Search

Reservoir analysis of seismic data is performed on migrated seismic images, which represent the spatial variability of the medium’s reflectivity. Intuitively, the process of migration rotates the wavelet so that it is normal to the imaged reflectors. Processes used in reservoir analysis such as deconvolution, inversion and warping for time-lapse analysis need to follow the structure of the data. The traditional 1D (vertical) convolutional approach does not honour this directivity. We introduce a wave equation based approach which provides an effective platform for structurally consistent reservoir analysis. This includes applications such as wavelet extraction, elastic inversion, warping and 4D time-strain inversion.

Induced polarization (IP) effects observed in airborne time domain electromagnetic (TEM) survey data offer information on the chargeability of the subsurface in addition to conductivity derived from TEM data. However the IP effect is generally weak and obscured in the total TEM response. As a result, the typical inverse transient associated with IP effect does not always manifest itself in the EM response. This causes difficulty for algorithms that rely on the inverse transient to estimate the chargeability of the subsurface. We have developed a robust method that decomposes the total electromagnetic response into a fundamental (inductive) and a polarization component and we estimate apparent chargeability from the polarization component. In this paper, we discuss the method and illustrate its effectiveness with examples

We propose a broadband processing workflow that is purely data driven and applicable to both conventional towed and variable depth towed marine streamer data. Three important ingredients of our workflow are highlighted: (1) deghosting; (2) FS-QTomo and A-QTomo; (3) prestack depth Q migration (Q-PSDM). We first demonstrate through a simple synthetic example how FS-QTomo can benefit from the removal of source and receiver notches in obtaining more accurate Q field. We then utilize a field data example from offshore Malaysia that exhibits severe absorption resulted from gas pockets to show how our workflow can preserve the fidelity of amplitude, frequency and phase of the data.

We demonstrate by example, a joint deterministic inversion – facies estimation procedure which makes minimal prior assumptions and obviates the need for the building of a low frequency model (LFM) which interpolates log curves between wells. We then apply the same procedure to geostatistical inversion and show that facies can successfully convey low frequency trend information between the two domains.

Contrary to conventional seismic that focuses in a narrow band in middle frequencies, broadband seismic aims at acquiring information within the whole bandwidth emitted by the source. A good knowledge of the source signature is thus of paramount importance in broadband processing. However, the measurement of source signature is difficult and the modeling results are not satisfactory in the low frequencies. Another source of wavelet dispersion is that seismic waves suffer from energy loss and phase dispersion due to intrinsic and apparent absorption which vary laterally and in depth. To address these challenges, a data-driven time variant non-Gaussian wavelet estimation which is sensible to phase is developed. As demonstrated on a North Sea broadband dataset, the proposed technique is able to produce zero-phase data on an enlarged bandwidth, both on low and high frequencies.

A great deal of effort has been expended to improve the amplitude reliability of migration. The similarity of reverse time migration (RTM) to the gradient of full waveform inversion (FWI) indicates that preserved amplitude RTM can help improve FWI. We develop the theoretical derivation of an improved gradient for FWI based on common shot preserved amplitude RTM. We validate our approach on the Marmousi II model and the Chevron SEG 2014 dataset, showing that it significantly improves the convergence rate of FWI.

Here we develop and present a 4D processing flow including ΔSΔR (the sum of the mis-positioning of the sources and receivers from baseline to monitor survey) thresholding and parallel pairwise binning with transfer operators to improve the quality of the 4D products while optimizing the 4D quality for all possible time in a multi monitor 4D context.

Simultaneous shooting increases acquisition efficiency by activating more than one source at the same time. This introduces blending noise that typically needs to be removed before data processing. We present a new deblending algorithm based on iterative annihilation filtering. The strategy attenuates coherent energy using a modified sparse τ-p transform following which the remaining energy is re-aligned to form the deblended output. The method is tested on a numerically blended dataset using real short and long offset narrow azimuth data. The results show the residual blending noise to be ~25 dB down.

Remotely detecting information about fractures and the stress field is an important objective in the development of unconventional and tight hydrocarbon reservoirs. Fractures and stress cause the earth to become anisotropic which is seismically observable. By observing the P-wave seismic amplitude variation with offset and azimuth (AVOAz) it is possible to infer the presence of fractures and their orientation. Unfortunately, the estimate of the fracture orientation is non-unique with two solutions 90 degrees apart. This issue is well known in the case of the near-offset AVOAz inversion, but is also true for the far-offset approximation. In the case of the far-offset approximation, the azimuth ambiguity also leads to biases in the remaining parameter estimates. This paper explores using geologic and rock physics constraints to resolve this issue. A priori information about the horizontal stress field and the form of the anisotropy are used to determine the symmetry axis orientation for both the near-offset and far-offset AVOAz inverse problems.