Amplitude Variation With Offset Attributes Research Articles

Quantitative evaluation of gas hydrate reservoir by AVO attributes analysis based on the Brekhovskikh equation

AVO (Amplitude variation with offset) technology is widely used in gas hydrate research. BSR (Bottom simulating reflector), caused by the huge difference in wave impedance between the hydrate reservoir and the underlying free gas reservoir, is the bottom boundary mark of the hydrate reservoir. Analyzing the AVO attributes of BSR can evaluate hydrate reservoirs. However, the Zoeppritz equation which is the theoretical basis of conventional AVO technology has inherent problems: the Zoeppritz equation does not consider the influence of thin layer thickness on reflection coefficients; the approximation of the Zoeppritz equation assumes that the difference of wave impedance between the two sides of the interface is small. These assumptions are not consistent with the occurrence characteristics of natural gas hydrate. The Brekhovskikh equation, which is more suitable for thin-layer reflection coefficient calculation, is used as the theoretical basis for AVO analysis. The reflection coefficients calculated by the Brekhovskikh equation are complex numbers with phase angles. Therefore, attributes of the reflection coefficient and its phase angle changing with offset are used to analyze the hydrate reservoir's porosity, saturation, and thickness. Finally, the random forest algorithm is used to predict the reservoir porosity, hydrate saturation, and thickness of the hydrate reservoir. In the synthetic data, the inversion results based on the four attributes of the Brekhovskikh equation are better than the conventional inversion results based on the two attributes of Zoeppritz, and the thickness can be accurately predicted. The proposed method also achieves good results in the application of Blake Ridge data. According to the method proposed in this paper, the hydrate reservoir in the area has a high porosity (more than 50%), and a medium saturation (between 10% and 20%). The thickness is mainly between 200m and 300m. It is consistent with the previous results obtained by velocity analysis.

Petrophysical inversion based on f-s-r amplitude-variation-with-offset linearization and canonical correlation analysis

The prediction of petrophysical properties, such as porosity and rock-fluid volumes, from partially stacked seismic data typically requires a rock-physics model that often is lithology dependent and difficult to calibrate. We adopt canonical correlation analysis (CCA) to infer the underlying relation between petrophysical properties and elastic attributes estimated from seismic data. We develop a two-step inversion approach: first, we predict elastic properties from partially stacked seismic data using a Bayesian linear inverse method based on an amplitude-variation-with-offset (AVO) linearization in terms of fluid, rigidity, and density factors, and then we predict petrophysical properties from the estimated AVO attributes using CCA. The novelty of our approach is the application of CCA to the fluid and rigidity factors, which avoids the calibration of an explicit rock-physics model by automatically deriving a linear relation in the lower dimensional space of the canonical variables. The parameterization of the linearization in terms of fluid, rigidity, and density factors maximizes the correlation with respect to the petrophysical properties of interest. Furthermore, the probabilistic approach is extended to the petrophysical inversion using Bayesian linear theory and the posterior distribution of petrophysical properties conditioned by seismic data is computed by combining the probability distributions obtained from seismic and petrophysical inversion to propagate the uncertainty from the seismic to the petrophysical domain. The inversion is validated on a synthetic case that finds high accuracy of our formulation. A case study with synthetic and real partially stacked seismic data also is presented and compared to a traditional inversion with an explicit rock-physics model.

Model-based approach to improve class 1 AVO attributes

We propose a model-based Amplitude Variation with Offset (AVO) approach to address the assumption of weak elastic contrast in the linearized approximations of Aki-Richard equation. Existing approximations, especially for the Class 1 AVO response, which indicates the positive elastic contrast, deviate significantly from the Zoeppritz equation in presence of large elastic contrast. Results of the proposed approach show that it can minimize the deviation from the Zoeppritz equation, improve AVO attributes, and is capable of providing the desired attribute for characterizing a reservoir. The method starts with two matrices. One matrix is of the simulated rock properties termed as the rock-property-matrix and the other of the Zoeppritz AVO responses for those rock properties termed as the response-matrix. A model-based AVO equation or the basis-function-matrix is computed utilizing the rock-property-matrix and the response-matrix. The inverse of the basis-function-matrix is applied to the real data to get the AVO attributes from this approach. The conventional AVO (Aki-Richards) and model-based AVO attributes are compared in a class 1 AVO environment from the offshore east coast of India. The curvature attribute based on model-based AVO shows significant improvement. The synthetic-seismic correlation of the curvature attribute improves from −0.03 to 0.9 while synthetic-seismic correlation of the gradient attribute improves from 0.5 to 0.77. The improved AVO attribute volumes add significant values in the reservoir characterization.

New AVO expression and attribute based on scaled Poisson reflectivity

A new expression for Amplitude Variation with Offset (AVO) is proposed. The expression introduces reflection amplitudes as a function of two very common AVO parameters, the Gradient (G) and scaled Poisson reflectivity (SPR). The SPR is proven to be an excellent indicator of hydrocarbon with more sensitivity to fluid than other ordinary AVO attributes. Another advantage of the SPR is that it works for all types of gas-saturated sands commonly targeted by AVO. From the new expression, a new AVO classification based on SPR is introduced. The classification involves two classes: 1) Class H, with negative SPR and negative G, represents rocks with high probability of hydrocarbon presence, and 2) Class N, with nearly zero or positive G and SPR, includes rocks with high probability of no-hydrocarbon presence. An AVO attribute combining SPR and Gradient is derived from the expression and is proposed as a good indicator of fluid presence. The approximation and the attribute are applied on Marmousi-2 model datasets and on a field data from Alberta, Canada. The attribute has demonstrated a good identification of the hydrocarbon-saturated reservoirs in theoretical model and the real field data. The new AVO approximation, classification and attribute are intended to complement other existing approximations and reduce uncertainty from their interpretations.

Determination of subsurface rock properties from AVO analysis in Konga oil field of the Niger Delta, Southeastern Nigeria

Amplitude variation with offset (AVO) analysis was carried out on Konga oil field, an onshore oil field in the Niger Delta, Southeastern Nigeria. The study consisted of forward modeling from rock parameters measured from well logs and AVO analysis of events on pre-stack time migrated 3D seismic gathers. Forward modeling predicted specific AVO behaviour of anomalous reservoir sands in the field. Density, compressional and shear wave velocity logs, combined with a wavelet extracted from recorded seismic gathers were used to generate synthetic seismic gathers and Gassmann fluid substitution models. These synthetics supported key observations made of the AVO response in the field data. AVO attributes (intercept and gradient) were derived from analysis of common depth point (CDP) gathers obtained from a 3D pre-stack seismic survey. The attributes were cross-plotted to establish trends against which anomalous amplitude behaviour were identified. Reflections related to shales and brine sands exhibit a relatively small range of orientations creating a dominant “background trend” against which anomalous events related to hydrocarbon-saturated reservoirs show clear deviations. On the basis of crossplot analyses (reflectivity versus offset/angle and intercept versus gradient) and modeled acoustic impedance, Class 1 type AVO anomalies observed are associated with non-hydrocarbon bearing clastic rocks that are most probably brine saturated in the field. Evidence from these results show that drilling in this field in search of hydrocarbon reservoirs poses a risky venture.

Energy-weighted Amplitude Variation with Offset: A new AVO attribute for low impedance gas sands

Amplitude Variation with Offset (AVO) is a technique that has been widely used as a reliable indicator of hydrocarbon expressions in seismic data. The technique uses mathematical approximations and approaches in order to estimate variations in seismic amplitude as a function of incidence angle, and related them to lithology and pore-fluids. In this study, a new AVO attribute is presented. The Energy-weighted AVO attribute uses Zoeppritz approximations and exploits the fact that hydrocarbon-bearing sediments show anomalous seismic responses relative to their backgrounds. The new AVO attribute emphasizes this hydrocarbon-associated effect and attenuates the background signal. In addition, the fact that the attribute integrates the signatures of conventional AVO attributes (intercept×gradient) with seismic data makes the discrimination of hydrocarbon anomalies from other anomalies caused by anomalous lithologies such as coal, carbonates or salt, straightforward. More interestingly, the new attribute can be applied to both prestack and poststack data. This indeed would solve several problems and difficulties encountered with prestack data, as prestack data are generally less readily available, and have a much lower signal to noise ratio.The process has been examined using synthetic data generated from Zoeppritz equations, and then applied to real prestack and poststack data. At all stages, promising results have been derived and the hydrocarbon effect and extent could be successfully predicted, also the AVO effect of hydrocarbons could be investigated even in the poststack data.

Probabilistic formulation of AVO modeling and AVO-attribute-based facies classification using well logs

We have developed a probabilistic formulation to derive the probability density function of amplitude variation with offset (AVO) attributes given the distribution of elastic properties, in the context of AVO analysis of well log data. The proposed probabilistic formulation includes the analytical expression of the posterior distribution and contributes to the correct propagation of the uncertainty through the AVO model. When this analysis is performed in each facies, the resulting posterior probability density functions of AVO attributes, conditioned by the elastic attributes and the underlying log-facies classification, can be used in a Bayesian inversion workflow to obtain the AVO-attribute-based facies classification. This classification is then compared with the petrophysical log-facies profile and can be extended to the entire reservoir model if inverted seismic attributes are available. We demonstrate the methodology on a data set from an onshore tight-sand gas reservoir in Texas. The main results of the application are the set of probability distributions of AVO properties as a function of elastic attributes in each facies and the corresponding AVO-based facies profile at the well location. A comparison of different statistical assumptions and a sensitivity analysis on the resolution of the elastic data set are presented.

Introduction to this special section—AVO

AVO (amplitude variation with offset) analysis was formerly introduced by Bill Ostrander (in a presentation at the 1982 SEG Annual Meeting), for confirmation of bright spots and other anomalous reflections seen on RAP (relative-amplitude preserved) sections. Since then, we have come a long way in extracting more information from prestack seismic amplitudes. A few crucial points that have emerged in all these analyses are: (1) an understanding of the assumptions underlying the theory being put to use and how well these assumptions are met in practice, (2) the realization that AVO reliability strongly depends on the quality of the processing, and (3) awareness of the importance that correct interpretation techniques be employed to extract meaningful information from the AVO attributes.

Amplitude analysis with an optimal model-based linear AVO approximation: Part II — Field data example

AVO analysis can be conducted by estimating amplitude variation with offset (AVO) attributes from seismic prestack data and by characterizing the measured amplitude responses by the position of their projection in the attribute space. The most common AVO attributes are the intercept (zero-offset reflectivity) and AVO gradient. We have constructed an optimized, model-based linear AVO equation that is more accurate than usual AVO approximations. The parameters of this equation represent new AVO attributes that are more directly related to the information contained in seismic reflection amplitudes. We use these new AVO attributes to classify reflector responses from field data. Five seismic facies are defined that are characterized by differentdistributions of seismic parameters. Nine reflector classes are formed by associating appropriate pairs of facies. The expected locations of the different reflector classes in the space of optimized attributes are found by modeling and are used to derive a classification scheme. This scheme is applied to sections of optimized attributes calculated from the prestack data, leading to a vertical section showing the distribution of most probable facies in an area containing a sand reservoir. We compare the new approach to classification with intercept and gradient. The new method is more robust and less sensitive to the number of attributes (two or three) used for classification. It offers an optimal, flexible, and robust way of extracting the information contained in reflection amplitudes by simple linear AVO equations.

Anisotropic geometrical-spreading correction for wide-azimuth P-wave reflections

Compensation for geometrical spreading along a raypath is one of the key steps in AVO (amplitude-variation-with-offset) analysis, in particular, for wide-azimuth surveys. Here, we propose an efficient methodology to correct long-spread, wide-azimuth reflection data for geometrical spreading in stratified azimuthally anisotropic media. The P-wave geometrical-spreading factor is expressed through the reflection traveltime described by a nonhyperbolic moveout equation that has the same form as in VTI (transversely isotropic with a vertical symmetry axis) media. The adapted VTI equation is parameterized by the normal-moveout (NMO) ellipse and the azimuthally varying anellipticity parameter [Formula: see text]. To estimate the moveout parameters, we apply a 3D nonhyperbolic semblance algorithm of Vasconcelos and Tsvankin that operates simultaneously with traces at all offsets andazimuths. The estimated moveout parameters are used as the input in our geometrical-spreading computation. Numerical tests for models composed of orthorhombic layers with strong, depth-varying velocity anisotropy confirm the high accuracy of our travetime-fitting procedure and, therefore, of the geometrical-spreading correction. Because our algorithm is based entirely on the kinematics of reflection arrivals, it can be incorporated readily into the processing flow of azimuthal AVO analysis. In combination with the nonhyperbolic moveout inversion, we apply our method to wide-azimuth P-wave data collected at the Weyburn field in Canada. The geometrical-spreading factor for the reflection from the top of the fractured reservoir is clearly influenced by azimuthal anisotropy in the overburden, which should cause distortions in the azimuthal AVO attributes. This case study confirms that the azimuthal variation of the geometrical-spreading factor often is comparable to or exceeds that of the reflection coefficient.

Prestack inversion of a Gulf of Thailand (OBC) data set

Seismic waveform inversion can be an important tool for reservoir characterization because high-resolution (within the seismic-frequency band) images of elastic parameters are obtainable from good quality seismic data. A detailed description of the velocity field with both high- and low-frequency variations is essential for delineating elastic properties of the medium for direct detection of hydrocarbon, lithologic discrimination, and estimation of fluid contents. Conventionally, amplitude-variation-with-offset (AVO) analysis determines fractional changes in P-impedance and Poisson's ratio from a linear fit of normal-moveout (NMO) corrected P-reflection amplitude to a two- or three-term approximation of reflection coefficients. Although, AVO analysis gives a useful estimation of elastic properties of the subsurface from AVO attributes, its limitation is well recognized; for example, it deals with “primaries only” models and is inadequate to account for mode-converted waves and internal multiples. In spite of several sources of error (such as random and coherent noise and a 1D earth-model assumption), a prestack seismic waveform inversion is generally desirable for more accurate mapping of the variation of the elastic properties of the subsurface. However, practical implementation of such an inversion scheme on a routine basis becomes discouragingly difficult as (1) the inverse problem is highly nonlinear, (2) the error function (a measure of data misfit in general) possesses multimodality, (3) the inverse problem is extremely computer intensive, and (4) inversion needs better data acquisition systems and data processing steps to preserve true amplitude in order to get precise estimates of elastic properties of the medium. Recently, Sen and Roy (2003) developed a regularized gradient-descent–based waveform inversion algorithm that is highly robust and addresses many of the problems encountered in wavefrom inversion. This paper is a sequel to that paper in that we demonstrate the applicability of the technique to a field data set from a gas field in the Gulf …