Output details
15 - General Engineering
Heriot-Watt University (joint submission with University of Edinburgh)
Bayesian optimization algorithm applied to uncertainty quantification
This paper is the first application in the oil industry of the Bayesian optimization Algorithm (BOA) to history matching reservoir models. This paper, and its companion on real field applications (DOI:10.2118/141161-PA) were funded by TSB (TP-BE056A £697k), and show that the studied algorithm outperforms evolutionary stochastic algorithms employed in history matching and uncertainty quantification on both synthetic and real cases. As a result of the performance demonstrated in this paper, the algorithm has been coded up in the commercial software Raven (www.useraven.com), with 3 licences sold to date (further details in the associated impact case study).