Automatic seismic and production history matching in Nelson using different updating schemes

A. Kazemi, K. D. Stephen

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    6 Citations (Scopus)

    Abstract

    Timelapse (4D) seismic is becoming a commonly used tool and helps reservoir engineers history match field data and make forecasts. Automated history matching processes are often restricted by large numbers of parameters that control changes to models. In this work we introduce and apply different parameter search strategies in the Nelson field. We reduce the volume of the parameter space heuristically by considering changes that are local within the reservoir and compare to those that are applied everywhere simultaneously. We find that we get similar results but local changes are applied more quickly if only one locality is modified at a time. We also investigate the improvements that are obtained through using time-laps seismic and report that seismic and production data are not always complimentary. In this study, production data is best used to modify reservoir properties near the well while seismic is more useful in the inter-well regions. Overall, improved seismic and production predictions were obtained and forecasts were also improved.

    Original languageEnglish
    Title of host publicationSociety of Petroleum Engineers - 71st European Association of Geoscientists and Engineers Conference and Exhibition 2009
    Pages1748-1752
    Number of pages5
    Volume3
    Publication statusPublished - 2009
    Event71st European Association of Geoscientists and Engineers Conference and Exhibition 2009 - Amsterdam, Netherlands
    Duration: 8 Jun 200911 Jun 2009

    Conference

    Conference71st European Association of Geoscientists and Engineers Conference and Exhibition 2009
    Country/TerritoryNetherlands
    CityAmsterdam
    Period8/06/0911/06/09

    Fingerprint

    Dive into the research topics of 'Automatic seismic and production history matching in Nelson using different updating schemes'. Together they form a unique fingerprint.

    Cite this