3rd International Workshop on Rock Physics th th 13 – 17 April 2015 Perth, Western Australia Relating static stiffness to ultrasonic, sonic and seismic velocities Erling Fjæra, Rune M. Holtb and Anna M. Stroisza. a SINTEF Petroleum Research, Trondheim, Norway; Norwegian University of Science and Technology, Trondheim Norway. Contact email: [email protected]. b Summary This paper outlines the complexity in relations between static stiffness and stiffness derived from elastic wave velocities, and shows results from modelling efforts aiming to fill some important gaps in the mathematical description of these relations. The paper also points to some interesting spin-offs of this modelling effort. Introduction Rock physics is to a large extent based on linear elasticity. Within this framework, there is a direct and simple relation between static stiffness and elastic wave velocities. Numerous observations [1-4] show however that this relationship is not so simple in real rocks. A consequence of this is that the estimation of static stiffness on the basis of sonic or seismic velocities – for prediction of reservoir compaction, for instance – is challenging and burdened with large uncertainty. Reliable observations of relations between static stiffness and elastic wave velocities are for the most part done in laboratories, on small samples at ultrasonic frequencies. This implies that we also need to understand the relation between ultrasonic and sonic or seismic velocities in order to build relations between static stiffness and sonic or seismic velocities. Several factors affect the relations between static stiffness and elastic wave velocities, and have to be controlled and accounted for [5]. The most important ones are: - strain rate, which differs by four to six orders of magnitude between seismic and ultrasonic frequencies, and also differ largely for different static deformations; - rock volume involved, which may differ by up to 12 orders of magnitude or even more, from a cube constrained by the ultrasonic wavelength to the volume of a reservoir; - drainage conditions, which for elastic waves are always undrained in a fully saturated rock volume, but are often drained under static deformations; - anisotropy, which implies that stiffness as well as velocities may differ largely with orientation of the rock; and – non-elastic processes, which may have a dominating impact on static stiffness, but nearly no impact on the wave velocities. A reliable routine for estimation of static stiffness from sonic or seismic velocities requires that all the factors mentioned above are considered and properly accounted for. Some of these elements are fairly 1 3rd International Workshop on Rock Physics, Perth, 13th–17th, April 2015 well understood and described mathematically in well-established models. Still, these effects are often ignored due to lack of information (as well as lack of attention). Other elements – in particular nonelastic processes and strain rate effects – have not yet reached the same level of understanding and mathematical modelling. Strain rate and strain amplitude The difficulties in measuring dispersion in the range from seismic to ultrasonic frequencies has for a long time hampered the development of static/dynamic relations, however several laboratories have now developed equipment and procedures that allow for such measurements [6-9]. Results obtained so far point towards partial saturation as a major cause for dispersion in this frequency range [5]. The strain rate of static deformations in laboratory tests is comparable to the strain rate of a seismic wave [5] (although in field situations the relevant static strain rates may be significantly lower). The strain rate is therefore primarily a source of dispersion, and to a lesser extent a cause for the differences between actual static stiffness and stiffness derived from sonic or seismic velocities. The most important differences between static deformations and the dynamic deformations associated with a passing elastic wave are the strain amplitude, and the fact that the dynamic deformations are oscillatory while the static deformations are monotonic. This implies that non-elastic processes may have a strong impact on static stiffness, while their impact on dynamic stiffness can usually be ignored. A model describing the impact of such processes during first, monotonic loading [4] has been shown to give a good description of the relation between static and dynamic stiffness for sandstone [10] and shale [11], and is used successfully to predict strength and static stiffness from wireline log data [12]. A recent extension of this model [13] also describes how the creep process may transform from transient creep through steady state creep into accelerating creep. This model has several interesting applications, for instance scaling between strain rates in laboratory and field, prediction of consequences of strain rate changes, and interpretation of complex laboratory tests. Two types of non-elastic processes are assumed to dominate this behaviour: crushing of pointed contacts, and frictional slip. Whereas both processes are expected to be active during first loading, only frictional slip will be significantly active during unloading and reloading [14]. Moreover, frictional slip activity builds up gradually over a period of monotonous loading or unloading, but starts from zero again when the loading direction is reversed. A consequence of this behaviour is that the relationship between static and dynamic stiffness, which changes slowly with monotonously increasing stress, changes abruptly when the stress path is reversed and maintains a somewhat complicated behaviour during an unloading/reloading cycle, as shown in Figure 1. The behaviour during unloading has been utilized for estimation of dispersion in the range from seismic to ultrasonic frequencies [5]. That method was derived on the basis of observations. Based on the model by Nihei et al. [15] we here establish a description of how the frictional slip process affects the static stiffness in the vicinity of a turning point on the stress path. This model predicts a.o. that the initial unloading gradient of the non-elastic compliance ( S H ) to stress ( σ z ) depends on the stress at the turning point. Figure 2 shows that these predictions match quite well with observations for a set of measurements on Berea sandstone. The results support the assumption that only frictional slip is active during unloading, at least initially. 2 3rd International Workshop on Rock Physics, Perth, 13th–17th, April 2015 This model description allows us to subtract the frictional slip process from the static compliance during loading, using the model calibrated during unloading. The crushing process which is also active during loading may then be studied separately. This work is in progress. Static 60 10 40 Unloading 5 First loading 0 40 70 100 20 Continued first loading 0 160 130 Axial stress [MPa] 80 15 Re-loading Stiffness [MPa] 100 Dynamic 20 Time [min] Figure 1. Static and dynamic stiffness, measured in a core plug of Castlegate sandstone. 1.4 ∆SH/∆σz [MPa-2] 1.2 1 0.8 0.6 0.4 0.2 0 0 20 40 60 80 Axial stress at turning point [MPa] Figure 2. Initial unloading gradient of non-elastic compliance to stress, versus stress at the point where the stress path is reversed. The data points originate from measurements on Berea sandstone. Solid line: model predictions. Conclusions A model describing the development of frictional slip during unloading appears to give a relevant description of the relation between static and dynamic stiffness, thus supporting the assumption that this is the only active non-elastic process during unloading. The model may be used to subtract frictional slip from the static compliance during loading, thereby allowing for separate studies of crushing processes that are active during loading. 3 3rd International Workshop on Rock Physics, Perth, 13th–17th, April 2015 References 1. Simmons, G. and Brace, W.F. (1965) Comparison of static and dynamic measurements of compressibility of rocks. Journal of Geophysical Research, 70, 5649-5656. 2. King, M.S. (1970) Static and dynamic elastic moduli of rocks under pressure. Proc. 11th US symposium on Rock Mechanics. 329-351. 3. Jizba, D. and Nur, A. (1990) Static and dynamic moduli of tight gas sandstones and their relation to formation properties. SPWLA 31st Annual Logging Symposium, Paper BB. 4. Fjær, E. (1999) Static and dynamic moduli of weak sandstones. In: Amadei B, Krantz RL, Scott GA, Smeallie PH (eds.) Rock mechanics for industry. Balkema, pp 675-681. 5. Fjær, E., Stroisz, A.M., Holt, R.M. (2013) Elastic dispersion derived from a combination of static and dynamic measurements. Rock Mech Rock Eng 46:611–618. DOI 10.1007/s00603013-0385-8. 6. Duranti, L., Ewy, R., Hofmann, R. (2005) Dispersive and attenuative nature of shales: multiscale and multifrequency observations. SEG Expanded Abstracts. 24:1577-1580. 7. Hofmann, R. (2005) Frequency dependent elastic and anelastic properties of clastic rocks. PhD Thesis, Colorado School of Mines. 8. Batzle, M.L., Han, D.-H., Hofmann, R. (2006) Fluid mobility and frequency-dependent seismic velocity – Direct measurements. Geophysics 71:N1-N9. doi: 10.1190/1.2159053. 9. Bauer, A., Szewczyk, D., Hedegaard, J., and Holt, R.M. (2015) Seismic dispersion in Mancos shale. Abstract submitted for 3IWRP, Perth, Australia. 10. Li, L., Fjær, E. (2012) Modeling of stress-dependent static and dynamic moduli of weak sandstones. Journal of Geophysical Research – Solid Earth, 117, WA103-WA112. 11. Holt, R.M., Nes, O.-M., Stenebråten, J.F., Fjær, E. (2012) Static vs dynamic behavior of shale. ARMA 12-542. 12. Woerl, B., Wessling, S., Bartetzko, A., Pei, J. and Renner, J. (2010) Comparison of methods to derive rock mechanical properties from formation evaluation logs. ARMA 10-167. 13. Fjær, E., Larsen, I., Holt, R.M., Bauer, A. (2014) A creepy model for creep. ARMA 14-7398. 14. Stroisz, A.M. (2013): Nonlinear elastic waves for estimation of rock properties. PhD Theses, NTNU. 15. Nihei, K.T., Hilbert Jr, L.B., Cook, N.G.W., Nakagawa, S., Myer, L.R. (2000) Friction effects on the volumetric strain of sandstone. Int J Rock Mech & Min Sci 37:121-132. 4
© Copyright 2026 Paperzz