Temperature and Pressure Dependence of Thermal Conductivity in Deep Marine Carbonates: Correction Models and...

The reconstruction of the ultradeep geothermal field is the foundation of deep oil and gas exploration, and the evolution of rock thermophysical properties under ultrahigh temperature and pressure environments directly affects the accuracy of geothermal prediction. Taking Well SDTK1 in the Tarim Basin as an example, this study employs a layered iteration algorithm based on geological parameters to conduct deep geothermal prediction and thermal conductivity model optimization under temperature-only, pressure-only, and coupled temperature-pressure conditions. Simulation results show that the dynamic correction effect of thermal conductivity exhibits significant depth-dependent differentiation. In Mesozoic-Cenozoic clastic rocks shallower than 6000 m, the prediction error generated by the constant thermal conductivity assumption is minor. However, upon entering the deep Paleozoic marine tight carbonate rocks, influenced by high temperatures, the uncorrected model leads to an underestimation of geothermal temperatures by up to 40-50 °C. Comparison among the correction models indicates that the compaction gain brought by formation pressure at a depth of 11000 m (a temperature decrease of 10.2 °C) is far insufficient to compensate for the thermal conduction attenuation caused by high temperatures (a temperature increase of 45.1 °C). The thermal resistance effect induced by high temperatures dominates the evolution of the deep geothermal field. Calibrated with measured data, the Chapman model, which possesses a larger temperature attenuation gradient, can better compensate for the thermal conduction attenuation of tight carbonate rocks at high temperatures, and its predicted geothermal temperature (206.7 °C) under the coupled correction model is closer to the actual value. This study demonstrates that in the evaluation of thermal history and hydrocarbon phase states in ultradeep basins, it is necessary to introduce a dynamic thermal conductivity correction model to effectively reduce prediction deviations of deep geothermal temperatures.