The tight gas sand reservoirs are low permeability accumulations, with ordinarily less than 0.1 mD, produced by hydraulic fracture treatment under satisfactory oil market conditions Though horizontal drilling wells over natural fractures are the norm, multilateral wells are the standard practice to stimulate reservoirs, increase efficiency, and reduce migration distances from the formation to the wells. Hence, the development of new technologies is a crucial factor to improve the detection of sweet spots and optimize horizontal well spacing
Geoscientists have contributed to the exploration, evaluation, drilling, and completion stages of tight gas reservoirs. By using seismic attributes to determine the fracture conductivity and complexity,and monitor the drilling. Consequently, seismic attributes collaborate in the sustainable economic development of tight gas reservoirs.
Post-stack seismic attributes
Seismic attributes are instruments to deduce geology from seismic by finding, revealing, or quantifying a specific feature, pattern, or property. Their nature could be geological, geophysical, or mathematical. Moreover, the most useful attributes are unique, comparable, easy to use, and geologically meaningful. Among them, the post-stack seismic attributes are easy to implement, present low complex computations, and are remarkable structural and stratigraphic properties’ quantifier. Nevertheless, they lack direct lithology indicator
Although nowadays exist more than 200 post-stack seismic attributes, there are two broad categories for a simplified classification. The morphological and reflectivity quantification
The morphological attributes derive the information from azimuth and dip of the reflector, and correlates with fractures, faults, channels, diapirs, and carbonates. The reflectivity attributes gather the information from the amplitude, waveform, and illumination angle of the reflector. They are related to lithology, reservoir thickness, and fluid content
The complex seismic trace analysis quantifies reflectivity. They locate and analyze sweet spots by blending color attributes. For instance, the weighted instantaneous frequency is a physical attribute that indicates longer wavelength variations, because gas-bearing sands attenuate the seismic signal, it can characterize tight gas sand reservoirs. Another example is the RMS amplitude attribute, that is a useful indicator of the sand tuning thickness effect. However, broader sandstone areas with high amplitude do not always correlate well with production sandstones because of tuning effect. Therefore, the integration of a higher number of attributes leads to more reliable discrimination.
On the other hand, seismic amplitudes have a strong correlation with porosity and fluid saturation because the velocity and density changes affect the properties of the reservoir. Accordingly, the inversion of seismic amplitudes born in the mid-1970s converts seismic traces into acoustic impedance, velocity and density information. Impedance values are a direct indicator of lithology, porosity and permeability features of the reservoir. Moreover, the inverted P-wave volume is transformed into permeability and porosity based on well log statistics.
Another relevant factor in the natural fractures characterization and occurrences because they give conduits for gas migration and enhance the permeability and productivity of the reservoir. Hence, reflectivity seismic attributes and morphological seismic attributes, like coherence and curvature, can jointly understand the fault system
The similarity or coherence attribute measures the seismic waveforms correspondence from one trace to another, highlighting locations of low consistency, seen as lateral differences in seismic response. The coherence detects and image faults, fractures, and changes in stratigraphic feature. Low-coherence relates to variations in structure, stratigraphy, lithology, porosity, or the presence of hydrocarbons
Additionally, case studies show that coherence combined with spectral decomposition providing further enhancement for structural feature detection. The spectral decomposition separates the seismic response into different spectral bands. Enabling the detection of subtle structural and stratigraphic features located at specific frequency components, ordinarily buried within broadband data
The curvature attributes measures the seismic reflection bending or dip changes along an specific azimuth, it is independent of the seismic amplitude or coherence and has a direct relationship with the fracture intensity. Between the different families, the curvature gradient attributes extract additional information for fracture prediction and obtain a better fault description in fractured reservoirs. Specifically, in conjugation with further attributes, curvature can locate faults and fracture zones, phase change attributes can depict sedimentary margins, and strong amplitude attributes can reflect the effective sand distribution on the area There are two widely useful attributes, the most positive curvature to highlight high structural features (i.e., mounds, folds, and flexures), and the most negative curvature identify sandstone channel edges.
The texture classification based on the Gray-level Co-occurrence Matrix (GLCM), measures the combinations of pixel brightness values in a digital image. This alternative statistic approach is less sensitive to the seismic waveform compared with coherence and curvature. It can delineate faults and fractures with preferred orientation.
Additionally, GLCM can show faults and fracture intensities, strikes, dips, and highlights features in 2D and 3D (e.g., rollover structures and listric faults). As well as, identify and illuminate seismic facies (e.g., channel structures and within sedimentary facies). Among the primary GLCM attributes, four statistical measurements (Energy, entropy, contrast, and homogeneity) are commonly use in seismic characterization and discrimination without redundancy Additionally, low values of GLCM entropy attribute are related to a sandstone response because sandstone channel textures are the acoustic expression of a fluvial-deltaic facies
Multi-linear regression and Neural Networks.