Retrieving virtual reflection responses at drill‐bit positions using seismic interferometry with drill‐bit noise

In the field of seismic interferometry, researchers have retrieved surface waves and body waves by cross‐correlating recordings of uncorrelated noise sources to extract useful subsurface information. The retrieved wavefields in most applications are between receivers. When the positions of the noise sources are known, inter‐source interferometry can be applied to retrieve the wavefields between sources, thus turning sources into virtual receivers. Previous applications of this form of interferometry assume impulsive point sources or transient sources with similar signatures. We investigate the requirements of applying inter‐source seismic interferometry using non‐transient noise sources with known positions to retrieve reflection responses at those positions and show the results using synthetic drilling noise as source. We show that, if pilot signals (estimates of the drill‐bit signals) are not available, it is required that the drill‐bit signals are the same and that the phases of the virtual reflections at drill‐bit positions can be retrieved by deconvolution interferometry or by cross‐coherence interferometry. Further, for this case, classic interferometry by cross‐correlation can be used if the source power spectrum can be estimated. If pilot signals are available, virtual reflection responses can be obtained by first using standard seismic‐while‐drilling processing techniques such as pilot cross‐correlation and pilot deconvolution to remove the drill‐bit signatures in the data and then applying cross‐correlation interferometry. Therefore, provided that pilot signals are reliable, drill‐bit data can be redatumed from surface to borehole depths using this inter‐source interferometry approach without any velocity information of the medium, and we show that a well‐positioned image below the borehole can be obtained using interferometrically redatumed reflection responses with just a simple velocity model. We discuss some of the practical hurdles that restrict the application of the proposed method offshore.     

Seismic interferometry by tangent‐phase correction

We present a modified interferometry method based on local tangent‐phase analysis, which corrects the cross‐correlated data before summation. The approach makes it possible to synthesize virtual signals usually vanishing in the conventional seismic interferometry summation. For a given pair of receivers and a set of different source positions, a plurality of virtual traces is obtained at new stationary projected points located along the signal wavefronts passing through the real reference receiver. The position of the projected points is estimated by minimizing travel times using wavefront constraint and correlation‐signal tangent information. The method uses mixed processing, which is partially based on velocity‐model knowledge and on data‐based blind interferometry. The approach can be used for selected events, including reflections with different stationary conditions and projected points with respect to those of the direct arrivals, to extend the interferometry representation in seismic exploration data where conventional illumination coverage is not sufficient to obtain the stationary‐phase condition. We discuss possible applications in crosswell geometry with a velocity anomaly and a time lapse.     

Seismic interferometry,intrinsic losses and Q‐estimation*

Seismic interferometry is the process of generating new seismic traces from the cross‐correlation, convolution or deconvolution of existing traces. One of the starting assumptions for deriving the representations for seismic interferometry by cross‐correlation is that there is no intrinsic loss in the medium where the recordings are performed. In practice, this condition is not always met. Here, we investigate the effect of intrinsic losses in the medium on the results retrieved from seismic interferometry by cross‐correlation. First, we show results from a laboratory experiment in a homogeneous sand chamber with strong losses. Then, using numerical modelling results, we show that in the case of a lossy medium ghost reflections will appear in the cross‐correlation result when internal multiple scattering occurs. We also show that if a loss compensation is applied to the traces to be correlated, these ghosts in the retrieved result can be weakened, can disappear, or can reverse their polarity. This compensation process can be used to estimate the quality factor in the medium.     

Retrieval of super‐virtual refraction by cross‐correlation

Topography and severe variations of near‐surface layers lead to travel‐time perturbations for the events in seismic exploration. Usually, these perturbations could be estimated and eliminated by refraction technology. The virtual refraction method is a relatively new technique for retrieval of refraction information from seismic records contaminated by noise. Based on the virtual refraction, this paper proposes super‐virtual refraction interferometry by cross‐correlation to retrieve refraction wavefields by summing the cross‐correlation of raw refraction wavefields and virtual refraction wavefields over all receivers located outside the retrieved source and receiver pair. This method can enhance refraction signal gradually as the source–receiver offset decreases. For further enhancement of refracted waves, a scheme of hybrid virtual refraction wavefields is applied by stacking of correlation‐type and convolution‐type super‐virtual refractions. Our new method does not need any information about the near‐surface velocity model, which can solve the problem of directly unmeasured virtual refraction energy from the virtual source at the surface, and extend the acquisition aperture to its maximum extent in raw seismic records. It can also reduce random noise influence in raw seismic records effectively and improve refracted waves’ signal‐to‐noise ratio by a factor proportional to the square root of the number of receivers positioned at stationary‐phase points, based on the improvement of virtual refraction's signal‐to‐noise ratio. Using results from synthetic and field data, we show that our new method is effective to retrieve refraction information from raw seismic records and improve the accuracy of first‐arrival picks.     

Synthesis of a seismic virtual reflector*

We describe a method to process the seismic data generated by a plurality of sources and registered by an appropriate distribution of receivers, which provides new seismic signals as if in the position of the receivers (or sources) there was an ideal reflector, even if this reflector is not present there. The data provided by this method represent the signals of a virtual reflector. The proposed algorithm performs the convolution and the subsequent sum of the real traces without needing subsurface model information. The approach can be used in combination with seismic interferometry to separate wavefields and process the reflection events. The application is described with synthetic examples, including stationary phase analysis and with real data in which the virtual reflector signal can be appreciated.     

Seismic interferometry experiment in a shallow cased borehole using a seismic vibrator source‡

We present the results of a seismic interferometry experiment in a shallow cased borehole. The experiment is an initial study for subsequent borehole seismic surveys in an instrumented well site, where we plan to test other surface/borehole seismic techniques. The purpose of this application is to improve the knowledge of the reflectivity sequence and to verify the potential of the seismic interferometry approach to retrieve high‐frequency signals in the single well geometry, overcoming the loss and attenuation effects introduced by the overburden. We used a walkaway vertical seismic profile (VSP) geometry with a seismic vibrator to generate polarized vertical and horizontal components along a surface seismic line and an array of 3C geophones cemented outside the casing. The recorded traces are processed to obtain virtual sources in the borehole and to simulate single‐well gathers with a variable source‐receiver offset in the vertical array. We compare the results obtained by processing the field data with synthetic signals calculated by numerical simulation and analyse the signal bandwidth and amplitude versus offset to evaluate near‐field effects in the virtual signals. The application provides direct and reflected signals with improved bandwidth after vibrator signal deconvolution. Clear reflections are detected in the virtual seismic sections in agreement with the geology and other surface and borehole seismic data recorded with conventional seismic exploration techniques.     

Application of the virtual refraction to near‐surface characterization at the Boise Hydrogeophysical Research Site‡

Seismic interferometry is a relatively new technique to estimate the Green's function between receivers. Spurious energy, not part of the true Green's function, is produced because assumptions are commonly violated when applying seismic interferometry to field data. Instead of attempting to suppress all spurious energy, we show how spurious energy associated with refractions contains information about the subsurface in field data collected at the Boise Hydrogeophysical Research Site. By forming a virtual shot record we suppress uncorrelated noise and produce a virtual refraction that intercepts zero offset at zero time. These two features make the virtual refraction easy to pick, providing an estimate of refractor velocity. To obtain the physical parameters of the layer above the refractor we analyse the cross‐correlation of wavefields recorded at two receivers for all sources. A stationary‐phase point associated with the correlation between the reflected wave and refracted wave from the interface identifies the critical offset. By combining information from the virtual shot record, the correlation gather and the real shot record we determine the seismic velocities of the unsaturated and saturated sands, as well as the variable relative depth to the water‐table. Finally, we discuss how this method can be extended to more complex geologic models.     

Passive seismic data primary estimation and noise removal via focal‐denoising closed‐loop surface‐related multiple elimination based on 3D L1‐norm sparse inversion

Passive seismic has recently attracted a great deal of attention because non‐artificial source is used in subsurface imaging. The utilization of passive source is low cost compared with artificial‐source exploration. In general, constructing virtual shot gathers by using cross‐correlation is a preliminary step in passive seismic data processing, which provides the basis for applying conventional seismic processing methods. However, the subsurface structure is not uniformly illuminated by passive sources, which leads to that the ray path of passive seismic does not fit the hyperbolic hypothesis. Thereby, travel time is incorrect in the virtual shot gathers. Besides, the cross‐correlation results are contaminated by incoherent noise since the passive sources are always natural. Such noise is kinematically similar to seismic events and challenging to be attenuated, which will inevitably reduce the accuracy in the subsequent process. Although primary estimation for transient‐source seismic data has already been proposed, it is not feasible to noise‐source seismic data due to the incoherent noise. To overcome the above problems, we proposed to combine focal transform and local similarity into a highly integrated operator and then added it into the closed‐loop surface‐related multiple elimination based on the 3D L1‐norm sparse inversion framework. Results proved that the method was capable of reliably estimating noise‐free primaries and correcting travel time at far offsets for a foresaid virtual shot gathers in a simultaneous closed‐loop inversion manner.     

Interferometric interpolation of sparse marine data

We present the theory and numerical results for interferometrically interpolating 2D and 3D marine surface seismic profiles data. For the interpolation of seismic data we use the combination of a recorded Green's function and a model‐based Green's function for a water‐layer model. Synthetic (2D and 3D) and field (2D) results show that the seismic data with sparse receiver intervals can be accurately interpolated to smaller intervals using multiples in the data. An up‐ and downgoing separation of both recorded and model‐based Green's functions can help in minimizing artefacts in a virtual shot gather. If the up‐ and downgoing separation is not possible, noticeable artefacts will be generated in the virtual shot gather. As a partial remedy we iteratively use a non‐stationary 1D multi‐channel matching filter with the interpolated data. Results suggest that a sparse marine seismic survey can yield more information about reflectors if traces are interpolated by interferometry. Comparing our results to those of f‐k interpolation shows that the synthetic example gives comparable results while the field example shows better interpolation quality for the interferometric method.     

Seismic preprocessing and amplitude cross-calibration for a time-lapse amplitude study on seismic data from the Oseberg reservoir

The cross‐calibration of different vintage data is an important prerequisite in attempting to determine the time‐lapse seismic effects induced by hydrocarbon production in a reservoir. This paper reports the preprocessing and cross‐calibration procedures adopted to modify the data of four seismic vintages (1982, 1989, 1992 and 1999) from the Oseberg field in the North Sea, for optimal conditions for a time‐lapse seismic amplitude analysis. The final results, in terms of time‐lapse variations, of acoustic impedance and of amplitude‐versus‐offset, are illustrated for selected data sets. The application of preprocessing to each individual vintage data set reduces the effects of the different acquisition and noise conditions, and leads to consistency in the amplitude response of the four vintages. This consistency facilitates the final amplitude cross‐calibration that is carried out using, as reference, the Cretaceous horizon reflections above the Brent reservoir. Such cross‐calibration can be considered as vintage‐consistent residual amplitude correction. Acoustic impedance sections, intercept and gradient amplitude‐versus‐offset attributes and coherent amplitude‐versus‐offset estimates are computed on the final cross‐calibrated data. The results, shown for three spatially coincident 2D lines selected from the 1982, 1989 and 1999 data sets, clearly indicate gas‐cap expansion resulting from oil production. Such expansion is manifested as a decrease in acoustic impedance and a modification of the amplitude‐versus‐offset trends in the apical part of the reservoir.     

Imaging pre‐existing natural fractures using microseismic data

In hydraulic fracturing treatments, locating not only hydraulic fractures but also any pre‐existing natural fractures and faults in a subsurface reservoir is very important. Hydraulic fractures can be tracked by locating microseismic events, but to identify the locations of natural fractures, an additional technique is required. In this paper, we present a method to image pre‐existing fractures and faults near a borehole with virtual reverse vertical seismic profiling data or virtual single‐well profiling data (limited to seismic reflection data) created from microseismic monitoring using seismic interferometry. The virtual source data contain reflections from natural fractures and faults, and these features can be imaged by applying migration to the virtual source data. However, the imaging zone of fractures in the proposed method is strongly dependent on the geographic extent of the microseismic events and the location and direction of the fracture. To verify our method, we produced virtual reverse vertical seismic profiling and single‐well profiling data from synthetic microseismic data and compared them with data from real sources in the same relative position as the virtual sources. The results show that the reflection travel times from the fractures in the virtual source data agree well with travel times in the real‐source data. By applying pre‐stack depth migration to the virtual source data, images of the natural fractures were obtained with accurate locations. However, the migrated section of the single‐well profiling data with both real and virtual sources contained spurious fracture images on the opposite side of the borehole. In the case of virtual single‐well profiling data, we could produce correct migration images of fractures by adopting directional redatuming for which the occurrence region of microseismic events is divided into several subdivisions, and fractures located only on the opposite side of the borehole are imaged for each subdivision.     

Numerical modelling of interface scattering of seismic wavefield from a random rough interface in an acoustic medium: comparison between 2D and 3D cases

Seismic wavefield scattering from a statistically randomly rough interface in a multilayered piecewise homogeneous medium is studied in 3D. The influence of the surface roughness on the scattered wavefield is analysed numerically by using a finite‐difference operator in the acoustic domain. Since interface scattering in the real practical sense is a 3D physical phenomenon, we show in this work that the scattering response of a randomly rough interface is not the same in 3D situations as in the 2D cases described in some earlier works. For a given interface roughness height in 3D, an interface roughness height at least three times greater is required to produce an equivalent phase scattering effect in 2D situations, for a given correlation length of the interface roughness scale. Based on observations from spectral analysis, we show that scattering results principally in de‐phasing and frequency band‐limiting of the incident wavefront, the frequency band‐limiting properties being comparable to cases reported in the literature for absorption and thin‐layer filtering. The interface scattering phenomenon should be critically considered when using amplitude and phase information from seismic signal during inversion processes.     

3D seismic residual statics solutions derived from refraction interferometry

We apply interferometric theory to solve a three‐dimensional seismic residual statics problem to improve reflection imaging. The approach calculates the static solutions without picking the first arrivals from the shot or receiver gathers. The static correction accuracy can be significantly improved by utilising stacked virtual refraction gathers in the calculations. Shots and receivers may be placed at any position in a three‐dimensional seismic land survey. Therefore, it is difficult to determine stationary shots and receivers to form the virtual refraction traces that have identical arrival times, as in a two‐dimensional scenario. To overcome this problem, we use a three‐dimensional super‐virtual interferometry method for residual static calculations. The virtual refraction for a stationary shot/receiver pair is obtained via an integral along the receiver/shot lines, which does not require knowledge of the stationary locations. We pick the maximum energy times on the interferometric stacks and solve a set of linear equations to derive reliable residual static solutions. We further apply the approach to both synthetic and real data.     

Scattered ground‐roll attenuation using model‐driven interferometry

Scattered ground roll is a type of noise observed in land seismic data that can be particularly difficult to suppress. Typically, this type of noise cannot be removed using conventional velocity‐based filters. In this paper, we discuss a model‐driven form of seismic interferometry that allows suppression of scattered ground‐roll noise in land seismic data. The conventional cross‐correlate and stack interferometry approach results in scattered noise estimates between two receiver locations (i.e. as if one of the receivers had been replaced by a source). For noise suppression, this requires that each source we wish to attenuate the noise from is co‐located with a receiver. The model‐driven form differs, as the use of a simple model in place of one of the inputs for interferometry allows the scattered noise estimate to be made between a source and a receiver. This allows the method to be more flexible, as co‐location of sources and receivers is not required, and the method can be applied to data sets with a variety of different acquisition geometries. A simple plane‐wave model is used, allowing the method to remain relatively data driven, with weighting factors for the plane waves determined using a least‐squares solution. Using a number of both synthetic and real two‐dimensional (2D) and three‐dimensional (3D) land seismic data sets, we show that this model‐driven approach provides effective results, allowing suppression of scattered ground‐roll noise without having an adverse effect on the underlying signal.     

The origin and nature of seismic reflections of sharp-based shoreface deposits (upper Jurassic Siliciclastics, northern France)

In order to advance understanding of the relationship between geological properties and their physical expression in reflection images, this study has focused expertise in reflection geophysics, petrophysics and sedimentology on the same geological object, in this case a succession of Upper Jurassic sharp‐based shoreface deposits embedded in offshore marine shales in northern France. This integrated approach to determine firstly the origin and nature of seismic reflections (calibration) and secondly to provide a means of extracting geological information from seismic imagery (inverse calibration) was built on the following analytical steps. Firstly, detailed and extensive petrophysical analyses of outcrop (plug) samples, continuous core and sonic well logs, in combination with a quantification of mineralogical and textural properties, allowed a direct conversion of acoustic properties (impedance) into sedimentological properties, resulting in a quantitative physical sequence stratigraphic model. Secondly, the integration of scale‐dependent acoustic measurements, ranging from 0.01 m and 320 kHz on cores up to the wavelength of field seismic data was established using an averaging algorithm (an effective‐medium‐theory type) as an upscaling approach. This alternative to a VSP or check shot allows an optimized depth–time conversion and hence determination of the origin of the seismic reflections with previously unattainable accuracy. Finally, the shape and scale dependence of impedance contrasts were integrated into so‐called singularity parameters that directly link depositional changes with information from seismic reflections: depositional changes in the shallow‐water domain are generally characterized by step functions, whereas those in more distal depositional environments are represented by spiky functions. This approach allows the recognition of the associated reflection events and, vice versa, it provides a unique opportunity to extract the character of impedance changes, and thus changes in depositional environment, from seismic reflection records in general. This integrated and multiscale characterization of sharp‐based shoreface deposits calibrates the typical reflection patterns for such sedimentary units. These include continuous high‐amplitude smooth and flat tops, discontinuous sharp basal reflections with variable amplitude, and complex sigmoidal high‐amplitude reflections within the compound shoreface deposits. In addition, the results of this study, by detailing the effects of scale and frequency on impedance changes, improve the identification of similar deposits in subsurface seismic data and the extraction of maximum amounts of geological information beyond seismic resolution.     

Drill‐bit seismic interferometry with and without pilot signals*

We use different interferometry approaches to process the seismic signals generated by a drill‐bit source in one well and recorded by seismic receivers located both in a second borehole and at the surface near the source well. We compare the standard interferometry results, obtained by using the raw drill‐bit data without a pilot signal, with the new interferometry results obtained by using the drill‐bit seismograms correlated with a reference pilot signal. The analysis of the stationary phase shows that the final results have different S/N levels and are affected by the coherent noise in the form of rig arrivals. The interferometry methods are compared by using different deconvolution approaches. The analysis shows that the results agree with the conventional drill‐bit seismograms and that using the reference pilot signal improves the quality of the drill‐bit wavefields redatumed by the interferometry method.     

Global-scale seismic interferometry: theory and numerical examples

Progress in the imaging of the mantle and core is partially limited by the sparse distribution of natural sources; the earthquake hypocenters are mainly along the active lithospheric plate boundaries. This problem can be approached with seismic interferometry. In recent years, there has been considerable progress in the development of seismic interferometric techniques. The term seismic interferometry refers to the principle of generating new seismic responses by cross‐correlating seismic observations at different receiver locations. The application of interferometric techniques on a global scale could create sources at locations where no earthquakes occur. In this way, yet unknown responses would become available for the application of travel‐time tomography and surface‐wave dispersion studies. The retrieval of a dense‐enough sampling of source gathers would largely benefit the application of reflection imaging. We derive new elastodynamic representation integrals for global‐scale seismic interferometry. The relations are different from other seismic interferometry relations for transient sources, in the sense that they are suited for a rotating closed system like the Earth. We use a correlation of an observed response with a response to which free‐surface multiple elimination has been applied to account for the closed system. Despite the fact that the rotation of the Earth breaks source‐receiver reciprocity, the seismic interferometry relations are shown to be valid. The Coriolis force is included without the need to evaluate an extra term. We synthesize global‐scale earthquake responses and use them to illustrate the acoustic versions of the new interferometric relations. When the sampling of real source locations is dense enough, then both the responses with and without free‐surface multiples are retrieved. When we do not take into account the responses from the sources in the direct neighborhood of the seismic interferometry‐constructed source location, the response with free‐surface multiples can still be retrieved. Even when only responses from sources at a certain range of epicentral distances are available, some events in the Green's function between two receiver locations can still be retrieved. The retrieved responses are not perfect, but the artefacts can largely be ascribed to numerical errors. The reconstruction of internal events – the response as if there was a source and a receiver on (major) contrasts within the model – could possibly be of use for imaging. With modelling it is possible to discover in which region of the correlation panel stationary phases occur that contribute to the retrieval of events. This knowledge opens up a new way of filtering out undesired events and of discovering whether specific events could be retrieved with a given source‐receiver configuration.     

Experimental estimation of velocities and anisotropy of a series of Swedish crystalline rocks and ores

To provide a guide for future deep (<1.5 km) seismic mineral exploration and to better understand the nature of reflections imaged by surface reflection seismic data in two mining camps and a carbonatite complex of Sweden, more than 50 rock and ore samples were collected and measured for their seismic velocities. The samples are geographically from the northern and central parts of Sweden, ranging from metallic ore deposits, meta‐volcanic and meta‐intrusive rocks to deformed and metamorphosed rocks. First, ultrasonic measurements of P‐ and S‐wave velocities at both atmospheric and elevated pressures, using 0.5 MHz P‐ and S‐wave transducers were conducted. The ultrasonic measurements suggest that most of the measured velocities show positive correlation with the density of the samples with an exception of a massive sulphide ore sample that shows significant low P‐ and S‐wave velocities. The low P‐ and S‐wave velocities are attributed to the mineral texture of the sample and partly lower pyrite content in comparison with a similar type sample obtained from Norway, which shows significantly higher P‐ and S‐wave velocities. Later, an iron ore sample from the central part of Sweden was measured using a low‐frequency (0.1–50 Hz) apparatus to provide comparison with the ultrasonic velocity measurements. The low‐frequency measurements indicate that the iron ore sample has minimal dispersion and attenuation. The iron ore sample shows the highest acoustic impedance among our samples suggesting that these deposits are favourable targets for seismic methods. This is further demonstrated by a real seismic section acquired over an iron ore mine in the central part of Sweden. Finally, a laser‐interferometer device was used to analyse elastic anisotropy of five rock samples taken from a major deformation zone in order to provide insights into the nature of reflections observed from the deformation zone. Up to 10% velocity‐anisotropy is estimated and demonstrated to be present for the samples taken from the deformation zone using the laser‐interferometery measurements. However, the origin of the reflections from the major deformation zone is attributed to a combination of anisotropy and amphibolite lenses within the deformation zone.     

Retrieval of local surface wave velocities from traffic noise – an example from the La Barge basin (Wyoming)

In regions where active source seismic exploration is constrained by limitations of energy penetration and recovery, cost and logistical concerns, or regulatory restrictions, analysis of natural source seismic data may provide an alternative. In this study, we investigate the feasibility of using locally‐generated seismic noise in the 2–6 Hz band to obtain a subsurface model via interferometric analysis. We apply this technique to three‐component data recorded during the La Barge Passive Seismic Experiment, a local deployment in south‐western Wyoming that recorded continuous seismic data between November 2008 and June 2009. We find traffic noise from a nearby state road to be the dominant source of surface waves recorded on the array and observe surface wave arrivals associated with this source up to distances of 5 kms. The orientation of the road with respect to the deployment ensures a large number of stationary points, leading to clear observations on both in‐line and cross‐line virtual source‐receiver pairs. This results in a large number of usable interferograms, which in turn enables the application of standard active source processing methods like signal processing, common offset stacking and traveltime inversion. We investigate the dependency of the interferograms on the amount of data, on a range of processing parameters and on the choice of the interferometry algorithm. The obtained interferograms exhibit a high signal‐to‐noise ratio on all three components. Rotation of the horizontal components to the radial/transverse direction facilitates the separation of Rayleigh and Love waves. Though the narrow frequency spectrum of the surface waves prevents the inversion for depth‐dependent shear‐wave velocities, we are able to map the arrival times of the surface waves to laterally varying group and phase velocities for both Rayleigh and Love waves. Our results correlate well with the known geological structure. We outline a scheme for obtaining localized surface wave velocities from local noise sources and show how the processing of passive data benefits from a combination with well‐established exploration seismology methods. We highlight the differences with interferometry applied to crustal scale data and conclude with recommendations for similar deployments.     

Feasibility of borehole ambient noise interferometry for permanent reservoir monitoring

The analysis of seismic ambient noise acquired during temporary or permanent microseismic monitoring campaigns (e.g., improved/enhanced oil recovery monitoring, surveillance of induced seismicity) is potentially well suited for time‐lapse studies based on seismic interferometry. No additional data acquisition required, ambient noise processing can be automatized to a high degree, and seismic interferometry is very sensitive to small medium changes. Thus there is an opportunity for detection and monitoring of velocity variations in a reservoir at negligible additional cost and effort. Data and results are presented from an ambient noise interferometry study applied to two wells in a producing oil field in Romania. Borehole microseismic monitoring on three component geophones was performed for four weeks, concurrent with a water‐flooding phase for improved oil recovery from a reservoir in ca. 1 km depth. Both low‐frequency (2 Hz–50 Hz) P‐ and S‐waves propagating through the vertical borehole arrays were reconstructed from ambient noise by the virtual source method. The obtained interferograms clearly indicate an origin of the ambient seismic energy from above the arrays, thus suggesting surface activities as sources. It is shown that ambient noise from time periods as short as 30 seconds is sufficient to obtain robust interferograms. Sonic log data confirm that the vertical and horizontal components comprise first arrivals of P‐wave and S‐waves, respectively. The consistency and high quality of the interferograms throughout the entire observation period further indicate that the high‐frequency part (up to 100 Hz) represents the scattered wave field. The temporal variation of apparent velocities based on first‐arrival times partly correlates with the water injection rate and occurrence of microseismic events. It is concluded that borehole ambient noise interferometry in production settings is a potentially useful method for permanent reservoir monitoring due to its high sensitivity and robustness.