## Modelling and Interpretation of Marine Controlled Source Electromagnetic Data

##### Doctoral thesis

##### Permanent lenke

http://hdl.handle.net/11250/239339##### Utgivelsesdato

2009##### Metadata

Vis full innførsel##### Samlinger

##### Sammendrag

This thesis aims to (i) interpret the marine CSEM data acquired over offshore Angola, Modgunn arch in the Norwegian Sea and Troll West Gas Providence (TWGP), offshore Norway, (ii) evaluate the effects of acquisition parameters and multi-resistor structures on CSEM responses, and (iii) investigate the time-lapse CSEM sensitivity with respect to changes in saturation and geometry of resistive pore-fluids. Three-dimensional (3D) finite difference time domain (FDTD) forward modelling is used to find correspondence between subsurface resistivity structures and measured data, and also to evaluate CSEM sensitivity and timelapse anomalies.
The electric field magnitudes measured above the proven HC reservoirs offshore Angola are 1.5–3 times higher compared to the synthetic background responses and “off-reservoir” measurements, while the responses above a salt structure is >3 times stronger compared to the background responses. The CSEM anomaly observed at Modgunn arch, corresponding to high-resistivity sills 1100 m below the seabed, is 2.5 times stronger compared to the background responses. The CSEM data measured above the TWGP (2.7 times stronger compared to “offreservoir” measurements), also indicate subsurface HCs. CSEM responses obtained from forward modelling, based on seismic and petrophysical data, show good agreement to the measured responses. Combined use of seismic, petrophysical and CSEM data improves the subsurface interpretation. Additionally, introduction of non-HC related high-resistivity structures, such as salt bodies, igneous sills and shallow resistors within the geo-resistivity model improves the subsurface interpretation.
Fine grid (1x1 km) acquisition geometry, irrespective of source orientation, provides precise definition of CSEM anomalies. Broadside data in grid geometry increase the data density (up to the factor of 3) and hence improve the CSEM attribute resolution. Source orientation oblique to the strike of an elongated resistor provides better structural definition for a coarse-grid geometry. A multiresistor anomaly is not simply the summation but a cumulative response with mutual interference between the constituent resistors. A gradual inverse variation of offset and frequency allows differentiation of CSEM anomalies for multilayered resistors. Similar frequency-offset variations for laterally persistent highresistivity facies show visual continuity with varying geometric expressions.
3D grid-modelling facilitates mapping of time-lapse CSEM anomalies. The time-lapse CSEM anomaly is not a simple function of the volumetric resistance, but the combined effect of variables associated with volumetric resistance including diameter, thickness, aspect ratio and resistivity. A 100% lateral expansion of a 1500 m wide CO2 plume results in 41% time-lapse anomaly, while a 300% vertical expansion of a 50 m thick plume gives only 7% anomaly. A typical HC reservoir (80 Wm) at 1000 m below seafloor gives 40% time-lapse anomaly when 5% of its reserve is replaced by brine. The un-drained compartments in a HC reservoir provide CSEM anomalies comparable to the baseline responses and hence the un-drained compartment may be detected by CSEM repeat surveying, given that the compartment diameter is comparable to the skin depth for the EM signal used. CSEM attribute analysis offers an early warning of CO2 migration in shallow traps.