Denoising Deep Learning Network Based on Singular Spectrum Analysis—DAS Seismic Data Denoising With Multichannel SVDDCNN

Distributed Acoustic Sensing (DAS) is a Fiber Optic (FO) cable based technology which is gaining importance for VSP surveys, especially for time-lapse monitoring of reservoirs. DAS offers distinct advantages over geophones but it also poses unique challenges, including receiver depth uncertainty and low signal-to-noise ratio. Here, we present an analysis of a DAS VSP acquired for monitoring purposes in a treatment well while a multi-stage hydraulic fracture treatment was ongoing in the same well. We describe the pre-processing needed before further time-lapse analysis of the VSP. We focus on two key steps - depth calibration and noise removal. Large spike noise, most likely from the DAS acquisition system, appears to be a major source of noise in the raw DAS data, but can be removed with careful processing. Other sources of noise include temperature fluctuations in the well and optical noise. We show that excellent quality VSP data can be obtained in a hydrofrac treatment well.

Summary Distributed acoustic-sensing technology is a diagnostic method that has been implemented in the oil and gas industry for flow monitoring during injection or production. One of the applications is to estimate fluid distribution during hydraulic fracturing treatment. This is an invaluable diagnostic tool for multistage fracture treatments because of the large number of parameters and high uncertainty involved in these fracture treatments. Distributed acoustic sensor (DAS) measurements are based on data extracted from a fiber-optic cable installed in a wellbore. Fiber-optic cable strain is sensitive to temperature and acoustic variations induced by fluid flow with different fluid properties and fracture geometries. These parameters are evaluated from the backscattered laser pulse through the fiber using coherent Rayleigh backscattering for DAS and Raman backscattering for distributed temperature sensing (DTS). The DAS interrogation system acquires strain response to acoustic variations along the measured depth of the wellbore for all periods of time. This study presents a method of interpretation of flow-rate distribution from acoustic signals from DAS measurements. Raw acoustic measurements are transformed into energy-attribute distribution for all necessary depth locations and time frame. It allows calculating energy for each perforation cluster location. This calculation demands proper depth-interval consideration for each perforation cluster. A depth-windowing procedure allowing location of intervals where DAS channel measurements must be integrated for each timestep is applied to all stages. Based on previous experimental and computational fluid-dynamic investigations, the correlation between acoustic signal and flow rate is applied to interpret the measured DAS data to flow distribution. In this paper, the system of equations that connects acoustic energy response with clusters flow-rate distribution during the injection period using these correlations is introduced. The solution of this system allows calculating cumulative volume for each perforation cluster on each timestep and at the end of fracture treatment. As additional verification of flow-rate distribution, the results of DAS interpretation were compared with the interpretation results of DTS. A field example is used to illustrate the interpretation procedure. From this work, it is concluded that, besides qualitative analysis, the DAS interpretation method provides a quantitative estimation of flow distribution. Based on current assumptions, the interpretation results from DAS and DTS are comparable with a satisfactory agreement. The combined DAS and DTS interpretations help in understanding cluster efficiency in multistage fracture treatments.

The industry treats the distributed acoustic sensing (DAS) system, which uses an optical fiber cable in vertical seismic profiling (VSP) data acquisition, as a new and fast-growing technology. The high-quality data set acquired from the DAS acquisition system can produce high-precision VSP images and obtain more detailed checkshots. However, in field data, the acquired VSP data set suffers from strong coherent DAS coupling noise. Many factors may cause coherent DAS coupling noise, such as the cable slapping and ringing due to the physical placement, the regular swing of the wireline in the well, and the uncoupling between the fiber cable and the casing. This DAS coupling noise reduces the signal-to-noise ratio and affects the subsequent processing and interpretation. Removing the DAS coupling noise can help to improve the quality of the VSP data set acquired with the DAS system. We have developed a sparse-optimization-based DAS coupling noise removing method. In the DAS-based VSP data set, the effective signal and the coupling noise have distinct morphological characteristics. The effective VSP signal has a wide bandwidth, whereas the DAS coupling noise appears in some narrow frequency bands in the frequency domain. The continuous wavelet transform and the discrete cosine transform can sparsely represent the effective VSP signal and DAS coupling noise, respectively. Therefore, we choose these two transforms as two sparse dictionaries and combine them to form an overcomplete dictionary. The morphological component analysis (MCA) can use the morphological difference between different components and the overcomplete dictionary to sparsely represent all components in the complicated signal. Based on the MCA theory, we use the block coordinate relaxation algorithm to separate the effective VSP signal and DAS coupling noise. Applications on a synthetic data set and two field data sets have validated the effectiveness of our method.

In all carbon capture, utilization, and storage (CCUS) projects, monitoring, verification, and accounting is required to span short to long periods to ensure plume extents and geologic integrity meet operational and regulatory objectives. Fiber-optic cables permanently deployed in the completions of injector wells enable time-lapse (4D) vertical seismic profile (VSP) imaging using distributed acoustic sensing (DAS). Most CCUS projects are onshore, and the majority of those utilize supercritical CO2 flooding for enhanced oil recovery (EOR). In these cases, fiber-optic installations, acquisition, and analysis are relatively straightforward. Offshore transport and storage projects, whether for EOR or sequestration in saline aquifers, may utilize subsea wells wherein fiber-optic deployment is not as straightforward. Unlike dry-tree wells, DAS of subsea wells requires advanced optical engineering solutions to compensate for the reduced acoustic bandwidth, optical losses, and back reflections accumulated through umbilicals, multiple wet- and dry-mate optical connectors, splices, optical feedthrough systems, and downhole fibers. In this paper, we simulate VSP shot records of a carbon storage site for a variety of proposed subsea DAS concepts and evaluate the 3D and 4D seismic image quality of each using Kirchhoff migration of upgoing seismic wavefields with their expected signal-to-noise ratios. We normalize interrogator performance, so we are evaluating the sensing topology only. We show that a subsea fiber topology consisting of dual transmission fibers, a remote circulator with selective amplification of the backscattered light, and enhanced backscatter sensing fiber results in high-quality seismic images suitable for both 3D and 4D VSPs of carbon storage.

Seismic surveys are widely used to characterise the properties of glaciers, their basal material and conditions, and ice dynamics. The emerging technology of Distributed Acoustic Sensing (DAS) uses fibre optic cables as seismic sensors, allowing observations to be made at higher spatial resolution than possible using traditional geophone deployments. Passive DAS surveys generate large data volumes from which the rate of occurrence and failure mechanism of ice quakes can be constrained, but such large datasets are computationally expensive and time consuming to analyse. Machine learning tools can provide an effective means of automatically identifying seismic events within the data set, avoiding a bottleneck in the data analysis process.Here, we present a novel approach to machine learning for a borehole-deployed DAS system on Store Glacier, West Greenland. Data were acquired in July 2019, as part of the RESPONDER project, using a Silixa iDAS interrogator and a BRUsens fibre optic cable installed in a 1043 m-deep borehole. The data set includes controlled-source vertical seismic profiles (VSPs) and a 3-day passive record of cryoseismicity.  To identify seismic events in this record, we used a convolutional neural network (CNN). A CNN is a deep learning algorithm and a powerful classification tool, widely applied to the analysis of images and time series data, i.e. to recognise seismic phases for long-range earthquake detection.For the Store Glacier data set, a CNN was trained on hand-labelled, uniformly-sized time-windows of data, focusing initially on the high-signal-to-noise-ratio seismic arrivals in the VSPs. The trained CNN achieved an accuracy of 90% in recognising seismic energy in new windows. However, the computational time taken for training proved impractical. Training a CNN instead to identify events in the frequency-wavenumber (f-k) domain both reduced the size of each data sample by a factor of 340, yet still provided accurate classification. This decrease in input data volume yields a dramatic decrease in the time required for detection. The CNN required only 1.2 s, with an additional 5.6 s to implement the f-k transform, to process 30 s of data, compared with 129 s to process the same data in the time domain. This suggests that f-k approaches have potential for real-time DAS applications.Continuing analysis will assess the temporal distribution of passively recorded seismicity over the 3 days of data. Beyond this current phase of work, estimated source locations and focal mechanisms of detected events could be used to provide information on basal conditions, internal deformation and crevasse formation. These new seismic observations will help further constrain the ice dynamics and hydrological properties of Store Glacier that have been observed in previous studies of the area.The efficiency of training a CNN for event identification in the f-k domain allows detailed insight to be made into the origins and style of glacier seismicity, facilitating further development to passive DAS instrumentation and its applications.

Seismic data are a significant facilitator for monitoring in carbon capture and sequestration projects, providing high-resolution images of fluid migration, using, for example, full-waveform inversion (FWI). Distributed acoustic sensing (DAS), a relatively novel technology for wavefield sampling, is well suited for this type of monitoring. Using noninvasive optical fibers, DAS allows for dense spatial sampling along the entire length of the wellbore, without disrupting operations. Permanently installed in the wellbore, typically behind casing, DAS offers highly repeatable and dense sampling of the transmitted wave modes crucial to seismic monitoring of injected carbon dioxide ([Formula: see text]). However, the DAS data consist of measurements of strain along the tangent of the fiber and therefore do not transfer directly to conventional FWI algorithms. Incorporation of DAS data in their native strain (or strain-rate) form in standard FWI algorithms, requires changing the definition of the receiver sampling operator to use geometric information about the fiber to supply tangential strain measurements to the FWI residual. The theoretical developments are applied to invert field vertical seismic profile data acquired with DAS fiber and accelerometers at a [Formula: see text] sequestration site in Newell Country, Alberta. Our method incorporates DAS data and accelerometer data in one objective function and allows us to tune the relative importance we wish to place on each data set. This method also transfers to noncollocated sensors, for example, surface-deployed geophones and borehole fiber. The inverted models contain features expected from the geology of the field site, and data modeled in the inverted models compare favorably with the field data for these sensor types. The models are derived from data acquired prior to [Formula: see text] injection, representing baseline models for future time-lapse studies planned at the field research station.

The advent of fiber optic technology in geophysics exploration has grown in its use in the exploration, production, and monitoring of subsurface environments, revolutionizing the way data are gathered and interpreted critically to speed up decision-making and reduce expense and time. Distributed Acoustic Sensing (DAS) has been increasingly utilized to build relationships in complex geophysics environments by utilizing continuous measurement along fiber optic cables with high spatial resolution and a frequency response of up to 10 KHz. DAS, as fiber optic technology examining backscattered light from a laser emitted inside the fiber and measuring strain changes, enables the performance of subsurface imaging in terms of real-time monitoring for Vertical Seismic Profiling (VSP), reservoir monitoring, and microseismic event detection. This review examines the most widely used fiber optic cables employed for DAS acquisition, namely Single-Mode Fiber (SMF) and Multi-Mode Fiber (MMF), with the different deployments and scopes of data used in geophysics exploration. Over the years, SMF has emerged as a preferred type of fiber optic cable utilized for DAS acquisition and, in most applications examined in this review, outperformed MMF. On the other side, MMF has proven to be preferable when used to measure distributed temperature. Finally, the fiber optic cable deployment technique and acquisition parameters constitute a pivotal preliminary step in DAS data preprocessing, offering a pathway to improve imaging resolution based on DAS measurement as a future scope of work.

Distributed acoustic sensing (DAS) is a new seismic monitoring technology. DAS generates a large amount of data, necessitating the development of new technologies to allow for cost-effective processing and handling. The raw seismic data is noisy and must be processed. The curvelet transform is an excellent choice for processing seismic data due to its localized nature, as well as its frequency and dip characteristics. However, its capabilities are limited in case of noise other than white. This paper proposes a denoising method based on a combination of the curvelet transform and a whitening filter, as well as a procedure for estimating noise variance. The whitening filter is included to improve the performance of the curvelet transform in both coherent and incoherent noise cases, as well as to simplify the noise variance estimation method and make it easier to use standard threshold methodology without delving into the curvelet domain. Two data sets are used to validate the suggested technique. Pseudo-synthetic data set created by adding noise to the actual noise-free data collection from the Netherlands offshore F3 block and the on-site data set (with ground roll noise) from east Texas, USA. Experimental results demonstrate that the proposed algorithm achieves the best results under various types of noise.

Summary Distributed acoustic sensing (DAS) technology offers the time-lapse capability of monitoring hydraulic fracturing and reservoir depletion processes. This paper uses a model-based study to explore the feasibility of using DAS vertical seismic profiling (VSP) to detect and characterize fracture-induced azimuthal variation in P-wave reflection amplitude. The study particularly focuses on seismic amplitude vs. offset and azimuth (AVOAz) to answer two questions. First, which fracture-induced seismic responses can be detectable by a three-dimensional (3D) DAS VSP survey. Second, how shots can be arranged to detect such seismic responses and image the extent of the fractures. Waveform data were simulated for two types of fractured reservoir models using full-elastic anisotropic algorithms. P-wave reflection amplitude was mapped in incidence angle intervals for the top of each reservoir model. Analyses of the amplitude variation with AVOAz revealed distinctive patterns for the two fracture models. These patterns of amplitude variation can be used to detect fractures, if the fractures have uniform density and similar orientation around the well.

Distributed acoustic sensing (DAS) is a rapidly evolving fiber-optic sensing technology that converts any standard fiber-optic cable into a distributed sensor capable of measuring acoustic events every meter over large (100 km+) distances. We describe a conveyance-agnostic DAS system for vertical seismic profiling (VSP) that has the capability to sense either single- or multi-mode sensing fibers, and encode all auxiliary seismic source signals directly onto the optical data stream using a series of piezo-electric fiber stretchers in-line with the sensing fiber. The DAS system can operate with latency-free synchronization via wireless telemetry between the seismic source controller and the DAS unit in either master or slave modes. Similar to geophone-based VSP data acquisition systems, the time-break signal is used to trigger data acquisition to ensure precise time synchronization at sub-millisecond accuracy between the DAS system and the seismic source. The system enables effective data management and the real-time generation of 1 millisecond sampled seismic and navigation data that conforms to the industry-standard seismic data formats, thus eliminating any latency between acquisition and field data deliverables.

Hydraulic fracture height and width are key parameters for completion design and evaluation of unconventional resources. Traditional measurement technologies like microseismic, tilt-meters, chemical or radioactive tracers, pressure temperature gauges, etc. either have low resolution or rely on sensitive models which can cause a high degree of uncertainty. Recent frac-hit measurement methods include the use of Distributed Acoustic Sensing (DAS) deployed in far-field wellbores offset from the treatment well. The DAS system directly measures the fracture propagation every second through the completion, with a few meters' spatial resolution sampled at 0.25-meter intervals along the monitored fiber well. Cross-well fiber optic monitored far-field strain (FFS) results suggested elastic stress effects, intense inelastic fracture expansion and closure events which provide identification and measurement of frac height on a TVD plot, as well as width by measured depth along the wellbore laterals. Vertical wells or the heel-section of horizontal wells are suitable for frac-hit height (FHH) measurement; fracture azimuth and wellbore geometries need to be considered for precise evaluation. A specially designed engineered constellation fiber cable was tested and utilized in this method in combination with a true phase coherence DAS interrogator with a 20 dB improved sensitivity (Signal-to-Noise-Ratio (SNR)) for both low and high frequency ranges DAS. The optic fiber can be either permanently installed outside the casing or temporarily deployed inside a monitor well. Comparable results can be achieved by the engineered fiber system and have been presented within case studies for both horizontal and vertical well sections. In addition, Distributed Temperature Sensing (DTS) and data from downhole gauge can confirm any temperature or pressure changes resulting from frac driven interactions (FDI). With this approach, fracture azimuth, frac-hit corridor (FHC) width and FHH can be determined with a high degree of accuracy and resolution. Completion engineers were able to optimize frac models in real-time and further change completion schedules during the frac treatment.

Distributed acoustic sensing (DAS) technology using fiber-optic cables offers a cost-effective and high-density approach for recording earthquake signals, particularly compared to conventional receivers. However, DAS recordings are often significantly degraded by strong ambient noise. Therefore, enhancing the signal-to-noise ratio of DAS data is crucial. We develop a data-driven solution, i.e., the moving-deep-learning-filtering (MDLF) framework, which uses learnable filtering coefficients (weights and biases of the neural network) to effectively extract weak DAS signals from significant noise. This MDLF framework incorporates a designed convolutional attention neural network (CANN) optimized with a synthetic data set, and a hybrid loss function consisting of a data distribution loss and a signal structure loss. The trained CANN within MDLF operates on DAS patches extracted using overlapping moving windows, allowing for direct generalization to field DAS data. Experiments on the synthetic data set, as well as two field DAS data sets from the Frontier Observatory for Research in Geothermal Energy and the San Andreas Fault Observatory at Depth projects, demonstrate the effectiveness of our MDLF method, which outperforms benchmark approaches.

HighlightsWhat are the main findings?The Distributed Acoustic Sensing (DAS) response is strongly influenced by the cable’s structural configuration, with the Microcable showing the highest strain sensitivity (≈0.025 a.u.) compared to the Duct and Anti-Rodent cables tested under identical laboratory conditions.Simplified and compact cable designs enable more effective strain transfer to the optical fiber, whereas multilayered or armored constructions significantly attenuate the transmitted strain signal.What are the implications of the main findings?The structural configuration of telecommunication cables should be carefully considered when repurposing them for DAS applications, as it directly impacts sensing efficiency and spatial response.Microcables are ideal for indoor or duct-protected environments requiring high sensitivity, while armored cables remain preferable for long-term or harsh-environment installations prioritizing durability.Distributed Acoustic Sensing (DAS) transforms conventional optical fibres into large-scale acoustic sensor arrays. While existing telecommunication cables are increasingly considered for DAS-based monitoring, their performance depends strongly on cable construction and strain transfer efficiency. In this study, the relative DAS signal amplitudes of three commercial telecommunication optical cables were experimentally compared using a benchtop Rayleigh backscattering-based interrogator under controlled laboratory conditions. By maintaining a constant temperature and ensuring no additional strain changes from the outside environment, we guaranteed that only strain-induced variations from acoustic excitations were measured. The results show clear differences in signal amplitude and signal-to-noise ratio (SNR) among the tested cables. The Microcable consistently produced the highest spatial peak amplitude (up to 0.029 a.u.) and SNR (up to 79), while the Duct cable reached 0.00268 a.u. with mean SNR ≈ 32. The Anti-Rodent cable showed low signal amplitude (0.0018 a.u.) but exhibited a high mean SNR (≈111) driven by an exceptional low noise floor in one of the runs. These findings reflect the variations in mechanical coupling between the fibre core and external perturbations and provide practical insights into the suitability of different telecom cable types for DAS applications, supporting informed choices for future deployments.

A complete comparison is made between a rig-source vertical seismic profile (VSP) recorded in a deviated well performed with conventional geophones data and with distributed acoustic sensing (DAS) fiber optics. The DAS acquisition was performed using a hybrid wireline cable during the logging descent of an ultrasonic tool for casing evaluation. This rig-source VSP project offers two rich data sets for which a one-to-one comparison of the most common quality indicators in borehole seismic can be made. The objective is to highlight commonalities, advantages, and potential shortcomings of both acquisitions when performed in a real oilfield operating environment. The comparison herein presented can be of interest to the industry, specifically looking at the overall cost reduction for gathering the well seismic information, versus the value of the information that can be achieved with both technologies.

This research presents a development of the affine projection algorithm (APA) in voice communication applications. A method of controlling the parameters of the APA is devised to improve the performance in cancelling various types of ambient noise that could possibly corrupt speech signals in voice communication systems. Indicators are used to identify the type of noise accompanying the target signal. Then the corrupted signal is processed in a noise a cancellation setup in such a way that three parameters of algorithm are changed according to the nature of the noise. The spreading of elements in the covariance matrix of the noise is used as an indicator for the type of noise so that the projection order, step-size and filter length are changed at the same time. This way the performance of the canceller is improved rendering lower estimation error with a moderate computational power. The method was tested under various types of noise and showed better convergence performance than the original APA as well as other commonly used algorithms in noise cancellation systems. The MSE of the proposed VPAPA method drops to -65 dB in steady state compared to -20 dB using NLMS and just below -30 dB using standard APA with projection order of 8, while the powerful RLS reaches around -60dB under the same environment. The method can be useful for clearer voice communication in variable environmental noise.