Abstract It is common to formulate the history-matching problem using Bayes’ theorem. From Bayes’, conditional probability density function (pdf) of uncertain model parameters proportional prior pdf parameters, multiplied by likelihood measurements. The static are random variables characterizing reservoir while observations include, e.g., historical rates oil, gas, and water produced from wells...

Reservoir simulation is ubiquitous in the petroleum industry, and conditioning models to production data (history matching) has become an essential job of reservoir engineers. Traditional history matching methods directly perturb reservoir properties without regard to the existing geological continuity. When the geological heterogeneity is destroyed, the result is often a history-matched model ...

In recent years, machine learning (ML) has become a buzzword in the petroleum industry with numerous applications that guide engineers toward better decision making. The most powerful tool production development decisions rely on is reservoir simulation modeling procedures, such as individual runs, history matching and forecast optimization. However, all these lead to considerable computational...

Reservoir history matching refers to the process of continuously adjusting parameters reservoir model, so that its dynamic response will match historical observation data, which is a prerequisite for making forecasts based on model. With development optimization theory and machine learning algorithms, automatic has made numerous breakthroughs practical applications. In this perspective, existin...