edge computing