Yuefeng Sun

Quantification of pore structure and its effect on sonic velocity and permeability in carbonates

Carbonate rocks commonly contain a variety of pore types that can vary in size over several orders of magnitude. Traditional pore-type classifications describe these pore structures but are inadequate for correlations to the rocks physical properties. We introduce a digital image analysis (DIA) method that produces quantitative pore-space parameters, which can be linked to physical properties in carbonates, in particular sonic velocity and permeability. The DIA parameters, derived from thin sections, capture two-dimensional pore size (DomSize), roundness (), aspect ratio (AR), and pore network complexity (PoA). Comparing these DIA parameters to porosity, permeability, and P-wave velocity shows that, in addition to porosity, the combined effect of microporosity, the pore network complexity, and pore size of the macropores is most influential for the acoustic behavior. Combining these parameters with porosity improves the coefficient of determination (R2) velocity estimates from 0.542 to 0.840. The analysis shows that samples with large simple pores and a small amount of microporosity display higher acoustic velocity at a given porosity than samples with small, complicated pores. Estimates of permeability from porosity alone are very ineffective (R2 = 0.143) but can be improved when pore geometry information PoA (R2 = 0.415) and DomSize (R2 = 0.383) are incorporated. Furthermore, results from the correlation of DIA parameters to acoustic data reveal that (1) intergrain and/or intercrystalline and separate-vug porosity cannot always be separated using sonic logs, (2) P-wave velocity is not solely controlled by the percentage of spherical porosity, and (3) quantitative pore geometry characteristics can be estimated from acoustic data and used to improve permeability estimates.

Changes of shear moduli in carbonate rocks: Implications for Gassmann applicability

In laboratory experiments we measured the saturation effects on the acoustic properties in carbonates and the results question some theoretical assumptions. In particular, these laboratory experiments under dry and wet conditions show that shear moduli do not remain constant during saturation. This change in shear modulus puts Gassmanns assumption of a constant shear modulus into question and also explains why velocities predicted with the Gassmann equation can be lower or higher than measured velocities.

Seafloor Properties From Penetrometer Tests

Quasi-static and freely falling dynamic penetrometers are currently in extensive use for measuring the mechanical properties of sediments composing the littoral seafloor. Sediments in this zone are often inhomogeneous both laterally and with depth so that it is difficult to predict burial of mines and other objects when relying on models that assume uniform, homogeneous sediment. The results of penetrometer tests discussed in this paper show that there can be a wide spread in the penetration resistance that is measured depending on the degree of sediment inhomogeneity and the rate of penetration. Moreover, the dilative response of granular strata appears to further complicate matters because of the sudden, large changes in shear strength that can occur. As a result, mine burial models currently in use, which often rely on simple strain-rate factors and shear strength determined from experiments utilizing uniform, reconstituted sediment, do not appear to be adequate to handle real in situ conditions in many cases. The objective of this paper is to obtain a better understanding of in situ properties and how they may be incorporated into various burial models.

Pore structure effects on elastic wave propagation in rocks: AVO modelling

Pore structure in rocks at reservoir pressures can affect strongly elastic wave velocities. For a reservoir where the fluid type and lithology are known, elastic wave velocity V can be described by a two-parameter model in terms of porosity and a pore structure parameter referred to as the frame flexibility factor (γ). Using this two-parameter elastic model V(, γ), the concept of pore structure type (PST) is introduced to quantify the pore structure effects on elastic properties of sedimentary rocks. To illustrate the concept, three PSTs are defined and they have their distinct characteristics on elastic wave propagation. Through solving exact Zoeppritz equations and elastic finite-element modelling, this study indicates that it may be feasible to detect pore structure variations in reservoir rocks from seismic data via amplitude variation with offset (AVO) analysis. AVO interpretation aimed at fluid detection can be complicated where the presence of large pore structure variations is expected because the effects of pore structure on wave propagation can be much stronger than the fluid factor, especially in carbonate rocks.

Estimation of aspect-ratio changes with pressure from seismic velocities

Abstract Seismic velocities are modeled as a function of rock mineralogy, porosity, fracture density, aspect ratio and fluid saturation. We compare the model with seismic velocity measurements made by Nur and Simmons in 1969 and predict quantitative aspect ratio changes as a function of differential pressure. The velocity change with pressure is initially caused by the deformation and collapse of pre-existing cracks and pores followed by the initiation of new cracks. For low porosity Casco granite (0.25% pore and 0.45% crack), we estimate that crack aspect ratio increases from 5.9e−3 to 0.01, then decreases to 5.0e−5 for a pressure increase from 0 to 300 MPa. Most pre-existing cracks and pores close at differential pressures < 40MPa, a lower pressure than predicted with standard approximations. These model results can be used to determine in situ porosity, fluid content, and pore pressure from seismic data by inversion and for reservoir monitoring from drilling and core data.

Seismic structure of the upper oceanic crust revealed by in situ Q logs

In situ seismic attenuation is computed through 2.1 km of the upper oceanic crust in the vicinity of ODP Hole 504B. The results strongly tie crustal properties to seismic measurables and observed geological structures: we find that the attenuation can be used to define seismic layer boundaries and is closely related to the intensity of vertical heterogeneity. The in situ attenuation Q−1 consists of both intrinsic and scattering contributions, but is dominated by the scattering attenuation unless porosities are near zero, when it approaches typical estimates from seismic refraction studies. The attenuation is analytically modeled by multiple backscattering from heterogeneities observed in a sonic Vp log and is found to decrease step-wise in relatively homogeneous layers from Q=25 to Q > 300 between the top of seismic layer 2A and a sharp discontinuity at 1.3 km depth. These changes correspond with heterogeneities at 1.0–1.3 m and at 5.6–10.0 m wavelengths that we interpret to be associated with fracturing and structure of pillow basalts and lava flows in seismic layers 2A and 2B. Although seismic velocity studies suggest that the layer 2/3 boundary also occurs at about 1.3 km, the large variation in Q (140 to 460) below this depth indicates that a seismically homogeneous and uniform layer 3 has not been reached in Hole 504B. We derive an empirical relationship between attenuation and porosity Q−1=Q0−1 eβϕ, where Q0−1=0.004 and β=25, that may be applicable at other oceanic crust locations and useful for constraining seismic inversion models.

ATTENUATION DIFFERENCES IN LAYER 2A IN INTERMEDIATE- AND SLOW-SPREADING OCEANIC CRUST

In situ seismic attenuationQ−1logs are derived from borehole velocity profiles and reveal sharp boundaries between morphologies of the extrusive volcanic layers in intermediate- and slow-spreading oceanic crust.Q−1logs are calculated from the scattering attenuation associated with vertical velocity heterogeneity in Ocean Drilling Program Holes 504B and 896A and in Hole 395A, located in 5.9–7.3 Ma crust on the Pacific and Atlantic plates, respectively. Our results strongly tie crustal properties to seismic measurables and observed geological structures: we find that the scattering attenuation can be used to identify the extrusive volcanic sequence because it is closely related to changes in the degree of vertical heterogeneity. We interpret a distinct decrease in the Q−1log at the transition below the extrusive volcanic layer to correspond with the seismic layer 2A/2B boundary. The boundary is located at 465 m depth below the sea floor in both Hole 395A and 504B, although this is likely to be a coincidence of the sediment thickness at these sites. Layer 2A is estimated to be approximately 150 m thick in Hole 504B and > 300 m thick in Hole 395A. Cyclic sequences of high-porosity pillows and low-porosity massive units in the uppermost 100 m of volcanics in Hole 395A result in large velocity heterogeneities which cause > 5 times more attenuation in this layer than in Hole 504B. In Hole 896A, by contrast, fewer pillows, more massive flows, and a greater volume of carbonate veins decrease the velocity heterogeneity and attenuation significantly over only 1 km distance from Hole 504B. We conclude that the attenuation in the extrusive volcanics of the ocean crust is largely controlled by variation in local heterogeneity and morphology as well as by subsequent hydrothermal alteration. The observed differences inQ−1profiles and layer 2A thickness at these sites may be attributed to variations in the volume and duration of volcanic activity at mid-ocean spreading centers for these Pacific and Atlantic ridge segments.

Effects of Pore Structure On 4D Seismic Signals In Carbonate Reservoirs

Carbonate reservoir rocks of different pore structure have quite distinct and detectable 4D seismic signatures. Using three defined pore structure types (PST), these seismic signatures are correlated with the permeability patterns of carbonate rocks. In both the high-permeability PST3 and intermediate-permeability PST2 regions, gas injection is preferred than water injection for effective 4D seismic monitoring. In the low-permeability PST1 region where lies most of the bypassed oil, both gas and water injection options result in similar 4D seismic signal magnitude. Fluid saturation changes under in-situ reservoir conditions can be better detected by offset domain 4D seismic analysis.

Improving Porosity–Velocity Relationships Using Carbonate Pore Types

Acoustic impedance in carbonates is influenced by factors such as porosity, pore structure/fracture, fluid content, and lithology. Occurrence of moldic and vuggy pores, fractures and other pore structures due to diagenesis in carbonate rocks can greatly complicate the relationships between impedance and porosity. Using a frame flexibility factor (γ) derived from a poroelastic model to characterize pore structure in reservoir rocks, we find that its product with porosity can result in a much better correlation with sonic velocity (Vp = A − B ∗ ϕ ∗ γ) and acoustic impedance (AI = C − D ∗ ϕ ∗ γ), where A, B, C and D is 6.60, 0.03, 18.3 and 0.09, respectively for the deep low-porosity carbonate reservoir studied in this paper. These new relationships can also be useful in improving seismic inversion of ultra-deep hydrocarbon reservoirs in other similar environments.

Characterization of the upper oceanic crust using high-resolution seismic amplitude modeling

We examine the links between structural and lithological changes in the upper oceanic crust and seismic amplitude changes by applying modeling techniques in order to extend the conventional uses and limits of seismic data. Using high-resolution velocity and density borehole profiles from Ocean Drilling Program Hole 504B, we evaluate modeling methods and study the mechanisms of seismic attenuation in this layered porous environment. Good agreement of full-waveform synthetic seismograms with field seismic data is observed in travel time as well as in waveform shape and amplitude variations. Large seismic energy losses in the upper oceanic crust are observed where the formation consists of cyclic sequences of high- and low-porosity units, which demonstrates that energy loss in the upper oceanic crust is dominated by the effects of scattering alone. A fault zone and the boundary between seismic layers 2 and 3 are observed by amplitude changes on both synthetic and observed reflection records. The seismic amplitude, therefore, may indeed be used to map porosity and other lithologic variations.