Thermal conductivity is a major influencing factor on subsurface conductive heat transport and resulting temperature distribution, which in turn is a key parameter in basin modeling. Basin modeling studies commonly use representative literature values of thermal conductivity despite their impact on modeling results. We introduce a workflow for quantifying the effect of uncertain thermal conductivity on subsurface temperature distribution and thus on basin modeling results and test this workflow on a two-dimensional generic model from the Nordkapp Basin; a prior ensemble of possible models is conditioned according to Bayes’ theorem to incorporate prior knowledge of temperature data. This conditional probability yields a posterior ensemble of temperature fields with a significantly reduced standard deviation. To verify our approach, we use five characteristic scenarios from the posterior ensemble for transient petroleum systems modeling. How considering uncertain thermal conductivity affects variance in hydrocarbon generation is assessed by modeling corresponding vitrinite reflectances (R_o).

Temperature uncertainty increases with depth. It also increases with increasing offset from the salt diapirs, which can be associated with a large lateral heat-flow component in the complex tectonic environment of the Nordkapp Basin. The introduced workflow can reduce temperature uncertainty significantly, especially in regions with high prior uncertainty. The R_o is very sensitive to changes in thermal conductivity because the onset depth of the gas window in the Nordkapp Basin may vary by up to 800 m (2600 ft) within the 95% confidence interval. This demonstrates the importance of quantification of the uncertainty in thermal conductivity on thermal basin modeling.