Estimating Flow Properties from the Onset of Time-Lapse Changes

Abstract:

Geophysical time-lapse observations are increasingly important for monitoring subsurface fluid flow.
Time-lapse data can even be used to characterize spatial variations in the properties of a porous medium.
A major impediment to such characterization is the difficulty in connecting flow-related changes to changes geophysical properties.
For example, physical models used to relate changes in fluid saturation and pressure to seismic velocity changes
often depend upon unknown parameters, or on the detailed distribution of the fluid at intermediate spatial scales.
The challange is particularly acute when one tries to relate the magnitudes of saturation and pressure changes to
the magnitude of a change in a time-lapse observation.
I present an alternative approach for the characterization of a porous medium, based upon the onset of changes
of a geophysical attribute, that is applicable when there are a sequence of geophysical surveys.
An onset time is the calander time as which a geophysical observable begins to deviate from its initial or background value.
In many cases onset times are sensitive to flow properties, such as permeability, and insensitive to the
details of the physical model governing the geophysical response.
Several examples of the use of onset times will be presented, including the seismic monitoring of
injected carbon dioxide and the use of surface deformation data to image fluid flow at depth.
Though the technique works best when there are numerous geophysical snapshots, numerical modeling
indicates useful results are possible from yearly seismic surveys.