High-resistivity anomalies at Modgunn arch in the Norwegian Sea

Transcription

1 first break volume 24, January 2006 High-resistivity anomalies at Modgunn arch in the Norwegian Sea Anwar Bhuiyan, 1 Tor Wicklund, 2 and Ståle Johansen 3 Introduction Typical seismic data provide information about subsurface stratigraphy and structure. Formation characteristics, such as lithology and fluid content, can also be predicted from seismic data. Well-log data can verify seismically extracted formation characteristics. However, well drilling is relatively expensive and the success rate of commercially viable exploration wells, depending on the seismic data, is only about 10 30% (Johansen et al., 2005). Additional remote sensing methods for the detection of subsurface formation properties (e.g. resistivity) can be used to minimize the uncertainties associated with drilling. The recently developed SeaBed Logging (SLB) method shows a very promising potential for the detection of deeply buried highresistivity layers (Eidesmo et al., 2002). Resistivity contrasts in the subsurface strata make SBL a potential tool for the detection of high-resistivity hydrocarbon reservoirs or other high-resistivity lithologies, such as salt domes, volcanic rocks or igneous sills. The first fullscale SBL calibration survey was conducted offshore Angola in 2000 (Ellingsrud et al., 2002), opening a new frontier in hydrocarbon exploration. Subsequently, several surveys were performed over known hydrocarbon fields offshore Norway. SBL calibration surveys from Ormen Lange and Troll Western gas province have been presented by Røsten et al. (2003) and Johansen et al. (2005), respectively. In this article, we present SBL data acquired across the Modgunn arch, which is located in the Norwegian Sea. The SBL data interpretation aims at finding the resistivity distribution within the seismically interpreted subsurface strata. The Modgunn arch is characterized by strong seismic anomalies, which may partially correspond to high-resistivity anomalies. The SBL data of this area, in parts, show strata with high resistivity. SBL data analysis can predict the presence of the high-resistivity layers and rocks, but due to low resolution, it is difficult to determine the exact geometry of the resistivity structure from the SBL data alone. To establish the quantitative relationship between the seismic anomalies and the resistivity distribution within the strata, SBL and seismic data interpretation play complementary roles. The integrated approach of seismic and SBL data interpretation provides a realistic subsurface resistivity distribution with fewer uncertainties. An interpretation study, based on electric field magnitudes taken from the same data set, has been presented by Bhuiyan et al. (2005). The SBL method The SBL method has been described in detail by Eidesmo et al. (2002) and Kong et al. (2002). It uses a horizontal electric dipole (HED) source, which generates ultra low-frequency (~ Hz) but powerful electromagnetic (EM) signals while being towed approximately m above the seabed. Energy propagates through the seawater, the seawater/air interface (air-waves) and the subsurface layers (Fig. 1). These signals, after propagating through different media, are measured by EM receivers located at the seafloor. According to Ward & Hohmann (1988), the propagation (α) and attenuation (β) constants in a conductive medium for frequencies below 10 5 Hz are defined as, where ω, μ and σ represent angular frequency, magnetic permeability and conductivity, respectively. For non-magnetic rocks, which we mostly encounter in sedimentary basins, μ μ 0 (the magnetic permeability in free-space). Therefore, in the case of fixed geometry, EM energy attenuation depends only on frequency, conductivity and source-receiver distance. For an ultra low-frequency system (~ Hz) as in SBL, the transmitting energy rapidly attenuates in seawater and seafloor sediments saturated with conductive saline water. Therefore, the direct energy (primary field) transmitted through seawater dominates the recordings only at short source receiver offsets. The air-wave (downgoing field) dominance depends on the source frequency, seawater depth, seawater and subsurface resistivity distribution, and source receiver distances. In the case of greater water depths (>1000 m), the air-wave usually starts to dominate at far offsets (e.g. >6 8 km). In a relatively thin (~10 50 m) 1 Norwegian University of Science and Technology (NTNU), Department of Petroleum Engineering and Applied Geophysics, Electromagnetic Geoservices, Geology and Geophysics, S. P. Andersens vei 15a, Trondheim 7491, Norway bhuiyan@stud.ntnu.no 2 Norwegian University of Science and Technology (NTNU), Department of Electromagnetic Geoservices and Geology and Geophysics, Norway 3 Norwegian University of Science and Technology (NTNU) Department Electromagnetic Geoservices, Norway 2006 EAGE 39

2 and high-resistivity subsurface layer (~ Ωm), EM energy propagates at a higher velocity as guided waves with less attenuation, and it is transmitted back (upgoing field) to the receivers at the seafloor. The upgoing field from a highresistivity subsurface layer will dominate over directly transmitted energy, i.e the primary field, when the source receiver offset is comparable to or greater than approximately twice the depth to this layer from the seafloor. The detection of the response from a relatively thin and high-resistivity subsurface layer is the basis of SBL (Ellingsrud al., 2002). The Modgunn arch The Modgunn arch is located in Blocks 6403/5 and 6403/6 in the Norwegian Sea; these blocks are operated by Statoil. The water depth in this area ranges from approximately Figure 1 Sketch of experimental geometry (top panel) and modelled SBL responses. Typical resistivities of air, seawater, sediments, and resistor are shown for reference. Black arrows denote downgoing fields (air-waves); primary fields (direct energy pathways) are indicated with blue arrows while the upgoing fields from a buried high-resistivity layer are shown with red arrows. EM responses are presented as magnitude vs. offset (MVO) and normalized MVO (middle panel) and phase vs. offset (PVO) and normalized PVO (bottom panel). Red and blue curves represent a model with and without resistor, respectively. 40 first break volume 24, January to 2200 m (Fig. 2). The Modgunn arch is a dome feature that developed due to tectonic inversion during the Oligocene time, like other major inversion structures in the Mid-Norway continental margin (Dore & Lundin, 1996). The structure is a three-way closure (defined from the seismic section) with spill point to the north-east. The overall lithology of this area is shale-dominated with occasional sand/shale alternations and few sandstone interbeds (Dalland et al., 1988; Swiecicki et al., 1998). The south-western part of the dome is characterized by discontinuous reflectors showing very strong anomalies compared to the regional reflectors, which can be interpreted as igneous intrusions within the strata of Lower Palaeocene to Upper Cretaceous age (Swiecicki et al., 1998). Contrasting anomalies mostly have crude parallelism, although some of them intersect the weaker hosts. A few strong reflectors cut across the horizontal to subhorizontal strata at high angles (~70 ), indicating that they are dikes. Hydrothermal veins are evidenced by the presence of hot spots, upward flows, and mismatches among the reflectors. Upper Tertiary sediments are characterized by continuous and parallel reflectors Figure 2 Upper panel: the SBL survey layout and the subsurface geological interpretation from 2D seismic sections are shown on a seabed topographic map. Lower panel: the seismic section (Line MB13-93) across the Modgunn area, corresponding to SBL towline (line-2) showing the receiver positions (yellow triangles). Geological time intervals used were taken from Swiecicki et al. (1998). The reference receiver used for normalization is indicated by a black arrow EAGE

3 first break volume 24, January 2006 occasionally disrupted by small-scale faults. The seabed and the shallow reflectors in the western part of the area are affected by the Storegga landslide and are characterized by steep escarpments and slumping (Bugge et al., 1988; Bunz et al., 2003). Data acquisition A speculative multi-client SBL survey was acquired by Electromagnetic Geoservices in 2003 with 31 EM receivers deployed from the vessel along seismic lines B3-85 (line-1 with 16 receivers) and MB13-91(line-2 with 15 receivers) over the Modgunn arch (Fig. 2). SBL line-2, on which this paper is focuses, was towed from SW to NE. An EM receiver had two sets of electric and one set of magnetic data-recording channels (sensors), each of which had orthogonally orientated x- and y-components. SBL data were recorded as time series (counts) throughout the survey. The HED antenna consisted of two electrodes (lead and tail), separated by approximately 220 m, with electrical contact to the seawater. The peak-to-peak current varied from zero to almost 1000 A. The HED antenna was towed from west to east over 16 receivers along line-1 and from SW to NE over 15 receivers along line-2, transmitting a continuous square-shaped signal at a base frequency of 0.25 Hz. The average towing heights were 40 and 30 m for lines 1 and 2, respectively. One possible reason for the large (10 m) sourceheight variation between lines 1 and 2 could be due to seafloor roughness (Fig. 2). Detailed descriptions of the SBL instruments and their operation can be found in Webb et al. (1985) and Sinha et al. (1990). Processing Prior to visualization and interpretation, the SBL data were processed through various steps, started from integration of the navigation and recorded data. A data length, representing the time that the HED source remained active, was extracted from all the recordings for further processing. This is called global data windowing, and it corresponds to the length of the towline. The windowed data were then transformed from the time domain to the frequency domain by performing short-window Fourier analysis. The offset values are spaced at regular intervals of 100 m after the Fourier transform. The frequency-filtered data were calibrated to relate the recorded signals (counts) to the physical field responses, i.e. the recorded data were converted to the corresponding electromagnetic field quantities. Source-signature variations influence the SBL responses, which can be quantified by current-moment (source-current amplitudes times the antenna length) normalization. Currentmoment-normalized data facilitate the comparison of SBL responses under changing source-signature conditions. This normalization was achieved by dividing the frequency-filtered complex receiver responses by the complex source-current moments for the same frequency. Proper channel dropping or summation of redundant electric channel responses is used to increase the signal-tonoise ratio for electric field components. If two channels have different noise levels, the noisy channel is dropped, and in the case of the same noise level, the two channels are summed. SBL receivers are allowed to sink freely down to the seabed while being deployed from the survey vessel. The x- and y-channel orientation could therefore be arbitrary. It is, however, necessary to rotate the data so that the x- and y- channels are respectively aligned with and perpendicular to the nominal towline direction. The in-line component (x) of the electric field (E) and the cross-line component (y) of the magnetic field (H) are found to be strong for an in-line orientated HED, whereas the y- and x-components of the E and H wavefields are weak and noisy for this case and should therefore be filtered out. Consequently, we focus here on the Ex and Hy components. The processed data quality was observed to be sufficiently good to be interpreted at source receiver distances of up to ~7 km. The data were displayed as the magnitude and the phase of the electric and magnetic fields versus offset. Modgunn SBL data Magnitude vs. offset (MVO) and phase vs. offset (PVO) responses for all receivers were normalized by a reference receiver response to highlight the changes from one measurement to the next. The MVO normalization is done by dividing the MVO responses of target receivers by the response of a reference receiver, while the PVO normalization is done by subtracting the PVO responses of target receivers from that of the reference receiver (Fig. 3). In the MVO normalization, the fluctuations in the response curve become very high due to the noise level variations at far offsets (>7 km). To avoid this variation in the response curve, a least-squares polynomial curve-fit (green curves) for the reference receiver (Fig. 3) is used for the MVO normalization. Normalized out-towing (source towed away from the receiver) electric and magnetic MVO and PVO responses of one representative receiver (M02_Rx01_out), located above the area of strong reflectors (SW part), are shown in Fig. 3. The reference receiver (M01_Rx01_out) from line-1 is located in the SE part of the study area, outside the strong reflectors (Fig. 2). Accumulated curves for all the receivers have been constructed by plotting normalized median-filtered values (each median-filtered value is calculated from every 20 consecutive measurements) as a function of range (receiver position ±3.0 km along the survey line, see Fig. 4). In order to display SBL variations throughout the entire survey (line-2), we have taken the normalized median-filtered MVO and PVO (both in-towing and out-towing) values at 6.0±0.5 km offset for all receivers and posted them at the source receiver midpoints (i.e. ±3.0 km). SBL data points for the summed plot are usually chosen from an intermediate interval of the response curve to avoid near-offset (~2000 m) primary field 2006 EAGE 41

4 first break volume 24, January 2006 effects and far-offset (>7000 m, in this case) air-wave effects and noisy data. Forward modelling simulations help to find the appropriate offset intervals. Normalized median-filtered MVO and PVO systematic results for electric and magnetic responses follow similar trends along the towline across the arch (Fig. 4). The normalized electric MVO responses are around 1.0 at the north-east part of the Modgunn arch (outside the region of intrusives), and increase gradually towards 3.0 over the apex of the arch where the strong reflectors are present. The normalized magnetic MVO responses vary from 1.0 to 2.3, following the similar trend of the normalized electric MVO responses. The normalized electric PVO values vary from 0 to 1.4 radians and normalized magnetic PVO variations range from 0 to 1.0 radians (from NE to SW). Modelling SBL forward-modelling simulations have been used to quantify the measured SBL responses in the context of resistivity distribution. In this study, we used 2.5D finite-difference timedomain forward-modelling software (Wang & Hohmann, 1993; Chen et al., 1997). Important factors to consider in SBL forward-modelling simulations are water depth, seafloor topography, burial depth of the significant layers in the overburden, electrical properties of the overburden, geometric and electrical properties of the targeted resistor, and electrical properties beneath the resistor. 2.5D modelling demonstrates how the EM response is affected by the geometry and resistivity distributions that vary in two dimensions (depth and in-line horizontal coordinates) within the subsurface strata, and the effect of varying receiver-layout geometries. The reciprocity principle was used to compare the real SBL experiment with the 2.5D modelling with a fixed source position at the seafloor and a regularly spaced (100 m) receiver array, 30 m above the seafloor. The structure-geometry models used in the modelling simulations were constructed from the 2D seismic interpretation, shown in Figs 5 and 6. These models were found to be adequate to fulfil the Figure 3 Measured SBL data presentation procedure. Upper left-hand panel: the electric MVO; upper right-hand panel: the electric PVO responses. Lower left-hand panel: the magnetic MVO; lower right-hand panel: the magnetic PVO responses. The arrows indicate the median-normalized outtowing responses at 6.0±0.5 km source receiver offsets as the basis for construction of Fig. 4. Figure 4 Median-filtered normalized responses for out-towing (red) and in-towing (blue) responses at 6.0±0.5 km source receiver offsets for all receivers along line-2. Upper panel: magnetic responses (rhombus for normalized MVO and the triangles for normalized PVO); middle panel: electric responses (squares for normalized MVO and the circles for normalized PVO); lower panel: the simplified geological model used along the survey line. The receivers and the corresponding structure-geometry models (framed) used for 2.5D modelling simulations are shown in the lower panel. Median-filtered values of the normalized responses are posted at the source receiver midpoints (i.e. ±3.0 km) EAGE

5 first break volume 24, January 2006 Figure 5 2.5D forward-modelling simulation results for different resistivity combinations (top). Modelled and measured SBL receiver data across the Modgunn arch (bottom). The red curve represents a typical response from a representative receiver measurement above the sill (M02_Rx01_out) divided by a reference receiver (M01_Rx01_out) outside the sill. The blue curve is the 2.5D modelled on-sill response normalized by the 2.5D modelled off-sill response. requirements for 2.5D modelling simulations. In fact, the background geology (used in the modelling) is more or less the same in and around the study area (Dalland et al., 1988; Swiecicki et al., 1998). However, detailed mapping of sills and/or other high-resistivity structures has to be limited due to lack of 3D seismic data. Average resistivity values used in modelling simulations are based on available well information around the study area (Wells 6404/11-1, 6305/4-1, 6305/5-1, 6305/7-1, 6305/8-1, etc., in Havsule and Ormen Lange Structures of the Mid-Norway Continental Margin). The resistivity value used here for the sills ranges from 100 to 1000 Ωm, whereas the resistivity used for the overburden and the half-space ranges from 1 to 5 Ωm. The resistivity value used for a moderate resistor, introduced 1000 m below the seafloor (Fig. 6), ranges from 10 to 50 Ωm. In the simulations, the air-layer is not included in the modelling. The EM energy trapped at the seawater/air interface (air-wave) will return as a downgoing field and will influence the SBL responses only at far offsets since the water depth is comparatively large ( m). Forward-modelling simulation results indicate good to fairly good fits between modelled and measured data for almost all the receivers. Figure 6 2.5D forward-modelling simulation results for different resistivity combinations for a moderate resistor 1000 m below the seafloor and fixed resistivity (300 Ωm) for the sills (top). Modelled and measured SBL data for the receiver M02_Rx07_out are shown in the bottom panel. The red curve represents a typical response from a representative receiver measurement above the sill (and possible moderate resistor) divided by a reference receiver (M01_Rx01_out) outside the sill. The blue curve is the 2.5D modelled on-sill response normalized by the 2.5D modelled off-sill response. Here, we present the simulation results for two representative receivers as examples. Figure 5 (top) represents the modelling responses (normalized electric MVO) of a structure-geometry model, constructed as an out-towing location for receiver-1 in line-2 with different resistivity distributions. Modelled SBL responses are normalized to a reference case (receiver-1 out-towing location in line-1) with no assumed subsurface high-resistivity anomalies. Figure 5 clearly shows that the MVO increases with resistivity in the sills region, as expected. The modelled SBL responses are more than 3.5 times stronger than the response from an off-sill case. Figure 6 (top) represents normalized electric MVO modelling responses of a structure-geometry model, constructed as an out-towing location for receiver M02_Rx07, with different resistivity values of the moderate resistor located 1000 m below the seabed. The resistivity value for the sill was set at 300 Ωm. Figure 6 shows that MVO increases with the resistivity of the modelled resistor, and the modelled SBL responses are more than 100% stronger than that of the off-sill case after 7 km offset. Comparison of modelled and measured data The geological model and the corresponding measured (red circles) and modelled (blue line) data are shown in Figs 5 and EAGE 43

6 first break volume 24, January 2006 (bottom panel). The measured responses start to increase after 2 km offset and the normalized values reach 4.0 for receiver M02_Rx01_out and 2.0 for receiver M02_Rx07_out at 7.5 km offset. There is fairly good agreement between the 2.5D forward simulations and the measured SBL data (Fig. 5). The resistivity value used for the sills in the model is 300 Ωm. However, normalized values of the measured MVO at intermediate offsets (2 5 km) are slightly higher than the modelled ones. A better match was achieved by introducing the 50 m thick (thickness taken from the seismic section), moderately resistive (10 Ωm) layer, buried at 1000 m below the seafloor in a model for receiver M02_Rx07_out (Fig. 6). In this model, the resistivity for the strong reflectors representing sills is also set at 300 Ωm. This indicates that the measured SBL response cannot be fully explained by the simplified geological model with sills alone. Likely candidates for the 10 Ωm layer could be either a low saturation hydrocarbon layer or a very thin, highly fractured and discontinuous sill. However, this resistivity for the sill layer (10 Ωm) is probably unrealistically low. If the resistivity is kept at 300 Ωm, the sill must be very thin and possibly not visible on the seismic section. We would need 3D seismics to generate a more detailed 3D geological model. Such data are not available to us. Conclusions The SBL data across the Modgunn arch directly detect deeply buried high-resistivity sills. A powerful dipole source induced SBL responses in the SW part of the study area that were increased by up to 250%, compared to the SE part. The strong SBL responses are due to reflection and refraction of EM energy from sills (300 Ωm) situated within the depth range 1100 to 2500 m below the seabed. The SBL measurements show fairly good agreement with the 2.5D SBL forward-modelling simulations. An even better match was achieved after introducing a 10 Ωm layer within the overburden above the sills (1000 m below the seafloor). This suggests that the measured SBL response cannot be fully explained by a simplified geological model with sills alone. Acknowledgements We thank Electromagnetic Geoservices for giving permission to publish this data set and Pål Gabrielsen for valuable help during the project. References Bhuiyan, A.H., Wicklund, T.A., and Johansen, S.E. [2005] Geophysical characterization of sub-seafloor strata of Modgunn arch from SBL data. 67 th EAGE Conference, Madrid, Spain, Extended Abstracts, Bugge, T., Belderson, R.H., and Kenyon, N.H. [1988] The Storegga slide. Philosophical Transactions of the Royal Society of London 325, Bunz, S., Mienert, J., and Berndt, C. [2003] Geological controls on the Storegga gas-hydrate system of the mid-norwegian continental margin. Earth and Planetary Science Letters, 209, Chen, Y.H., Chew, W.C., and Oristaglio, M.L. [1997] Application of perfectly matched layers to the transient modeling of subsurface EM problems. Geophysics 62, Dalland, A., Worsley, D., and Ofstad, K. [1988] A lithostratigraphic scheme for the Mesozoic and Cenozoic succession offshore mid- and northern Norway. Norwegian Petroleum Directorate Bulletin, 4, Dore, A.G. and Lundin, E.R. [1996] Cenozoic compressional structures on the NE Atlantic margin: Nature, origin and potential significance for hydrocarbon exploration. Petroleum Geoscience, 2, Eidesmo, T., Ellingsrud, S., MacGregor, L.M., Constable, S., Sinha, M.C., Johansen, S.E., Kong, F.N., and Westerdahl, H. [2002] Sea Bed Logging (SBL), a new method for remote and direct identification of hydrocarbon filled layers in deepwater areas. First Break, 20(March), Ellingsrud, S., Sinha, M.C., Constable, S., MacGregor, L.M., Eidesmo, T., and Johansen, S.E. [2002] Remote sensing of hydrocarbon layers by Sea Bed Logging (SBL): results from a cruise offshore Angola. The Leading Edge, 21, Johansen, S.E., Amundsen, H.E.F., Røsten, T., Ellingsrud, S., Eidesmo, T., and Bhuiyan, A.H. [2005] Subsurface hydrocarbons detected by electromagnetic sounding. First Break, 23(March), Kong, F.N., Westerdahl, H., Ellingsrud, S., Eidesmo, T., and Johansen, S.E. [2002] SeaBed Logging: A possible direct hydrocarbon indicator for deep sea prospects using EM energy. Oil & Gas Journal, May 13. Røsten, T., Johansen, S.E., Ellingsrud, S., Amundsen, H.E.F., and Brevik, I. [2003] A Sea Bed Logging (SBL) calibration survey over the Ormen Lange gas field. 65 th EAGE Conference, Stavanger, Norway, Extended Abstracts, P058. Sinha, M.C., Patel, P.D., Unsworth, M.J., Owen, T.R.E., and MacCormack, M.G.R. [1990] An active source electromagnetic sounding system for marine use. Marine Geophysical Research, 12, Swiecicki, T., Gibbs, P.B., Farrow, G.E., and Coward, M.P. [1998] A tectonostratigraphic framework for the Mid-Norway region. Marine and Petroleum Geology, 15, Wang, T. and Hohmann, G.W. [1993] A finite difference time-domain solution for three dimensional electromagnetic modelling. Geophysics, 58, Ward, S.H. and Hohmann, G.W. [1988] Electromagnetic theory for geophysical applications. In: Electromagnetic Methods in Applied Geophysics (Ed. M.N. Nabighian), pp Investigations in Geophysics, No. 3 (Volume 1, Theory). Webb, S.C., Constable, S.C., Cox, C.S., and Deaton, T.K. [1985] A seafloor electric field instrument. Journal of Geomagnetism and Geoelectricity, 37, EAGE