Vertical and lateral contrast in thermal conductivity produce refracted isotherms and anomalies in heat flow. For simple steady-state heat conduction, three models of rift basins with various fault attitudes (vertical, high-angle, and listric) and a thermal conductivity ratio of 1:2, basin material:country rock, are analyzed numerically using the computer program CCC which utilizes the Integrated Finite Difference Method. The isotherms are refracted near the contrast in thermal conductivity and heat flow anomalies are present on both sides of the fault. An analytical solution for the temperature distribution across a vertical fault is found by representing the contact as two plates in perfect thermal contact and solving Laplace's equation in two dimensions. The normalized heat flow is calculated for the models and is also approximated by an empirically derived equation. The equation for normalized heat flow can be derived from the analytical solution to Laplace’s equation.

From the analytical solution for the temperature distribution, the maximum and minimum heat flow near a thermal conductivity contrast is determined from the average thermal gradient and the conductivities of the rocks. The analytical solution also yields an equation which when solved iteratively can be used to estimate the distance to the fault. The analytical and numerical results shows excellent agreement.

A linear equation which represents the horizontal normalized heat flow distribution is empirically derived. This equation is based on the distance from the fault, the minimum heat flow in the basin and the horizontal change in normalized heat flow. The minimum normalized heat flow for listric faulting is found to lower in value than for the vertical and high-angle fault models. If a heat flow determination is lower than the calculated minimum heat flow, the fault attitude is different from vertical. The linear equation for heat flow is simple enough for field use.

Thermal equilibrium data can supplement interpretation of structure as deduced from other geological and geophysical data sets.