EAGE Workshop on Land and Ocean Bottom - Broadband Full Azimuth Seismic Surveys

Land broadband is a challenge when compared to marine broadband as the seismic bandwidth is limited by many factors such as coherent noise, sampling, our ability to preserve bandwidth in data processing and last but not least the near surface effects.

We will address these issues with the vibroseis source and bandwidth can be also limited by our ability to generate sufficient low and high frequency energy that can penetrate the near surface. The key strep in a successful spreading of onshore broadband seismic, especially in the Middle East will be our ability to successfully deal with the near surface effects limiting the bandwidth and we will see that the key element to solve this problem is to dramatically increase the sampling density of the 3D acquisition designs in order to generate data sets where all the noise are properly sampled and never aliased.

Borehole seismic can also be used as a tool to study and quantify the surface seismic bandwidth high end with the optimum one achieved in borehole seismic by Walkaways and or 3D VSPs.

KOC conducted a feasibility study to assess the maximum acceptable limit to relax the point-receiver sampling in order to move from the macro receiver line geometry to the single (skinny) receiver line geometry and study the impact of applying the advanced noise attenuation techniques to effectively remove any further aliased noise. The results suggested that it is safe to relax the point-receiver spacing up to 12.5m and effectively remove the coherent noise, even if it is aliased. Using such geometry design will allow better utilization of the available single-sensors; consequently will lead to a cost effective seismic data acquisition surveys.

Ocean Bottom Node (OBN) acquisition is increasing in popularity due to its inherent benefits such as full azimuth coverage, long offsets, rich low frequencies, and high signal-to-noise ratio. In order to obtain a good reservoir image, the multiple wavefield must be properly removed from OBN data. However, due to the high acquisition cost, OBN surveys usually have sparse receiver sampling. Additionally, the difference between their source and receiver datums is significant. These two factors make conventional SRME inapplicable to OBN surveys alone.

By incorporating an existing wide azimuth towed streamer (WATS) survey, we are able to overcome the OBN obstacles for traditional SRME application. In this paper, we present a practical 3D OBN SRME flow, which combines OBN and WATS data for OBN free-surface multiple prediction. Our flow does not require any velocity and reflectivity information, and naturally predicts the whole set of free-surface related multiples with the data’s full bandwidth preserved. Its effectiveness in removing OBN multiples from complex subsalt settings has been proven in many practical applications.

We show the application of our flow on an OBN survey from the deep-water, Gulf of Mexico. The multiples have been effectively removed, and subsalt images become clearer and more coherent.

Several new technologies. These technologies can be divided in three main fields: the seismic source, the seismic receiver and the processing algorithms. We will discuss these new technologies, where the main focus will be on the source- and receiver configurations that are requires for broadband full azimuth surveys.

In this work we present principles and results of the vibroseis technology, which is based on complex nonlinear frequency modulated signals.

Improving the efficiency of ocean bottom sensor (OBS) surveys has been a long time goal of both oil companies and acquisition contractors. Full azimuth and offset OBS surveys have historically been time consuming to acquire due to limitations in the amount of equipment and the efficiency of the crews. A large increase in efficiency can be realized through the use of multiple independent source vessels working simultaneously. The field of Compressive Sensing (CS) provides the necessary framework to allow the vessels to work simultaneously and be able to produce comparable data sets to a traditionally acquired OBS survey. We describe the operational issues of acquiring a survey with multiple source vessels including the CS techniques used.

Increasing the vibratory seismic bandwidth especially on the low end brought new challenges. To reach deeper signal penetration, to improve reflection continuity and to enhance inversion solutions, broadband non-linear sweeps were introduced to overcome the mechanic and hydraulic limitations of a standard seismic vibrator. The main disadvantage is the substantially longer non-linear sweep length compared to the length of an equivalent energy linear sweep. Pseudo-random sweeps are also tested and applied with success, but have not gathered enough momentum for widespread use yet. We have developed a global optimisation method and defined constraints to produce broadband pseudo-random sweep sequences, which have the potential to replace broadband non-linear sweeps with satisfactory quality while reducing sweep length to increase productivity. The optimisation process could incorporate the need for higher sweep energy, lower spectral fluctuations and also keeping vibrator limitations, the control parameters which are unavailable in the standard software given by the vibrator manufacturers. Our numerical calculations show that optimised broadband pseudo-random sweeps have the potential for a few percentage productivity increase in normal applications, but several hundred percentage enhancement in urban seismic measurements, where resonance effect reductions are also on our side.

Azimuth Moveout (AMO) is wave-equation operator that can be effectively applied to interpolate and regularize 5-D seismic data and improve the accuracy, illumination and imaging of structurally complex targets. AMO is strictly derived from the wave equation and therefore carries the correct kinematic, phase and amplitude transformation. The dipping events are moved correctly when transforming or interpolating the data and diffractions are preserved in a manner that is consistent with the wave equation. This property sets AMO apart as a seismic interpolator from more conventional ones. The AMO operator rotates the azimuth and modifies the offset of 3-D prestack data. It is analytically derived by cascading the forward and inverse 3-D DMO.

Traditionally, AMO was designed to address the marine acquisition shortcomings, and regularize common azimuth data. Typical implementation in time-space domain or frequency-wavenumber domain were primarily designed for 4-D data output. The recent focus by exploration and production companies on wide azimuth data both onshore and offshore creates the need to extend the AMO implementation to 5-D, by using offset vector tiles (OVT) to control the output geometry.

Ocean bottom node (OBN) surveys are gaining popularity in deep-water environments and are frequently acquired to enrich streamer data for field development. However, the benefits of OBN surveys are often restricted by high acquisition costs and are thus acquired with very limited shot and receiver coverage, limiting the capability to build an adequate PSDM velocity model based on OBN data alone. Combining streamer and OBN surveys for model building and imaging can overcome the restricted aperture of OBN itself, resulting in a more accurate velocity model and, consequently, better reservoir images. We present a case study of wide azimuth towed streamer (WATS) and OBN joint model building and imaging on a subsalt reservoir located in a deep-water region of the U.S. Gulf of Mexico.

Surface-wave inversion and early-arrival full-waveform inversion can be integrated to extract more information and value from recorded field data. We show how we integrated these technologies and why this integration is beneficial. In fact, these techniques are complementary one to each other; surface waves provide a high-resolution near-seabed model in a depth range where acoustic full-waveform inversion techniques typically produce suboptimal results, and where reflection tomography using the up-going wavefield is typically compromised by a lack of traces that have recorded pre-critical reflections and by multiple and noise contamination at near offsets. The resulting high-resolution P- and S-wave near-seabed velocity models can be obtained after completion of seismic acquisition, and then used for short- to long-wavelength perturbation corrections, for general earth model building, and for geohazard assessment and other geotechnical applications. Obviously, tomography at deeper overburden and reservoir depth range, as well as PSDM imaging quality of the reservoir targets is expected to benefit from the correct velocity modelling of small-scale and complex structural and stratigraphic features in the near surface. For shallow hazard assessment, the integrated near-surface model can be used as interpretation product, and would improve mirror migration of the down-going wavefield.