|
WEB APPLICATION BASED ON COMPUTER VISION IN DETECTING CHAN WATER CONTROL DIAGNOSTIC PLOT Department of Petroleum Engineering, Faculty of Engineering, Pattimura University, Jl. Ir. M. Putuhena, Poka, Kec. Tlk. Ambon, Kota Ambon, Maluku, 97233 Abstract Chan water control diagnostic plots introduced by K. S. Chan were used to determine the mechanism of excessive water production, as well as observe the log plots of WOR and WOR^ vs. time to find out problems with water behaviour in the well which can occur with a normal displacement of oil by water, multilayer channelling, and rapid channelling. In identifying water behaviour problems in wells, the root of the problem is the inconsistency of human judgment and the absence of criteria in appropriate classification patterns, which are the key to identifying water behaviour problems in wells. While relying on professional human judgment is common, there is significant value in seeking consistency, and it may be difficult to differentiate clearly. This shows that there is an opportunity to find out the problem of water behaviour in the well if it is carried out correctly with the application of Artificial Intelligence (AI) to learn patterns from data sets and excel at predicting outcomes, in contrast to conventional software engineering, where rules are defined more explicitly. This paper developed a Web application with models of computer vision that can overcome problems of data accuracy and complexity, as well as create efficient and accurate visualizations and classification patterns, which can help take proactive monitoring decisions that require identifying signature patterns in Chan plots, so it can usefully automatically classify whether a well shows a particular Chan plot signature, to flag it for review in a broader petroleum engineering decision framework. Keywords: : Chan water control diagnostic, Artificial Intelligence, Web Application Topic: Reservoir engineering |
| ICPMGET 2024 Conference | Conference Management System |