Acoustic and elastic properties of calcareous sediments across a siliceous diagenetic front on the eastern U.S. continental slope

Any reference should be made as "Guerin, Gilles 2000. Acoustic and Thermal Characterization of Oil Migration, Gas Hydrates Formation and Silica Diagenesis, PhD. Thesis, Columbia University"
 

1.1 Abstract

The Ocean Drilling Program drilled Hole 904A on the upper continental slope offshore New Jersey through 230 m of Eocene chalks and an opal-A/opal-CT diagenetic front. In addition to a suite of standard logging tools, a dipole sonic tool was deployed to determine the in situ shear sonic velocity and the elastic properties of these slow, highly porous, sediments. While porosity decreases by 20% across the diagenetic front, the compressional velocity and the density are observed to increase by 30%, the shear sonic velocity and the bulk modulus by about 60%, and the shear modulus by almost 300%. It is shown that shear and compressional velocities are both controlled by porosity and consolidation, and that existing models for the prediction of elastic properties and shear velocity from standard logs are still valid through the diagenetic front. The strong effect of diagenesis on rigidity makes the shear velocity log a highly sensitive indicator of pore shape and pore filling transformations.

1.2 Introduction

The Ocean Drilling Program Leg 150 drilled and collected logging data at 4 Sites (902, 903, 904 and 906) on the eastern U.S. continental slope where it recovered a thick section of Eocene chalks [Mountain, Miller, Blum, et al., 1994]. At all four sites, a diachronous high amplitude seismic reflector was recovered which had previously been identified as a siliceous diagenetic front below which biogenic opal-A has been converted to porcellanite [Tucholke, Vogt et al., 1979; Poag, Watts, et al., 1987]. In this paper we investigate the in situ elastic properties of the chalks through the diagenetic front and their relationship to porosity and consolidation.

Compressional velocity (Vp) and density (r) are both routinely measured in situ and on core samples. These data are critical for log-seismic correlation (synthetic seismograms) and for the identification of seismic sequence boundaries [e.g. Goldberg et al., 1987], but in order to describe the elastic behavior of the formation, its dynamic elastic constants must also be known. The shear modulus, µ, and the bulk modulus, K, can be defined by:

     and      (1)

where Vs is the shear wave velocity. Because neither of the elastic moduli can be directly measured in situ, the measurement of r, Vp and Vs is necessary for the resolution of equation (1). When using standard acoustic logging tools, however, there is no detectable shear wave in slow, poorly consolidated formations such as on the New Jersey continental slope, and it is not possible to measure Vs directly [Toksöz and Cheng, 1991]. Dipole sonic logging technologies, which initiate flexural waves in the formation using a directional pressure source [Chen, 1988], usually allow for estimates of Vs through waveform analysis, over a wide range of wave speeds and formation porosity. During ODP Leg 150, a commercial dipole tool was used in addition to a suite of standard tools which made possible the estimation of Vs and of the elastic moduli in these high-porosity sediments.

In previous work on the New Jersey continental slope at adjacent DSDP Sites 612 and 613, Poag, Watts, et al. [1987] and Wilkens et al. [1987] identified a sharp change in core and log physical properties related to the conversion of biogenic Opal-A to porcellanite. Wilkens et al. [1987] linked the density and porosity changes to the transformation in the micro-structure of the chalks: the re-precipitation of the dissolved biogenic silica (opal-A) into silica lepispheres (opal-CT). This diagenetic transformation cements the matrix of the sediments and partially fills the large calcareous fossils, reducing porosity by about 20% over just a few meters and without affecting the overall chemical composition of the grains or their density.

The dipole sonic tool used during Leg 150 enabled direct measurement of Vs in the chalks and through this same diagenetic front. We discuss the results at Site 904 and at nearby Site 902 and the suitability of current models for velocity-porosity relationships and sediment elastic properties.

1.3 Logging Results

In Hole 904A, an extensive suite of logs was run over a 230 m-thick calcareous chalk section, including porosity (F), r, Vp and Vs measurements. The borehole conditions were excellent over this interval and the data clearly show the opal-A/opal-CT diagenetic front penetrated at 525 meters below seafloor (mbsf). In Figure 1.1 porosity (F) decreases by 20 % across this boundary, density and Vp increase by about 30% and Vs by 60%. Mountain, Miller, Blum, et al. [1994] observe that the grain density in core samples remains unchanged, or even decreases slightly, below 525 mbsf, which suggests that the increase in bulk density and compressional velocity are controlled uniquely by changes in the porosity and the nature of the pore space.

Figure 1.1

The strong correlation between velocities and porosity in Figure 1.2a and 2b underlines the velocity dependence on porosity. The chalks drilled in Hole 904A have been divided into three distinct subunits, identified by Mountain, Miller, Blum et al., [1994]: above (+) and below (o) an abrupt drop in clay content at 419 mbsf; and below (*) the diagenetic boundary at 525 mbsf (see Figure 1.1). Least square regressions distinguish different trends between these three groups of data (Table 1.1). The increase in slope with increasing depth and decreasing porosity can be attributed mainly to compaction. Part of this increase between (+) and (o) may also be attributed to a clay content reduction which increases the contribution of higher-velocity calcite and silica minerals. Data recorded below the diagenetic front (*) are particularly distinct, showing reduced scatter (increased correlation coefficients in Table 1.1), but no significant change in chemical composition. This behavior is interpreted to be characteristic of the change in mechanical state of the sediments.

Figure 1.2
In the past, a lack of direct Vs measurements has precluded the development of empirical relations between Vs and porosity. The shear velocity and the rigidity of marine sediments have not been easily computed from their physical components and properties because of the diversity in grain sizes, shapes, and other parameters affecting particle interlocking and sediment rigidity [Hamilton, 1971]. Figure 1.2b shows, however, that a consistent correlation exists between F and Vs. The relationships for 1/Vs and 1/Vp vs. Fthrough the chalks and the diagenetic front are similar with a reduction in scatter and a steepening of slopes as F decreases with depth (Table 1.1). Figures 2a and 2b show that Vs increases faster than Vp with decreasing porosity, although this is more apparent in the evolution of their ratio Vp/Vs as a function of depth. Figure 1.3 (column a) shows Vp/Vs, representative of the Poisson's ratio and of the stiffness of the material, steadily decreasing with depth as a result of compaction, and dropping sharply from 3.0 to 2.5 at the diagenetic front. This change is consistent with the stiffening of sediments during the early stages of diagenesis also observed by Wilkens et al.[1992].

Having acquired log data for Vp, Vs and r, equation (1) can be used to estimate µ and K, the shear and bulk moduli, as a function of depth. Figure 1.3 (column b) illustrates the sharp increase in the two moduli and the mechanical change occurring through the diagenetic front at 525 mbsf. The relative increase in bulk modulus is about 60%, and the shear modulus increases by about 300 %. Therefore a five times greater relative increase in rigidity than in incompressibility results from diagenesis in these sediments.

Figure 1.3







1.4 Velocity-Porosity Models

1.4.1 Models description

Several models have been proposed for the determination of elastic moduli in sediments from porosity, in the absence of shear velocity measurements. Particularly, Wood [1941] and Gassmann [1951] derived equations that have been used by different authors to address the changes in elastic properties of marine sediments through consolidation. Wood assumes that the shear modulus of unconsolidated sediments, considered to be high porosity aggregates of fluid and grains, is zero and that the bulk modulus is the geometric average of the bulk moduli of the grains (Ks) and of water (Kw):

  and .                     (2)

Hamilton [1971], Hamilton et al.[1982] and Wilkens et al. [1992] suggest that this formulation is not valid for marine sediments, which have some rigidity, but still can be used to obtain a maximum estimate of the shear velocity. Assuming that the difference between compressional velocities estimated from Wood's equation and measured velocities can be attributed to non-zero rigidity, the dynamic shear modulus predicted by Wilkens et al.[1992] is µ = 3/4(rVp2 - KWood).

In Gassman's equation [1951], the bulk modulus is considered as the combination of three moduli: Ks, Kw and the frame bulk modulus (Kf) which represents a rigidity component introduced by grain-to-grain contacts:

   with .                     (3)

Hamilton [1971] and Hamilton et al.[1982] defined an empirical relation between Kf and F for calcareous sediments similar to those encountered on the eastern U.S. continental slope:

log(Kf (Pa)x 10-9) = 1.87355 - 3.41062F .                     (4)

Once K has been estimated from porosity using equation (2) (Wood's model) or using equations (3) and (4) (Gassmann-Hamilton's model), µ and Vs can be deduced from Vp and r measurements.

Using the logging data recorded at Site 904, we have compared the results of these two models to the Vs log data in Figure 1.3 (column c). Both models assume a grain bulk modulus Ks= 74.74 x 109 Pa (mostly calcite) and Kw= 2.25 x 109 Pa (after Hamilton et al., [1982]). As expected, the shear velocity estimated from Wood's model is higher than the log data, while on average the results from the Gassmann-Hamilton model agree with the log data within 10 %. In the absence of other in situ measurements of Vs, this observation supports the prediction of Vs in these sediments using the Gassmann-Hamilton model. We use this model for the estimation of Vs in adjacent Hole 902D, where no direct shear waves were recorded.

1.4.2 Hole 902D

Hole 902D was drilled 7 km upslope from Site 904 through similar, uniform lithologies and with good hole conditions. Only a monopole sonic tool was used in this hole and measured Vp, but no direct shear waves were detected. The second arrival detected was almost invariably the fluid wave at a constant velocity of Vf=1.6 km/s, as shown in Figure 1.4. In order to estimate Vs at this site, we applied the Gassmann-Hamilton model using the density, Vp and F logging data measured in Hole 902D. The results are also shown in Figure 1.4. Based on the comparison with Hole 904A, this estimate reasonably represents the shear velocity in Hole 902D because: (1) the limited range of r, F or Vp data is similar at Site 904; (2) Vs is clearly less than Vf and could not have been measured with a monopole source in this hole; and (3) in one interval where Vs is higher than Vf (at about 613 mbsf), a refracted shear wave was measured and agrees with the model.

Figure 1.4

A limitation of this method for estimating Vs is that it requires similar lithologies. When direct Vs measurements are not available, however, a reasonable estimate can be provided from standard logs by a well-constrained formulation. With more direct measurements of Vs (in combination with Vp, r and F) in different types of sediments, the empirical parameters in equations (3) and (4) could be defined broadly over a wide range of minerals and porosity, which could eventually allow for estimates of Vs in any environment using conventional logging data.

1.5 Elastic Moduli and Porosity

Figure 1.5 displays the relationships between porosity and the bulk and shear moduli calculated from logging data at Site 904, and the curve representing the Gassmann-Hamilton model in calcareous sediments. The same symbols are used as in Figure 1.2, except for µ below the diagenetic front (filled squares). The bulk moduli data agree reasonably well with the model and confirm its validity at this site, even through the diagenetic front. In itself, such a strong correlation points to the porosity as a dominant parameter in the elastic properties of the chalks.

Figure 1.5

This figure also illustrates the difference in the elastic moduli across the diagenetic front. While K and the gradient dK/dF increase progressively with decreasing F, m increases steadily and dµ/dF changes sharply at F= 45%, from 2.19 to 8.37 (Table 1.2).

1.6 Discussion

The discontinuity in rigidity may be explained by the formation of silica lepispheres in open pore space due to diagenesis. The abrupt increase in stiffness at the diagenetic front is clearly illustrated by the increase in Vp/Vs ratio (Figure 1.3a) and the discontinuous change in dµ/dF suggests a coincident discontinuity in the mechanical properties of the chalks and in the lithification process. Whether this process is actually discontinuous or not, however, the increase in rigidity is the most significant change in the elastic properties of the chalks through diagenesis.

The Gassmann-Hamilton model accurately predicts shear sonic velocities through the entire chalk interval and through the diagenetic front in the absence of direct Vs measurements. This suggests that there is no major change in the nature of the consolidation process, or in the grain structure, due to silica diagenesis. If there is a change, its quantitative effect on the elastic properties of the sediments is limited, so that they remain predictable by the same algebraic formulation shown in equations (3) and (4).

Wood's model, by comparison, predicts velocities less satisfactorily than observed by Wilkens et al. [1992] for the same lithological formation a few miles downslope at Site 613, and the velocities which result from using this model are much higher than the velocities measured in situ by the dipole sonic tool. However the difference between the model and the Vs log increases with depth, or greater consolidation, as also observed by Wilkens et al. [1992]. The greater burial depth and higher consolidation of the chalks at Site 904 than at Site 613 may partially explain the difference in the application of Wood's model at the two sites. There is also a considerable difference between the results of the inversion described by Wilkens et al. [1992] and the Vs logging data, particularly above the diagenetic front where the inversion process produces shear velocity values much higher than the present logging data. Wilkens et al. [1992] note that the inversion process breaks down at very low shear wave velocities (i.e. lower than about 800 m/s) which accounts for most of the interval above the diagenetic front both at Sites 904 and 613. Consequently we suggest that the results of the inversion study overestimate Vs at Site 613 above the diagenetic front, and that Wood's model overestimates Vs at both sites. Dipole log measurements and alternatively the Hamilton-Gassmann model are more representative of Vs in these marine sediments.

1.7 Conclusion

The primary role of porosity on the evolution of the mechanical properties of carbonates has been here observed and documented over an intermediate porosity range from 30 to 65% on the eastern U.S. continental slope. The Hamilton-Gassmann model provides a reliable empirical estimation of elastic moduli and Vs from porosity and conventional logs and agrees well with the dipole sonic log data. More in situ shear velocity measurements are needed to calibrate similar models in a variety of sedimentary environments.

At Site 904, the infilling of foraminifer chambers by diagenetic re-precipitation of Opal-CT from dissolved Opal-A generates a noticeable change in the lithification process, whose most apparent mechanical effect is a sharp increase in sediment rigidity. Shear velocity logs can indicate, and to some extent quantify, this specific diagenetic transformation in the absence of sediment core analyses. With adequate data for calibration, these observations could be extended to other mechanical or geochemical processes that affect grain-to-grain contacts and alter the pore space and the matrix through different types of transformations, such as the formation of ice (gas hydrates) or the precipitation of diverse types of minerals. The shear velocity log is a highly sensitive indicator of such processes that primarily affect the sediment rigidity.



References Chen, S.T., Shear wave logging with dipole sources, Geophysics, 53, 659-667, 1988.

Gassmann, F., Elastic waves through a packing of spheres, Geophysics, 16, 673-685, 1951.

Goldberg, D. S., Wilkens, R. H., and Moos, D., Seismic modeling of diagenetic effects in Cenozoic marine sediments at Deep Sea Drilling Project Sites 612 and 613, in Poag, C. W., Watts, A. B., et al., Init. Repts. DSDP, 95, 589-599, 1987.

Hamilton, E. L., Elastic properties of marine sediments, J. Geophys. Res., 76, 579-601, 1971.

Hamilton, E. L., Bachman, R. T., Berger, W. H., Johnson, T. C., and Mayer, L. A., Acoustic and related properties of Calcareous deep-sea sediments, J. Sed. Pet., 52(3), 733-753, 1982.

Mountain, G.S., Miller, K.G., Blum, P., et al., Proc. ODP, Init. Repts., 150, College Station, TX (Ocean Drilling Program), 1994.

Poag, C. W., Watts, A. B., et al., Init. Repts. DSDP, 95, Washington (U.S. Govt. Printing Office), 1987.

Toksöz, M.N., and Cheng, C.H., Wave propagation in a borehole, in Shear waves in marine sediments, edited by J.M. Hovem, M.D. Richardson, and R.D. Stoll, pp. 257-265, Kluwer Academic Publishers, 1991.

Tucholke, B.E., Vogt, P.R., et al., Init. Repts., DSDP, 43, Washington (U.S. Govt. Printing Office), 1979.

Wilkens, R. H., Schreiber, B. C., Caruso, L., and Simmons, G., The effects of diagenesis on the micro structure of Eocene sediments bordering the Baltimore Canyon Trough, in Poag, C. W., Watts, A. B., et al., Init. Repts. DSDP, 95, Washington (U.S. Govt. Printing Office), 527-536, 1987.

Wilkens, R. H., Cheng, C. H., and Meredith, J.A., Evaluation and prediction of shear wave velocities in calcareous marine sediment and rocks, J. Geophys. Res., 97, 9297-9305, 1992.

Wood, A.B., A textbook of sound, G. Bell and Sons, Ltd, London, 1941.
 


Tables
 
 
 

Table 1.1: Linear regressions for slowness vs. porosity (with correlation coefficients).

Unit
1/Vp
1/Vs
+
0.475 + 0.159F (0.441)
1.217+ 1.079F (0.404)
o
0.378 + 0.310F (0.636)
0.621 + 2.064F (0.586)
*
0.231 + 0.586F (0.817)
-0.324 + 3.897F (0.855)

 
 
 
 
 

Table 1.2: Linear regressions for elastic moduli vs. porosity (with correlation coefficients in brackets).

Unit
Bulk Modulus (K, in 109Pa)
Shear Modulus (m, in 109Pa)
+
7.21 - 4.50F [0.410]
1.05 - 0.95F [0.429]
o
10.31 - 10.17F [0.642]
1.73 - 2.19F [0.636]
*
15.26 - 19.76F [0.777]
4.69 - 8.37F [0.821]