Uncertainties in quantitative time-lapse seismic analysis

Most seismic time-lapse studies so far have been of a qualitative nature. Identification of areas with minor or no seismic changes has been used to plan new infill drilling targets. Increased accuracy in seismic acquisition methods, in both conventional streamer surveys and newer methods such as multicomponent sea-bed seismic and permanent sensors, opens possibilities for the next step: quantitative time-lapse analysis. Quantitative methods here mean the estimation of, for instance, a change in fluid saturation from 20% water to 90% water or the estimation of a pore pressure change of 5 MPa. Explicit expressions for the uncertainties associated with estimated changes in, for instance, reservoir pressure and fluid saturation are derived. These formulae can be used to compare relative uncertainties between estimated parameters as well as to identify the critical factors in various estimation techniques. The importance of accurate rock physics input, as well as that of highly repeatable time-lapse seismic data, is emphasized. Furthermore, uncertainty analysis can be used to find optimal weight factors when the same parameter (e.g. saturation change) is estimated by two or three different techniques.