Distributed acoustic sensing (DAS) that uses optical fibres as sensing units is attracting increasing interest for micro-seismic monitoring in geothermal projects. Standard optical fibres provide one-component measurements along the fibre and this pose challenges in determining certain characteristics of the source, such as its azimuth and its full moment tensor. Full source characteristics can be obtained via offset downhole measurements and/or measurements from horizontal well sections but these come with substantial extra costs. This paper proposes a single-well dual-cable DAS configuration to reduce the need for drilling additional wells or sections, where two DAS cables are assumed to be positioned within a single vertical well at opposite sides of the well. Synthetic DAS signals are generated by an open-source code that assumes plane-layered media and are used to study the feasibility of the dual-cable DAS for localising a seismic source and resolving its moment tensor. A localisation procedure is presented, and a sensitivity analysis of localisation accuracy is conducted with respect to source parameters and noise levels. In addition, an analysis is performed to assess the resolvability of the moment tensor components from the dual-cable DAS configuration. Results suggest the source location can be fully determined, yet low signal-to-noise ratio and azimuth close to 0∘ (North, aligned with the two cables) lead to a decrease in accuracy. The full moment tensor can be resolved only if the epicentral distance is 5 m or less, while non-double-couple components can be reliably resolved with an epicentral distance up to 20 m, showing improvement compared to installations with a single cable. Consequently, near-borehole failures, regardless of the source mechanisms, can be characterised within an epicentral distance of 5 m. With epicentral distance increasing, resolvability of the mix-mode failures is reduced first, followed by the resolvability of the pure shear or tensile failures, which depends on the azimuth. Overall, the results demonstrate that a single-well dual-cable configuration has the potential for monitoring and understanding near-borehole micro-seismic events induced during geothermal reinjection and stimulation operations.

Distributed acoustic sensing has been limited in its use for surface-seismic reflection measurements due to the fiber’s decreased broadside sensitivity when the fiber is deployed horizontally. Deploying the fiber in a helically wound fashion has the promise of being more sensitive to broadside waves (e.g., P-wave reflections) and less sensitive to surface waves than a straight fiber (SF). We examine such claims and compare the responses of SFs and helically wound fibers (HWFs) with different wrapping angles, using standard and engineered fibers. These fibers have been buried in a 2 m deep trench in a farmland in the province of Groningen in The Netherlands, where we performed an active-source survey. We observe in our field data that using HWF has a destructive effect on the surface-wave amplitudes. Our data confirms the effect of the wrapping angle on the polarity of the surface-wave arrival and the dampening effect of the helical winding, behaving in quite a predictable fashion. Apart from the effect of the wrapping angle, the different design choices, e.g., cable filling and material type, do not show a significant effect on the amplitude of the signals. As for P-wave reflections, we observe that engineered SF and HWF provide reflection images comparable with those obtained from simultaneously deployed geophones at the surface despite the SF’s decreased broadside sensitivity. A polarity reversal and an amplitude difference between the SF and HWFs are observed. Finally, we demonstrate that the combined use of SF and HWF proved to be useful because SF showed better sensitivity in the shallower part and HWF in the deeper part.

Retrieving accurate microseismic source locations induced by hydraulic-fracturing operations is an important step to gain insights into the hydraulically stimulated reservoir volume. Recently, deep neural networks have been proposed that directly recover source locations from the seismic waveforms. The optimal performance of the proposed deep neural networks usually requires large training sets. The need for a large training set can be circumvented if a previously trained deep neural network can be used to start the training process with its weights instead of randomly initialized weights. These weights can then be fine-tuned using a smaller training set, which is also known as transfer learning. In this work, we implement a transfer learning workflow to update the weights of a deep neural network that was initially trained on a large synthetic dataset to localize microseismic events. We present two methods of processing, namely one post-monitoring mode and one continuous mode where the processing takes place during the monitoring period. We apply the methods to field data from a hydraulic fracturing site in Texas, USA. In the first scenario, a subset of the field data from the entire monitoring period is used to update the weights of the deep neural network, which is then applied to the remaining data resulting in mean and median distances of 227 and 182 m, respectively, compared to the results of a good localization method. In the second scenario, the deep neural network is updated daily with previously detected and located events and applied to the events detected the following day. Since the observed data used for training generally do not cover a wide range of source locations, we enrich the training set with synthetic data. The addition of synthetics for transfer learning ensures that the updated deep neural network provides accurate source locations for events with locations far from locations used during transfer learning. Transfer learning combining synthetic and real data performs significantly better (more consistent) locations than transfer learning without synthetics.