Heat flow studies of the Exmouth Plateau, offshore North West Australia

Abstract

The surface heat flow pattern for the Exmouth Plateau region offshore North West Australia has been defined from 86 new, widely spaced heat flow measurements using conventional marine heat flow techniques. The average surface heat flow is 59 mW/m² , which is the expected level for continental crust. The heat flow frequency has a Gaussian (bell shaped) distribution about this mean within the range of 19 mW/m² to 106 mW/m². Overall the areal heat flow variation is rather smooth, and there is a gradual trend from high heat flow, in the order of 100 mW/m², from a confined region in the south east, to average heat flow values on the plateau. It is within this background heat flow on the plateau that a single region of low heat flow low is situated. The area of the heat flow low, which has levels below 40 mW/m² , is about 4000 km² and the centre of the low is about 230 km from the heat flow high. The confined heat flow high and the heat flow low are the only major heat flow anomalies (relative to the average heat flow) for the region. The nature of these anomalies, how and why they have come into existence and the general consequences to our understanding of heat transfer in sedimentary basins is the main stimulus for this thesis. As surface heat flow patterns are an expression of subsurface thermal processes; i.e. diffusion, convection and internal heat generation, the surface heat flow pattern can be interpreted in order to describe the dominant methods of heat transfer in the crust. The presence of the heat flow low provides the key to interpreting the regional pattern. As the heat flow level is less than base crustal level, it indicates that natural (thermally driven) pore fluid convection together with conduction are probably the main means of heat transfer in the continental crust within the region. It is reasonable to assume that the single heat flow low and high are linked in some way. So a single convection cell, based on the absence of other anomalies, is postulated and thought to be largely confined to the uppermost 10 kilometres of sediment where there is sufficient permeability. There are however two major problems that need to be addressed, firstly the separation of the surface heat flow low and high is 230 kilometres, consequently the cell has an extremely low aspect ratio of 0.04. Secondly the system appears to be operating in very low energy conditions and so what is the cause of the convection? Confirmation of the existence of a natural convection cell in the sedimentary section together with the problems of pore fluid convection in such a low energy environment and low aspect ratio cells are addressed by numerical modelling. The mathematics for two dimensional non-linear heat and mass transfer in heterogeneous and anisotropic porous medium that is applicable to the geological environment has been developed. The equations governing fluid motion are coupled to the diffusive and convective heat equation, so enabling an emphasis on free or natural convection rather than forced convection. A numerical model, based on Alternate Direction Implicit method, has been developed around the mathematical formulation to simulate heat and mass transfer in a porous media. The results of the modelling support the proposition that natural pore fluid convection is a dominant process for heat transfer in the Exmouth Plateau region. Some independent evidence exists that supports the proposition of natural pore fluid convection in the region, most notable is the temperature and vitrinite reflectance data from the many oil exploration wells drilled on and near Exmouth Plateau. A new thermal geohistory analysis, based on numerical methods has been developed to take full advantage of this data. The results from 13 offshore wells confirm the overall heat flow pattern and lend support to convective flow. However the major support for the postulated pore fluid convection pattern is derived largely from numerical modelling. The heat flow data and the results from the numerical modelling stimulated a study on the necessary and sufficient conditions for natural pore fluid convection to occur in sedimentary basins. There exist two essentially independent aspects important in the study: the first is the driving force, which largely depends on the lateral variation of rock thermal conductivity, and the second is the modifying force, which depends on the permeability distribution and the fluid viscosity. It is concluded that lateral variations in rock thermal conductivity give rise to horizontal temperature gradients that are the major driving force of natural convection. By the intrinsic nature of sedimentary basins, they contain rocks of low thermal conductivity and high permeability and are surrounded by or contained within rocks that have a relatively higher thermal conductivity and lower permeability, together with the temperature dependency of fluid viscosity, natural convection is probably an active process in most sedimentary environments. Discussion on the consequences and application of these findings is focused on hydrocarbon maturation, migration and trapping. Convection usually gives rise to areas of anomalously high or low temperature within the sedimentary section and this has direct consequences to hydrocarbon maturation. Furthermore as the pore fluids are moving, horizontal path lengths may be significant, so the study also has application to hydrocarbon migration. As the fluid is moving it will alter the hydrostatic pressure field to a hydrodynamic pressure field. Hydrodynamic hydrocarbon traps may be generated as a consequence of pore fluid flow.