Research challenges

i. present uplift and Neogene tectonism

The uplift rate residuals identified by Fjeldskaar et al. (2000) are significantly larger than the time-averaged Neogene rates. In this project we will undertake a refined rebound modeling, which will constrain these residuals of the tectonic component. The computations will be done by global models described in Cathles (1975) and regional models described in Fjeldskaar et al. (1997) with high spatial resolution (10 km; cf. Fig. 1b) and glacial history of every 1000 years. We will search for similar anomalies in Novaya Zemlja and Svalbard areas (Fig. 1b), in an attempt to infer  the lateral extent and magnitude of the present rate of the Neogene tectonic movements. This objective will be supplemented by 2D basin modeling with computation of vitrinite reflectance and quartz cement (method of  Walderhaug et al., 2000) and calibrated to well information.

ii. mapping of erosion and isostasy in 3 glacial stages

Previous attempts to map the regional Plio-Pleistocene erosion and sedimentation were made e.g. by Riis & Fjeldskaar (1992) and Rasmussen & Fjeldskaar (1996). More recently, A. Amantov (unpublished) has done similar compilations (Fig. 2). The challenge of estimating the amount of sediments that are removed are significant because we have only indirect data that can be used; one indirect measure being accumulated sediment volumes. Careful mass-balance control (Fig. 2) is thus important. One important, but also difficult task of this project is to divide the total Plio-Pleistocene erosion/deposition into three stages (cf. next paragraph).

Figure 2. Mapped Plio-Pleistocene sedimentation (left) and erosion of Northern Europe (right). From Amantov (unpublished).

iii. stress effects on faults due to rapid erosion

To our knowledge very few publications report modelling of stress due to glaciations and glacial erosion  in the Barents Sea, for enhancing the understanding of spillage of hydrocarbons. Løtveit et al. (2009) is an exception; they have used simple models for calculating the stress effects of erosion on the reservoirs at Loppa High and Snøhvit. In this project these models will be further developed by including more geophysical and geological data, so that the models become more realistic. The plan is to use a more realistic approach regarding the reservoir geometries, and more accurate locations and geometries for the faults in the vicinity of the reservoir. The properties of the faults are not well known in this area (the faults are avoided during drilling), so the idea is to study different scenarios for various fault properties and to estimate the change in fault permeability due to the erosion-related stress changes.

iv. Thermal conductivities of sedimentary rocks

A better estimation of the thermal conductivity is important for better predictions of the amount of erosion. A lot of approaches have been proposed to determine the effective thermal conductivity as some explicit expressions of concentrations and physical properties of the compounds: arithmetic mean, harmonic mean, geometric mean, etc. Homogenization theory proves that all these approaches are not universal and the dispersion of the results of their application is usually enormous.

Because of this unsatisfactory situation we have produced a database from measurements of horizontal and vertical thermal conductivities, total organic carbon and normative mineral composition of a set of about 350 sedimentary rock samples (TCD^TM; now a trademark of TGS Nopec). The thermal conductivities (in horizontal (K_h) and vertical (K_v) direction) have been measured on standard core plugs by a needle probe technique, the mineralogical composition has been determined by XRD and thin section petrography, and porosity measured by He-injection

From the database it is clear that there is a considerable variation of thermal conductivity even when lithology is known. We have shown that changes in the thermal conductivity are related to differences in porosity, texture and mineralogy (Fig. 2a). There can thus be a significant variation of the mineralogical content within one lithological unit. A core measurement is therefore not a good representation of an entire  geological unit. Estimating a mean value for a geological unit based on only few point measurements of thermal conductivity is subjected to large uncertainty.

Development of better tools is therefore needed. The problem can be viewed as a question of upscaling thermal conductivity from sample level to formation level. We propose to approach this by correlating thermal conductivity and well log data. Changes in the thermal conductivity are related to differences in porosity, texture and mineralogy. Common petrophysical logs such as spectral gamma, density, neutron, and sonic are in some way also affected by the same parameters.

 

Research approach/methods

Glacial Erosion within Plio-Pleistocene stages

Dividing the Plio-Pleistocene erosion/sedimentation into several stages will be done by modeling; the modeling technique described in more detail in Amantov et al. (2009). The margins of the glacial ice sheets are the starting point for the analysis. The ice margins in the three periods are shown in Figure 3, with present day topography. A number of tools are used to simulate erosion under the ice cover, and sedimentation under, at the margin and outside the ice. The tools are computation modules that allow geological analysis; computations that include sampling of gridded data (e.g. sub-ice lithology), connecting sparse kinds of data with a best-fitting surface, inferring velocity fields from the distance to an ice depocenter and topography, subtracting surfaces to determine the material removed, etc. 

Erosion under the ice sheets is estimated using these tools by requiring that the long-term glacial erosion rates are reasonable and the pattern of erosion conforms to the concentric (radial) changes in erosion as well as the 'spider's web' pattern of grounded ice sheets movement (ice streams). This is illustrated in Figure 4A-C, which shows the erosion and sedimentation that might occur if only the ice velocity were considered.  The concentric pattern results from the low ice velocity under the center of the continental glaciers and the more rapid basal ice velocity near the margins.  Figure 4B shows how this simple pattern is modified if the likely effect of the 'spider-web' pattern with the enhanced erosional capacity of ice streams is taken into account. Figure 4C illustrates the effect of different erodability of sedimentary bedrock and basement lithologies.

The above modeling technique was used to compute the Late Weichselian erosion and sedimentation (Amantov et al., 2009). This accumulated sediment mass must, of course, equal the mass of material eroded.  Our analysis assures that this is the case, not only today, but also for every increment of erosion that occurred over the entire glacial cycles. This technique will be used to divide the erosion/sedimentation into 3 Plio-Pleistocene stages, based on the three styles of glaciation shown in Figure 3.

Stress effects on faults due to erosion

Changes in local stresses and associated fluid pressures in petroleum reservoirs generated by rapid glacial erosion are topics of great practical importance, since sudden stress changes may reactivate or initiate faults and other fractures, allowing oil and gas to escape from reservoirs.

The result of a numerical model of a fluid reservoir subject to a fluid overpressure of 5 MPa with normal faults close to its lateral ends (Fig. 5), shows that the shear stress near to the ends of the faults is more than 40 MPa, which by far exceeds the shear strength of the rock (which is normally within the range of 1-12 MPa). This indicates that the normal faults are likely to be reactivated as reverse faults. The shear stress also concentrates at the lateral ends of the reservoir, indicating that the reservoir is likely to expand laterally. The expansion of the reservoir will lead to an increase in volume, which in turn leads to lowering of the fluid pressure giving rise to a possible underpressure.

Figure 5. Numerical model showing the resulting von Mises shear stress around a reservoir subject to a fluid underpressure of 5 MPa.

Løtveit et al. (2009) have used simple models for calculating the stress effects of erosion on the reservoirs at Snøhvit. Analytical models show that a rapid removal of 900 m sediments results in horizontal compressive stresses of magnitude 11.5 MPa. Analytical tunnel-crack and elliptical plate-bending models show that horizontal compression may have influenced the reservoir fluid pressure by first lead to overpressure and expansion, then to fluid underpressure as a consequence of the reservoir expansion and volume increase. The result of a similar model where the reservoir is subject to underpressure shows a slight subsidence of the block above the reservoir, and high shear stresses near to the tip of the faults, indicating a likely reactivation of the reverse faults.

In this project these models will be further developed by including more geophysical and geological data, so that the models become more realistic. We will use a more realistic approach regarding the reservoir geometries, locations and geometries of the faults, and a better representation of the sedimentary rock layers in the area surrounding the reservoirs. Different scenarios for various fault properties will be studied in for the estimation of the permeability changes induced by the erosion-related stress changes. These results will be of great importance to the discussion on the effects of the glacial erosion, and the possible leakage of gas and oil in the Barents Sea. The developed models will be implemented in basin modelling software (BMT).

Thermal conductivity of sedimentary rocks

The project will be split into 2 tasks. The first task describes development of methods to correlate well logs with thermal conductivity of the rocks along the well bore. The second task's focus is upscaling from sample level to formation level taking into account information obtained from the logs.

In order to obtain the input data for the upscaling of thermal conductivity it is necessary establish a relation between well log data and thermal conductivity. Popov et al. (2003a) have been able to do a quite good correlation between predicted thermal conductivities from petrophysical logs and measured values (Figure 2b). This methodology will be further developed, by use of multivariate data analysis. We will base this on a data set consisting of core plug samples. The plugs will be analysed for mineralogy (X-ray), porosity, TOC, textural properties (thin section, digital image analysis), thermal conductivity (K_h and K_v). In addition, corresponding data from several relevant types of petrophysical logs will be used.

For measuring the thermal conductivity we will use a non-contact and non-destructive optical scanning instrument that is developed (at Moscow State Geological Prospecting University) to provide a large number of high-precision measurements of thermal conductivity tensor components in samples of sedimentary and impact rocks. Studies show a reasonable correlation between (1) thermal conductivity and electric resistivity measured on core samples from the Em-Egovskoe oil-gas field, (2) thermal conductivity and sonic and electric logging data for the Ries impact structure (Popov et al., 2003b), and (3) parameter k_sat/k_dry and electric resistivity. Some of this work suffers from bad well log quality; it is, however, a good start of our research.

We will approach the development of thermal conductivity correlations in two ways. (1) interpretation of the well log data to derive a detailed description of the small scale variation of the thermal conductivity along the well bore, and (2) upscaling the detailed thermal conductivity tensor to the formation scale. The first step will be performed using the correlations developed in the previous task, see section 0between conductivity and well log signals. The computed conductivity will be a diagonal 2D tensor representing effective conductivies in the directions normal to the well and along the well.

 

REFERENCES

Amantov, A., Fjeldskaar, W. & L. Cathles, 2009. Glacial erosion in the Baltic Sea region: Effect on the post-glacial uplift. Accepted for publication in "The Baltic Sea Basin", Harff, Björck and Hoth (eds); Springer.

Cathles, L.M., 1975. The Viscosity of the Earth's Mantle, 386 pp., Princeton Univ. Press, Princeton, N. J.

Ekman, M., 1991. A concise history of postglacial land uplift research (from its beginning to 1950), Terra Nova 3, 358-365.

Fjeldskaar, W., C. Lindholm, J. F. Dehls & I. Fjeldskaar, 2000. Post-glacial uplift, neotectonics and seismicity in Fennoscandia, Quaternary Science Reviews, Vol. 19, 1413-1422.

Fjeldskaar, W., Christie, O. H. J., Midttømme, K., Virnovsky, G., Jensen, N. B., Lohne, A.,  Eide, G. I. and Balling, N., 2009. On the determination of thermal conductivity of sedimentary rocks and the significance for basin temperature history. Petroleum Geoscience, Vol. 15, 367-380. DOI 10.1144/1354-079309-814.

Knies, J., J. Matthiessen, C. Vogt, J. S. Laberg, B. O. Hjelstuen, M. Smelror, E. Larsen, K. Andreassen, T. Eidvin & T. O. Vorren (2009). The Plio-Pleistocene glaciation of the Barents Sea-Svalbard region: a new modelbased on revised chronostratigraphy. Quaternary Science Reviews 28, 812-829

Løtveit, I.F., Gudmundsson, A., Leknes, L., Riis, F., & W. Fjeldskaar, 2009. Effects of glacial erosion on the state of stress and fluid pressure in petroleum reservoirs in the Barents Sea. Accepted for publication in Journal of Geological Society of London, special issue, Pascal (ed.).

Mörner, N.-A., 1979. The Fennoscandian uplift and Late Cenozoic geodynamics: geological evidence, Geo Journal 33, 287-318.

Popov, Yu., V.Tertychnyi, R.Romushkevich, D.Korobkov & J.Pohl (2003a). Interrelations Between Thermal Conductivity and Other Physical Properties of Rocks: Experimental Data. Pure and Applied Geophysics 160, 1137-1161.

Popov, Yu., J.Pohl, R.Romushkevich, V.Tertychnyi & H.Soffel (2003b). Geothermal characteristics of the Ries impact structure. Geophysical Journal International 154 (2), 355-378.

Rasmussen, E. & W. Fjeldskaar, 1996. Quantification of the Pliocene-Pleistocene erosion of the Barents Sea from present-day bathymetry, Global and Planetary Change, v. 12, p. 119-133.

Riis, F. & W. Fjeldskaar, 1992: On the magnitude of the Late Tertiary and Quaternary erosion and its significance for the uplift of Scandinavia and the Barents Sea. In: Larsen, Brekke, Larsen, and Talleraas (eds): Structural and tectonic modelling and its application to petroleum geology, Norwegian Petroleum Society, Elsevier, pp. 163-185.

Walderhaug, O., Lander, R.H., Bjørkum, P.A., Oelkers, E.H., Bjørlykke, K., & P.H. Nadeau, 2000. Modeling quartz cementation and porosity in reservoir sandstones: examples from the Norwegan continental shelf. Spec. Publ. Int. Ass. Sediment., 29, 39-49.