首页|'Reservoir Modeling And Well Placement Using Machine Learning' in Patent Applica tion Approval Process (USPTO 20240110469)
'Reservoir Modeling And Well Placement Using Machine Learning' in Patent Applica tion Approval Process (USPTO 20240110469)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
The following quote was obtained by the news editors from the background informa tion supplied bythe inventors: “In conventional systems, several workflows incl ude calibrating model parameters to select amodel that accurately represents da ta. For instance, history matching is a process of calibrating reservoiruncerta inty parameters with an objective of obtaining simulation production data as clo se as possible to historical data. The process is time-consuming and requires he avy computations, which pose a bottleneckfor further steps in a reservoir model ing workflow.“Further, once one or more models are selected, the models may be used to design wells, drillingparameters, production plans, etc. However, given the number of models that may be simulated in orderto accurately represent the reservoir, th e number of potential locations for the wells, and the number ofdifferent desig n parameters for a well at a given location, the process of using the wells to p rovide usefulinformation can take a large amount of time, expensive or potentia lly unavailable processing power, orboth. Thus, in addition to streamlining the reservoir modeling process, systems and methods for moreefficiently implementi ng the models to provide useful insights, e.g., for well placement analysis, wou ld bea welcome addition.”
NETL
NSTL
