Publications

One of the key focus areas of oilfield digitalization efforts is the design and development of integrated production optimization workflows. The goal is to have a seamless and scalable solution that automates reservoir and well performance tracking, coupled with manage-by-exception capabilities that optimize the entire network. The current manual and interpretive methods pose a challenge in reaching this desired state due to the complexity of integrating disparate data sources and streamlining cross-functional workflows. The potential impact of an automated reservoir and production surveillance solution is significant, where operators stand to gain significant productivity increases and maximize the production of their assets. This paper presents field case studies of implementing such digital solutions on 3 different assets: an offshore gas field, an offshore oil field, and an onshore gas field, outlining the workflows deployed and the value realized through their implementation. A digital solution integrating multiple data sources and timescale resolutions was implemented to automate multiple reservoir and production workflows. Under reservoir workflows, daily flowing bottom-hole pressure calculations and event detection algorithms were used to identify automatically and catalog planned and unplanned shut-in events on each well. A hybrid model (physics-informed data-driven approach) was developed to perform pressure build-up analysis to estimate key diagnostic parameters such as average reservoir pressure and productivity index. Steady-state detection algorithms were then used to update well inflows with changing operating conditions. The calibrated well and reservoir model outputs were then fed to an integrated network model that continuously identified production optimization opportunities and performed short-term forecasting under different scenarios and constraints to maximize field production. In addition, an exception-based surveillance workflow encompassing a wide range of anomaly detection capabilities, such as ML (Machine Learning)-based choke erosion monitoring and liquid loading detection, was deployed to capture under-performance at the well level autonomously.