Multicomponent distributed acoustic sensing: Concept and theory

Abstract

Distributed acoustic sensing (DAS) data are increasingly used in geophysics. Lower in cost and higher in spatial resolution, DAS data are appealing, especially in boreholes in which optical fibers are readily available. DAS has the potential to become a permanent reservoir monitoring tool with a reduced sensing time interval. To accomplish this goal, it is critical that DAS can record all wave modes to fully characterize reservoir properties. This goal can be achieved by recording the complete strain tensor consisting of 6C. Conventional DAS provides projections of these components along the optical fiber by observing deformation along the fiber. To obtain the entire 6C strain tensor, we have developed an approach using multiple strain projections measured along optical fibers with judiciously chosen geometry specifically. We evaluate designs combining multiple helical configurations or a single helical configuration together with a straight optical fiber that allow access to multiple strain projections. We group multiple strain projections in a given spatial window to perform reconstruction of the entire strain tensor in a least-squares sense under the assumption that the seismic wavelength is larger than the analysis window size. We determine how optimal optical fiber parameters can be selected using a scan of the entire configuration space and analyzing the condition number associated with the geometry of the optical fibers. We develop our method through synthetic experiments using realistic fiber geometry and wavefields of arbitrary complexity.