Study on Deconvolution Well Test Interpretation Model for Low-Permeability Oil Wells in Offshore Reservoir

2019 
In order to avoid reservoir damage and productivity impact caused by shut-in, some offshore low-permeability oil wells have a short shut-in test time and slow pressure recovery, which leads to unapparent or missing pressure characteristics in well test interpretation. In order to improve the accuracy of well test interpretation results, obtain more reservoir information, and solve the problem of missing radial flow characteristics and boundary features on conventional well test pressure and pressure derivative curves, one convolution and deconvolution well test model is established based on the dynamic variation law of oil wells’ bottom-hole flow pressure with variable production. Duhamel principle is used to obtain the equivalent constant flow pressure response for the entire production period, and the error caused by incomplete production or pressure history is reduced, which is verified by oil field examples. The result shows that the deconvolution well test interpretation model has the advantages that the conventional interpretation method does not have. By integrating the production and pressure information in the whole test process, it can eliminate the interference of wellbore storage effects on radial flow, enlarge the detection range of pressure fluctuation, update interpretation results of conventional well test and ensure consistent geological understanding with real reservoir information. The deconvolution model can also explain the distance between oil well and edge-bottom water invasion position. This study can interpret reservoir information that cannot be reflected by conventional well test, providing a new theoretical interpretation method for offshore well test interpretation with large production changes and short shut-in time. It is of great significance to provide important reservoir parameters and test well boundary types for reservoir development.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    10
    References
    0
    Citations
    NaN
    KQI
    []