Extended Internal Multiple Prediction

Ensured integrity of prestack inversion and reservoir characterization analyses

Removes multiples while protecting primaries

Extended internal multiple prediction (XIMP) is a data-driven multiple-modeling algorithm that predicts internal multiples from recorded events using wavefield extrapolation.

The method does not require a prior velocity function to differentiate between primaries and internal multiples.

XIMP uses the onset times of the multiple-generating events in the seismic data to provide an accurate model of the multiple energy that can then be removed using adaptive subtraction.

Seismic section and acoustic impedance inversion examples after multiple attenuation with 3D GSMP algorithm and XIMP.

XIMP is applicable in any geologic setting with three modes of operation based on available geologic information.

Horizon-based mode

Recommended for known or shallow internal multiple generators

Horizon-guided, layer-based mode

Recommended for thin, structurally variant rock formations

Layer-based mode

Prediction of all internal multiples generated within a given time interval

after multiple attenuation using 3D GSMP general surface multiple prediction

after multiple attenuation using both 3D GSMP prediction and XIMP

Predicts internal multiples at true azimuth

In areas of moderate structural complexity, XIMP can handle any acquisition geometry and predict multiples at true azimuth. It is applied fully in 3D and works with both land and marine datasets. The workflow addresses challenging internal multiples that are difficult to attenuate while also preserving primaries.

Move the slider to see poststack time migration after multiple attenuation using 3D GSMP algorithm (left) and after multiple attenuation using both 3D GSMP prediction and XIMP (right). (Data examples courtesy of Kuwait Oil Company.)