Integrated Asset Management (IAM) is highly complex and requires the combined effort of several disciplines and technological tools. The orchestration of disciplines, workflow tools, and available data are paramount issues in the oil and gas industry today. Resource negotiation, communication language, and decision-making protocols are minor issues that exacerbate the problem, resulting in poor and delayed decision making.
This study approaches these challenges through the implementation of innovative, distributed artificial intelligence (AI)-based architecture, designed for automated production management. This architecture, known as the Integrated Production Management Architecture (IPMA), has three layers: a connectivity layer, which allows access to the process information sources; a semantic layer, which establishes an ontological framework to guarantee the process-information integrity during the data interchange process developed between the applications that belong to the Enterprise Technology Information Platform; and a management layer, which automates the production process workflows using oilfield multi-agent systems and electronic institutions.
A virtual oilfield, based on a commercial-integrated production model (IPM) and history-matched data, was used to show the benefits of the proposed approach. The IPM had the following configuration: Three reservoirs, eight oil wells, one flow station, and several gathering pipelines. The IPMA's objective was to maximize asset revenue under different constraint scenarios and changing operational events. The reactive capacity of the architecture, the effective communication between the agents, and the proposed oilfield ontology were tested successfully. In this sense, the well function of the three layers of IPMA was demonstrated.
This study also outlines the use of ontologies and AI techniques that are important factors in future developments of IT solutions for the oil production industry.