Multicriteria hydraulic fracturing optimization for reservoir stimulation

Mohammed Motiur Rahman, M. K. Rahman, S. S. Rahman

    Research output: Contribution to journalArticlepeer-review

    20 Scopus citations

    Abstract

    A complete model is presented in this article for design optimization of hydraulic fracturing to enhance reservoir production. The model inegrates reservoir properties, various design constraints to avoid induced formation damage, a hydraulically driven fracture geometry model, a hydrocarbon production model, and an economic model in deciding on the optimum values of various treatment parameters. The design consraints are intended to ensure that an optimum design obtained from the model can be executed in the field using the specified surface equipment and without causing uncontrolled fracture growth and/or initiation of multiple secondary fractures. A pseudo-3D fracture model is improved to establish better compatibility relationships between reservoir properties, treatment parameters, and fracture growth. Features of Genetic Algorithm and Evolutionary Operation are combined to develop a robust optimization algorithm that executes complex interactions between the above-mentioned modules to achieve an optimum design. The capability and robustness of the proposed model is demonstrated by applications to a tight-gas reservoir. By trade-off analysis between production/NPV and treatment cost, it is shown that only a 13% sacrifice of the maximum possible production/net present value (NPV) over ten years can save up to 52% of the fracturing treatment cost. Various other design issues are also investigated by sensitivity analyses.

    Original languageBritish English
    Pages (from-to)1721-1758
    Number of pages38
    JournalPetroleum Science and Technology
    Volume21
    Issue number11-12
    DOIs
    StatePublished - 1 Nov 2003

    Fingerprint

    Dive into the research topics of 'Multicriteria hydraulic fracturing optimization for reservoir stimulation'. Together they form a unique fingerprint.

    Cite this