Strain microseismics: Radiation patterns, synthetics, and moment tensor resolvability with distributed acoustic sensing in isotropic media

Author(s):

Ismael Vera Rodriguez, NORSAR

Andreas Wuestefeld, NORSAR

Year:2020
Type:Journal
ISBN / DOI:10.1190/geo2019-0373.1

We have derived analytical formulations for the strain field produced by a moment tensor source in homogeneous isotropic media. Such formulations are important for microseismic projects that increasingly are monitored with fiber-optic distributed acoustic sensing (DAS) systems. We find that the spatial derivative of displacement produces new terms in strain proportional to 1/rn1/rn with n2n≥2. In viscoelastic media, the derivative also produces an additional far-field term that is scaled by a frequency-dependent factor. When comparing with full wavefield synthetic data, we observe that the new terms proportional to 1/r21/r2 can be considered part of a near-field in strain, similar to the practice with the displacement formulation. Analyses of moment tensor resolvability show that full moment tensors are resolvable with P-wave information from two or more noncoplanar vertical DAS cable geometries if intermediate- and far-field terms are considered and that S-wave information alone cannot constrain full moment tensors using only vertical wells. These results mirror previous observations made with displacement measurements. Furthermore, the addition of the new terms proportional to 1/r21/r2 in strain improves the moment tensor resolvability but only in the case of a single vertical array. In the case of a single deviated/horizontal well, we can, in theory, resolve a full moment tensor but a case-by-case analysis is necessary to identify regions of full constraint around the well and the necessary noise conditions to guarantee reliable solutions. Real DAS measurements also are affected by the gauge length and interrogator details. In the case of the gauge length, we observe that this operator does not change the resolvability of the problem but it does affect inversion stability. The results derived here represent theoretical limits or in some cases specific examples. Practical implementations require analyses of conditioning, noise, coupling, and the effect of gauge length on a case-by-case basis.

http://pub/_PDFs/VERARODRIGUEZ/2020/4348_VeraRodriguez+Wuestefeld2020.pdf

Gunnar Randers vei 15
N-2007 Kjeller, Norway
MVA: 974374765

  • +47 63 80 59 00