David Standingford of Bristol, UK-headquartered Zenotech reports on trials of a cloud-based combination of reservoir fluid-flow simulation with optimization technology from ‘masters of uncertainty’ Epistemy. Jonathan Carter, former CTO with E.On, initiated the study that involved tweaking multiple input parameters to Rock Flow Dynamics’ tNavigator flow simulator to quantify model uncertainty and optimize well placement. tNavigator was controlled from Epistemy’s Raven Bayesian history matching and optimizing toolset using a 10-dimensional Latin hypercube. The trial involved 31 models and 2,000 runs, more than in-house resources could handle.
Carter turned to Zenotech to use its EPIC front-end to compute resources in the Amazon web services cloud. EPIC (elastic, private, interactive, cloud) provides ‘easy and secure’ access to AWS resources. Standingford emphasized the flexibility of cloud-based HPC in oil and gas and also the ‘huge economic benefits’ to accrue from improved modelling and simulation. The whole optimization task was completed in12 days. The AWS bill amounted to $750.
© Oil IT Journal - all rights reserved.