Summary We present a novel interferometric method to reconstruct zero-offset VSP data from drill-bit vibrations that have been passively recorded with a seismic ocean-bottom array placed in the vicinity of a drilling rig. The method has been successfully applied during drilling of an exploration well for a 4 km deep target in the North Sea. At this borehole, significant depth uncertainty of the drilling target below a high-velocity formation required a check-shot velocity update and seismic look-ahead during drilling for optimal placement of the lowermost casing shoe. Drill-bit seismic can produce this information in real time without any impact on the drilling operation but is often hampered by lower signal-to-noise at zero offset due to the radiation pattern of PDC bits. We show that we can overcome such issues with the novel interferometric method, by reconstructing zero-offset data from a wide-aperture array. The resulting dataset is much denser than traditional VSP data. With about a tenfold increase in trace levels along the borehole, it is comparable to DAS VSP data. In addition, we obtain a high-fold 3D VSP image of the vicinity around the wellbore that can be used for reservoir characterization during the appraisal phase.

Seismic-while-drilling (SWD) by drill-bit source has been successfully used in the past few decades and is proven using variable configurations in onshore applications. The method creates a reverse vertical seismic profile (RVSP) data set from surface sensors deployed as arrays in the proximity of the monitored wells. The typical application makes use of rig-pilot reference (pilot) sensors at the top of the drill string and also downhole. This approach provides while-drilling checkshots as well as multioffset RVSP for 2D and 3D imaging around the well and prediction ahead of the bit. For logistical (sensor deployment) and cost (rig time related to technical installation) reasons, the conventional drill-bit SWD application is typically much easier onshore than offshore. We have developed a novel approach that uses a network of passive-monitoring sea-bottom nodes predeployed for microseismic monitoring to simultaneously and effectively record offshore SWD data. We study the results of a pilot test in which we passively monitored the drilling of an appraisal well at the Wisting discovery in the Barents Sea with an ocean-bottom cable deployed temporarily around the drilling rig. The continuous passive recording of vibration signals emitted during the drilling of the well provides the SWD data set, which is treated as an RVSP. The study is performed without rig-pilot signal. The results are compared with legacy data and demonstrate the effectiveness of the approach and point to future applications for real-time monitoring of the drilling progress, in terms of geosteering the drill bit and predicting formation properties ahead of the bit by reflection imaging.

Summary In order to limit the global warming of the planet to below 2o C, models show that net-zero release of anthropomorphic CO2 must be achieved by the mid-century. Since for the foreseeable future the most of the world's energy will still be provided by fossil fuels, other methods besides expanding the contribution of renewable energy are called for. According to the Intergovernmental Panel on Climate Change (IPCC), Carbon (short for carbon dioxide) Capture and Sequestration (CCS) is one such method. To achieve this climate goal current CCS efforts must increase by approximately 100-fold within the next 20 years. Geophysical simulations on suitable geologic models will provide an important tool to streamline and accelerate the vast expansion of site characterization and long term monitoring tasks to ensure the success of such large-scale CCS application.

Disclaimer Text and Data Mining. Any automated analytical technique aimed at analysing text and data in digital form in order to generate information including - but not limited to - patterns, trends and correlations (so-called Text and Data Mining) may only be applied if it is done by a research organisation or cultural heritage institution solely for the purpose of (non-commercial) scientific research as mentioned in Clause 3 of Directive (EU) 2019/790 (the Digital Single Market Directive) and Clause 15n of the Dutch Copyright Act [Auteurswet]. Text and Data Mining for commercial purposes is not allowed; all rights in this respect are reserved pursuant to Clause 4 paragraph 3 of the Digital Single Market Directive and Clause 15o paragraph 1 of the Dutch Copyright Act.

Abstract We monitor the seismic signal emitted from a rotating drill bit in real time with an array of seismic sensors at the seafloor. Drill-bit seismic signals provide information to locate the drill-bit position itself and to image geological objects ahead and around the drill bit for geosteering purposes during drilling operations. The data can be obtained in real time without the need to stop drilling for logging and without any additional downhole instrumentation in the bottom hole assembly. Drill-bit positioning accuracy is independent of measured depth and with meter level lateral precision. This is significantly better than conventional downhole gyro-based methods, especially for long horizontal wells. With sources along the drilled well path approaching a target reservoir we obtain a 3D reverse VSP (RVSP) image around the well for prediction ahead of the drill bit. This paper presents a case study from the Grane reservoir in the North Sea, where we utilize a permanent reservoir monitoring (PRM) array for listening to signals emitted from drilling with a PDC bit. We present imaging results from a highly deviated well and compare them to 3D seismic. The field example shows the ability to look ahead several hundreds of meters below the drilled well trajectory.

Summary We extract seismic-while-drilling (SWD) data from passive seismic recordings acquired with the Grane PRM array. The drill-bit pilot trace used for correlation is estimated by means of array beam forming. We observe a strong coherent signal emanating from the drill bit during times of actual drilling, when the rate of penetration (ROP) is high. For one PRM node near the wellhead, we assemble the while-drilling signals into regular depth intervals along a portion of the deviated well path. The resulting dataset is equivalent to a reverse vertical seismic profile (RVSP). We apply a check-shot VSP processing flow to extract reflected signals from below the drill bit that can be used for interpretation. We observe a good match of the RVSP corridor and corridor stack with a crossline section of the 3D seismic at Grane. This validates that passively collected seismic-while-drilling data at Grane may be utilized to look ahead of the drill bit. Using an ocean bottom cable, we can produce VSP information in real-time while drilling without the need to stop and pull the drill string for wireline deployment.

Disclaimer Text and Data Mining. Any automated analytical technique aimed at analysing text and data in digital form in order to generate information including - but not limited to - patterns, trends and correlations (so-called Text and Data Mining) may only be applied if it is done by a research organisation or cultural heritage institution solely for the purpose of (non-commercial) scientific research as mentioned in Clause 3 of Directive (EU) 2019/790 (the Digital Single Market Directive) and Clause 15n of the Dutch Copyright Act [Auteurswet]. Text and Data Mining for commercial purposes is not allowed; all rights in this respect are reserved pursuant to Clause 4 paragraph 3 of the Digital Single Market Directive and Clause 15o paragraph 1 of the Dutch Copyright Act.

We present the results of a pilot test where we passively monitored the drilling of an appraisal well in the Wisting discovery in the Barents Sea with an ocean-bottom cable deployed temporarily around the drilling rig. The continuous passive recording of drilling vibration signals emitted during the drilling of the well, provides a seismic while drilling (SWD) dataset, which is investigated as a reverse vertical seismic profile (RVSP). The study demonstrates the effectiveness of the approach and points to future applications for real-time monitoring of the drilling progress, both in terms of geosteering the drill bit and predict formation properties ahead of the bit. Presentation Date: Wednesday, September 18, 2019 Session Start Time: 8:30 AM Presentation Start Time: 9:20 AM Location: 303B Presentation Type: Oral

Kirchhoff depth migration is a frequently used method to transform measured seismic reflection data from the time into the depth domain. The results of migrating seismic ZO sections obtained by the CRS Stack method reveal a clearer image of the subsurface compared to the depth-migrated seismograms obtained by the conventional NMO/DMO/Stack approach. Demigration is the asymptotic inverse process to migration and aims at reconstructing a seismic time section from a depth-migrated image. By combining Kirchhoff migration and demigration, several imaging problems can be solved. One application of cascading these processes is the interpolation of missing traces in seismic sections. INTRODUCTION Kirchhoff depth migration is a widely investigated and frequently used tool in the world of seismic exploration to transform measured reflection data from the time into the depth domain. Kirchhoff migration assumes the subsurface to be built up by potential diffraction points. The migration result for such a point is obtained by a summation of seismogram amplitudes along the diffraction traveltime surface (Huygens surface) of the considered depth point. In the last decades, the originally purely kinematic migration schemes were extended in order to relate the amplitudes in the migrated images to physical subsurface properties. This is achieved by compensating for the geometrical spreading effects by means of applying suitable weight functions during the stacking process. Amplitudes in a depth image that are free of spherical divergence effects are called “true” amplitudes. If other effects on the reflection amplitudes (as, e.g., transmission loss, scattering, or source and receiver effects) are negligible or corrected for, these true amplitudes in the migrated images are a measure of the angle dependent reflection coefficient. Therefore, true-amplitude prestack depth migration allows to extract AVO/AVA (amplitude versus offset/angle) information which can be of great use in the search for hydrocarbon reservoirs. The asymptotic inverse process to migration is demigration. That is, demigration means the transformation of a migrated image into a seismic section in the time domain. The principle of Annual WIT report 2001 141 Kirchhoff demigration is analog to Kirchhoff migration: the value that is assigned to a point in the time domain is obtained by a stack of amplitudes in the migrated image which lie on the isochron of the point under consideration. While a true-amplitude migration compensates the geometrical spreading loss of the seismograms, true-amplitude demigration as its asymptotic inverse process has to re-introduce this effect back into the data. This can again be realized by applying an appropriate weight function during the stack. Defined in this way, true-amplitude demigration results in seismograms that are close to the actual recorded ones. True-amplitude migration and demigration can thus be applied one after the other without altering the amplitude information of the result. The basic idea of the Unified Approach Theory presented by Hubral et al. (1996) (basic concepts) and Tygel et al. (1996) (theory) is to combine Kirchhoff migration and demigration. If the macro-velocity model, the considered ray code (e.g., PP, SS, PS) or the measurement configuration is changed in between these processes, a multitude of imaging problems can be solved. Migration and demigration can either be applied in sequence (cascaded solution) or they can be analytically chained which leads to a single-stack solution for each specific imaging problem. Tasks that can be addressed in this way include • remigration, i.e., the updating of a migrated image according to a better velocity model or taking anisotropy information into account, • ray-code transformations, e.g., the simulation of S-wave seismograms from a given P-wave data set, • configuration transformations, i.e., the simulation of seismograms pertaining to certain measurement configurations, • redatuming, i.e., the simulation of seismograms that would be recorded on a chosen (horizontal) datum plane using data recorded on a given topography. Based on the Unified Approach Theory an imaging tool named “Uni3D” was developed by a working group at the Geophysical Institute, Karlsruhe University (see, e.g., Hertweck et al., 2001; Jäger, 2001). This program is currently able to handle 2.5D and 3D zero-offset data sets, and 2.5D multicoverage data. Due to the amount of data that would have to be processed and the limited computing resources, 3D prestack algorithms are not yet implemented. The main components of Uni3D are true-amplitude Kirchhoff migration and demigration for arbitrary macrovelocity models in 3D and 2.5D. By cascading these processes, we are able to address further imaging tasks. Due to practical reasons (Hertweck et al., 2001) the chained solution is, up to now, not considered. Kirchhoff (de)migration algorithms allow the computation of (de)migration results independently for all outputor input sample points. Thus, they are very suitable for parallel computation on multi-processor systems. MIGRATION OF NMO/DMO/STACK AND CRS STACK RESULTS Kirchhoff migration was applied to a real data set provided by the Federal Institute for Geosciences and Natural Resources (BGR), Hannover, Germany. These data were acquired over the 142 Annual WIT report 2001 Figure 1: ZO section obtained by the CRS Stack Chile Trench near 22◦S during the CINCA (Crustal Investigations onand offshore Nazca/Central Andes) program in 1995. The sampling rate of the data is 4 ms and the total record length is 15 s. The streamer consisted of 120 receivers placed in intervals of 25 m. Thus, the largest acquired offset is about 3 km. This is relatively small compared with the water depth of around 7 km. Figure 1 shows the simulated zero-offset (ZO) section obtained by the Common-ReflectionSurface (CRS) Stack method (see, e.g. Mann et al., 1999; Jäger et al., 2001). The sampling rate is 4 ms and the CMP spacing is 12.5 m. A subsurface structure beneath the clearly visible ocean bottom is hardly observable in the seismograms. The only strong event below the seafloor can be identified as a water multiple, its two-way traveltime is exactly twice the traveltime to the ocean bottom. It was attempted to attenuate this multiple during the CRS Stack, with more success on the right than on the left part of the image, as can be seen in the result. The entire seismic section is strongly dominated by diffraction patterns stemming from the rugged seafloor. For a subsequent poststack migration a velocity model had to be created. We used a macrovelocity model constructed by C. Ranero from Geomar (Kiel, Germany). It covers the right part of the data set starting at x = 58 km and reaches down to a depth of 20 km. Figure 2 shows the smoothed version of the original model, and is the one that was actually used to calculate the relevant part of the Green’s functions. The simulated zero-offset section was depth-migrated by means of a 2.5D weighted Kirchhoff migration to a target zone identical to the size of the macrovelocity model. To speed up the migration process and to avoid operator aliasing, the size of the stacking operator was restricted by limiting its maximum dip to 35◦. To attenuate boundary effects due to this limited aperture the operator was not truncated but tapered smoothly to zero over an additional range of 10◦ using a two-sided Hanning window. Annual WIT report 2001 143

Kirchhoff depth migration is an imaging process that transf orms reflection seismic data into the depth domain in order to obtain a structural image of the subsurfac e. Mathematically, it can be formulated as a weighted diffraction stack which is related to the Kirchhoff integral representation of the scalar acoustic wave equation and, hence, usually only form ulated for imaging of compressional or, more generally, monotypical waves. In this paper, the scala r approach to Kirchhoff imaging based on zero-order ray theory is extended to handle the full elast ic wavefield recorded with multicomponent receivers by considering the polarization of respecti v wave modes scattered at an interface. The weight functions that remove the effect of geometrical spre ading from the recorded amplitude change for each scattering mode and have thus to be extended for the c ase of an elastic multicomponent migration. The extended weight functions presented here are a lso v lid for converted waves. It is shown, that the method allows to retrieve the full elastic scatteri ng matrix of target reflectors. INTRODUCTION The graphical migration scheme proposed by Hagedoorn (1954 ) can be formulated as a summation over the Green’s function of the medium. This leads to the so-call ed Diffraction Stack Integral (DSI) as a wave-equation related integral representation of the grap hic l image construction technique (Bleistein and Gray, 2001). In the last two decades, several representatio s of the DSI were developed with slightly different approaches to approximate the (unknown) Green’s fu ction of the medium (see, e. g. Bleistein, 1987; Beylkin and Burridge, 1990). These approaches lead to a weighted diffraction stack in which a weight function accounts for the amplitude loss due to geome trical spreading along the propagation path. If other effects on the recorded amplitude can be neglected, this enables the recovery of the angle-dependent reflection coefficient from the migrated image, and is thus ca lled true-amplitude migration. The approach presented by Schleicher et al. (1993) makes exp licit use of zero-order ray theory in order to describe reflections of smoothly-curved first-order disc ontinuities as the part of the wavefield relevant for imaging. This approach can also be extended to the asympt otically inverse process of demigration, which leads to the so-called unified approach theory to seism ic imaging (Hubral et al., 1996; Tygel et al., 1996). In this paper, the work of the latter three authors, which has so far been strictly formulated only for monotypical reflections (i. e., it did not account for the pos sibility of mode conversions at an interface) is extended to include the vectorial properties of the elast ic wavefield, which can also be described by zero-order ray-theory (̌ Cervený, 2001). This approach can then also handle mode-con verted reflections. RAY-THEORETICAL DESCRIPTION OF ELASTIC AMPLITUDES The correct treatment of the vectorial properties of an elas tic wave within the framework of zero-order ray theory requires a formulation of the problem in ray-centere d coordinates. Assuming that the elastic wavefield has been recorded in three Cartesian components on a (fo r simplicity) planar measurement surface, I Annual WIT report 2002 119 can obtain the principle components in ray-centered coordi nates by means of a rotation ( Červený, 2001): ~ U(~ ξ, t) =   USV USH UP   = H   Ux Uy Uz   . (1) H denotes the local rotation matrix from the ray-centered coo rdinate system to the Cartesian measurement coordinate system. The amplitude ~ U(~ ξ, t) stands for the displacement field and subscript P refers to an incoming P-wave that is polarized perpendicular to the wa vefront1 andSV andSH refer to the two principal S wave components within the local tangent plane t o the wavefront. The correct rotation of the recorded seismograms into the ray-centered coordinate sys tem will thus enable a separation of the wavefield into its constituting wave modes. The rotation itself requi res the knowledge of the local emergence angle of a wavefront; a property which is, as we shall see, also necess ary for the calculation of the weight function. SMR ray branch MRR ray branch

Seismic images obtained by Kirchhoff time or depth migration are always accompanied by some artifacts known as “migration noise”, “migration boundary effects”, or “diffraction smiles”, which may severely affect the quality of the migration result. Most of these undesirable effects are caused by a limited aperture if the algorithms make no special disposition to avoid them. Likewise, strong amplitude variation along reflection events may also cause similar artifacts. All these effects can be explained mathematically by means of the Method of Stationary Phase. However, such a purely theoretical explication is not always easy to understand for applied geophysicists. By relating the terms of the stationary-phase approximation to simple geometrical situations, a more physical interpretation of the migration artifacts can be obtained. A simple numerical experiment for poststack (zero-offset) data indicates the problem and helps to develop an intuitive understanding of the effects and the methods to avoid them. INTRODUCTION Since the early work of Hagedoorn (1954), migration concepts have strongly improved and are now an important tool in the world of seismic imaging, either as prestack or poststack time and depth migration. A frequently used method is Kirchhoff migration (Schneider, 1978) that treats each depth point M on a sufficiently dense grid like a diffraction point. In an a-priori given macrovelocity model, the relevant part of the Green’s function of a point source at any single diffraction point M in the depth domain is calculated. The kinematic part of this Green’s function is the configuration-specific diffraction-traveltime surface, also called “Huygens surface”. The amplitudes of the input seismograms (or, to be more specific, of their derivatives) are stacked along the Huygens surface and assigned to the depth point M . This explains why the Kirchhoff migration scheme is also called a “diffraction stack”. If so desired, the effect of geometrical spreading can be removed from the output amplitudes by multiplying the data during the stack with a true-amplitude weight factor that is calculated from the dynamic part of the Green’s function. Annual WIT report 2001 87 Ideally, the extent of the Huygens surfaces, that is, the migration aperture, should be limitless so that no contributions due to the abrupt truncation of the sum occur. In practice, of course, the aperture is always limited by the region over which seismic data have been acquired. In other words, because of the finiteness of the survey area, Kirchhoff migration will always be a “limited aperture migration” (LAM) (Sun, 1998). This is, however, not the only reason why we have to deal with the effects of a finite migration aperture. In practical migration implementations, even ranges of source and receiver positions might be excluded where data actually have been acquired. Such a procedure can be advantageous because • less traces to sum leads to a speedup of the whole migration process, • a smaller operator excludes steeper dips, which helps to avoid operator aliasing (see, e.g., Abma et al., 1999), • less summation of data away from the signal reduces the stacking of unwanted noise. For the best possible reduction of aliasing and noise as well as the best computational efficiency, one would like to use a model-based aperture restriction, i.e., one would like to make use of the (projected) Fresnel zone (see, e.g., Schleicher et al., 1997; Sun and Bancroft, 2001). Unfortunately, it is difficult to determine the exact center and size of the Fresnel zone for each depth point prior to or during migration. A reasonable compromise between accuracy and practicability is to specify a common maximum migration aperture radius or a maximum stacking operator dip. These aperture reductions lead to dip-restricted migration operators as, for example, a 45◦ migration. With these kinds of operators, higher dips cannot be imaged. In regions where dips are known to be restricted, this is a very convenient way of reducing aliasing and improving computational efficiency at the same time. It should, however, be kept in mind that close to the maximum dip, these dip-restricted migration operators will achieve only kinematically correct images (see, e.g., Schleicher et al., 1997; Sun, 1998). For true-amplitude migration, the maximum operator dip must always be chosen somewhat larger than the maximum reflector dip to be imaged. The fact that the migration aperture is limited causes artifacts known as migration noise, boundary or aperture effects, or migration smiles. In this paper, we relate the mathematical explanation of the migration artifacts by means of the Method of Stationary Phase (see, e.g., Bleistein, 1984; Sun, 1998; Bleistein et al., 2001) to simple geometrical situations. This more physical interpretation leads to a more intuitive insight into these effects. Of course, since the stacking operations are the same in Kirchhoff time and depth migration, the corresponding artifacts are conceptually identical in both processes. Thus, we restrict our present discussion to Kirchhoff depth migration. It should, however, be kept in mind that everything said and shown in this paper with respect to an image in depth holds in the same way for an image in time. KIRCHHOFF MIGRATION Mathematically, the Kirchhoff migration process is expressed as an integration over the recorded wavefield and reads in 3D (Tygel et al., 1996)

To achieve high recovery rates, modern-day production management can benefit from not only snapshot images of the state of the reservoir at regular time intervals, but also continuous monitoring of the dynamic processes induced by pressure changes and fluid movement during production. Production management using time-lapse 4D snapshots is reactive, i.e., adjustments addressing the sweep efficiency or reservoir integrity can only be instigated once the next snapshot image is available after acquisition, processing and interpretation, often years later. For a more proactive reservoir management, it is important to have dynamic reservoir information in real time between the seismic time-lapse snapshots. Such information is contained in microseismic monitoring data and in surface or borehole deformation measurements. If sensors are permanently installed, this information comes at a negligible additional cost, provided that the data can be transferred to shore in real-time and processed automatically.Time-lapse 4D snapshot images are typically obtained over a period of years and are inadequately sampled for capturing dynamic reservoir changes taking place over much shorter time intervals, from hours to days. Such changes can include variations in the permeability caused by scaling or compaction (Barkved and Kristiansen, 2005), changes of the fluid phase owing to pressure variations (e.g., Osdal et al., 2006), unintended alterations of the flow paths owing to out-of-zone injections and fault reactivation (e.g., Schinelli et al., 2015), or movement in the overburden, potentially compromising the integrity of infrastructure in the form of casing failures or seafloor subsidence (e.g., Yudovich et al., 1989; Hatchell et al., 2017). While 4D seismic data can capture the cumulative effect of such processes by evaluating differences in still images every few years, they provide little information about when exactly the associated dynamic changes occurred and how they relate to changes in flow rate and pressure that may have been captured through continuous measurements in the wellbores accessing the reservoir.Microseismic events from within or around a producing reservoir can be indicative of reservoir fluid pathways and sub-seismic reservoir compartmentalization (e.g., Maxwell and Urbancic, 2001), or stress changes and associated production- related deformations in the vicinity (e.g., Teanby et al., 2004; Zoback and Zinke, 2002; Wuestefeld et al., 2011, 2013). Continuous monitoring of seismicity can also help in assessing deformation-related risks to infrastructure over the life of a field. Combined with pressure and flow rate, such data can provide the necessary information to capture dynamic processes in the reservoir right when they happen. In conjunction with geomechanical flow modelling, production optimization strategies can thus be validated and adjustments can be properly planned at an early stage. The result will be an improved sweep efficiency with a further increased recovery factor, as well as better risk assessment with the avoidance of potentially costly mitigation actions. A better understanding of, and continuous information about, the reservoir dynamics may even help to plan a 4D seismic strategy better. This can include better definition of suitable intervals for the acquisition of time-lapse images. These intervals could be irregular, depending on the state of reservoir development and type of recovery method. Continuous monitoring may also provide a means to high-grade areas of the reservoir for partial 4D imaging at lower cost and faster turnaround in between ‘full’time-lapse surveys (Hatchell et al., 2013).In this paper we discuss the general ability to monitor microseismic events in an offshore setting and presents results from a real-time monitoring pilot in Norway. We validate the concept of continuous real-time monitoring from a fibre-optic deep-water installation by comparing our automatic detections with data from a regional seismograph network.

Time-lapse information can be used for different purposes, depending on whether it spans several years or only a few months between snapshots, or even reflects producton/injection-induced seismicity in real time. Synchronizing the availability of such 4D data with production-relevant decision gates, and ensuring sufficient resolution for these decisions, is the key to improve efficiency and optimize impact of 4D. PRM systems need to reliably function over the entire production lifetime of an asset and provide the sensitivity to both detect small 4D signals over short repeat intervals and continuously monitor tiny production-induced deformations. A fiberoptic solution is best capable of providing the required endurance in the seafloor environment over the entire asset lifetime. We present the next generation Optoseis pressure-balanced technology, and some novel procesing techniques that are very useful for PRM with sparse receiver sampling.

At the end of 2012, Petrobras installed the first deep-water optical permanent reservoir monitoring (PRM) system over the Jubarte oilfield in the Campos basin offshore Brazil. The PRM pilot project covered approximately 9 sq. km with a fully fiberoptic 4-component (4C) sensor array. All topside optoeletronic equipment was installed on FPSO P-57. The primary objective was to validate the fiberoptic sensing technology to detect subtle impedance changes in the reservoir. This goal was achieved, allowing interpreters to apply an interpretation strategy to improve the understanding of flow patterns and to update the geological scenarios of the area based on very high quality 3D/4D images provided by the permanent system.

Summary Small waveform and directivity variations of marine airgun signatures due to waves interacting with the source float are a source of 4D noise. We are assessing the magnitude of this noise by first measuring the amount of variability from near-source auxiliary data and then modeling synthetic time-lapse ocean bottom seismic data with realistic source variations based on the measured statistics and standard ocean wave models. We quantify the contribution of source variations to 4D noise as a function of sea state by calculating the NRMSD attribute in the image domain. We find that up to 4% NRMSD can be attributed to source variations under realistic scenarios, with two main contributing effects: variations of individual gun signatures due to pressure changes, and array directivity variations due to the wave-induced pitch and roll of the source floats. The latter effect has a larger impact on the 4D noise in our simulations and depends more on the wave steepness rather than the wave height. While waveform variations can be addressed by a nearfield-based shot-by-shot designature, directivity variations are difficult to correct without knowledge of the sea surface shape.

Despite a tremendous leap in efficiency and wavefield sampling over the past two decades, it is sometimes still difficult to achieve adequate coverage and resolution with marine streamer acquisition. It is therefore necessary to carefully study the acquisition geometry, especially with respect to resolution and image quality in the crossline direction. We have extended the existing focal beam method to analyze marine streamer geometries with and without using all multiples. The focal beam analysis method provides a direct link between the acquisition parameters at the surface (number of sources, source locations, number of detectors, detector locations, and which sources are sensed by which detectors) and the image quality at a target location in the subsurface. We have further derived the concept of weighted focal beams to assess the angular aperture available for a specific acquisition configuration at the considered depth point. So far, this method has only addressed illumination by primaries. Multiples are traditionally suppressed but can instead be used in the imaging process, in which they often contribute positively by opening up the available aperture. Therefore, we have extended the focal beam analysis concept to account for multiple-reflected waves. Focal beams are more densely sampled in the ray-parameter domain when using the full wavefield (primaries and multiples). We especially addressed coverage deficiencies that can occur when the survey geometry deviates from the ideal, for example, due to feathering. As a result, extra infill lines must be acquired. We have determined how infill analysis can be performed in the depth domain via the focal beam theory. Our analysis provides the geophysical-based infill specifications to assess the impact of coverage holes on data quality. This leads to more accurate infill decisions compared with traditional common midpoint-based criteria.

Summary A significant number of microseismic events were detected over 120 days of passive monitoring with a deepwater PRM pilot array offshore Brazil. The array is installed in 1240–1310m water depth and consists of over 700 four-component stations. Recording occurred during two consecutive two-month periods in between active seismic surveys. The passive monitoring detected distinct event swarms that are highly clustered in space and time. These events occur at an estimated depth of about 5 km with moment magnitudes ranging from 0.2 to 1.9. The seismicity occurs in a depth interval near a currently undeveloped deeper reservoir and is possibly of natural origin. The capture of such seismicity is valuable input for long-term risk assessment and development planning of the lower reservoir.

Summary Despite a tremendous leap in efficiency and wave field sampling over the last two decades, it is sometimes still difficult to achieve adequate coverage and resolution with marine streamer acquisition. It is therefore necessary to carefully study the acquisition geometry, especially with respect to resolution and image quality in the cross-line direction. In this paper, we extend the focal beam method to analyze marine streamer geometries with and without using all multiples. Multiples are traditionally suppressed, but can instead be used in the imaging process where they often contribute positively by opening up the available aperture. We especially address coverage deficiencies that can occur when the survey geometry deviates from the ideal, for example due to feathering. As a result, extra ’infill’ lines must be acquired. We present how infill analysis can be performed via the focal beam theory to assess the impact of coverage holes for primaries as well as multiples.

Abstract Mitigation measures to protect marine mammals from sound emissions can be implemented more effectively if the sound output of a seismic source is known beforehand. We present a forward modelling workflow to assess the environmental impact of seismic exploration in areas with marine mammal activity at the survey planning stage. The focus is on modeling the output sound exposure level created by the seismic source, typically an airgun array. The employed physical model considers the hydrodynamics of marine airguns, including air bubble oscillations, near-source interactions with other airguns in array-configurations as well as interaction with, and impact of, the surface ghost reflection. Sound pressure and/or sound exposure levels can be obtained as a function of distance from the source through consideration of different geometrical spreading models and a hearing threshold term for different marine mammal species. Combined, these results allow for the estimation of sound output and propagation properties in a number of different scenarios, in line with most recent regulatory requirements. The presented model allows an accurate estimation of sound output of seismic sources up to about 1 kHz, which marks on average the frequency at which airgun sources have decayed by 50 dB from their respective spectral maximum. We present example modeling results that illustrate how the workflow can be employed for the planning of specific mitigation measures, such as exclusion zones, safety radii, and soft-start procedures.

Summary The current implementations of marine source modeling theory have been calibrated and adjusted against measured signatures with a goal of high modeling accuracy within a limited frequency band. As multicomponent streamers and source de-ghosting allows for utilizing a significantly broader range of frequencies in seismic imaging, adjustments to the modeling are necessary in order to achieve a better match between measured and modelled signatures over the expanded frequency band. This includes significant changes to the calibration process such as considering de-ghosted measurements and avoiding the historically rooted standard DFS V filtering. The modeling results after applying the improved calibration show a very good match with measured array signatures over a wide frequency range.