BI Dashboarding Application in Reservoir Simulation

Abstract

A Business Intelligence (BI) tool is a type of software that is used to gather, process, analyze and visualize a large volume of data, whether it is historical data, live data or forecasting data for the future. The objective of implementing BI tool is to create interactive reports, generate actionable business insights, and to simplify / accelerate the decision-making process. Depending on the size and the maturity of their fields, reservoir engineers often have to deal with a tremendous quantity of data from various categories such as simulation input or output, within challenging timelines. It is not uncommon that simulation REs spend the majority of their time in data pre/post-processing, grouping, filtration and setting up visualization templates, before being able to finally perform some value-adding results analysis and eventually improve the model forecasts. This paper focus on the applications of dashboarding in reservoir engineering and simulation work using a popular BI software (Spotfire®) that is outperforming in many ways some industry-standard software, with objective of promoting the BI dashboarding culture within reservoir engineer population. Depending on the purpose, various types of dashboard could be built, which allow RE users to better discover patterns and unveil the real meaning behind the data. In this paper, three templates (including history matching quantitative assessment, scenario comparator and PVT data QC) currently adopted by Abu Dhabi National Oil Company (ADNOC) Onshore asset are illustrated. Benchmarking against some conventional industry-standard tools is also performed to highlight their added-value. As a result, in the context of a 2G & R integrated model review for a multi-billion-barrels reservoir, some concrete examples focusing on BI dashboarding assisted well-by-well history match is illustrated, which showcases how simulation REs could boost their daily work efficiency and create added value to the organization.

Benchmarking Human-AI Collaboration for Common Evidence Appraisal Tools

“…(3) Human consensus vs combined LLMs: consistency approach Combining multiple assessments from the same LLM in a "self-consistency" approach improves performance in biomedical and general contexts beyond using only a single assessment. 28,33 We combined a total of 9 LLM assessments: 2× Claude-3-Opus, 2× Claude-2, 1× GPT-4 (due to high costs), 2× GPT-3.5, 2× Mixtral-8x22B. 34 Deferring fraction: Inconsistent responses (i.e., without alignment among 6-12 out of the 12 LLM assessments) were considered uncertain and thus deferred to human raters, meaning the LLMs could not provide a consistent rating and a human must decide.…”

Research and Practice on the Construction of Municipal Industry-Education Consortium Based on the Integration of Industrial Chain, Innovation Chain, Talent Chain, and Education Chain

“…Universities are the "guiding force" at the junction of education, technology, and talent. It is necessary to strengthen systematic thinking, "step out of education and technology," promote the deep integration of the four chains of "industrial chain, innovation chain, talent chain, and education chain," and contribute scientific and educational wisdom to the high-quality development of the economy and society [3] . It is essential to build a circular ecological chain of industry-education integration that empowers education, cultivates talents, and supports industries, to truly promote the "four-chain integration," as well as reflect, consolidate, and optimize the positioning of vocational education types, thereby stimulating and unleashing greater vitality, enabling a bright future for students, transforming from "Made in China" to "Created in China," and achieving iterative upgrading of national industrialization.…”