| Abstract |
Accurate modeling and forecasting of potential geothermal resources requires accurate measurements or predictions of the physical properties of the crust. The conductive thermal transport properties of rocks, thermal conductivity (k) and thermal diffusivity (D), have been studied extensively for over a hundred years, but it has been shown that results obtained by commonly used techniques contain systematic errors associated with imperfect contacts and ballistic radiative transfer (heat transport by radiation that does not interact with or warm the sample, and is directly generated by the heating element). Using Laser Flash Analysis (LFA), which limits these systematic errors, we have measured the temperature-dependent thermal diffusivity of major rock types across relevant lithospheric temperatures. These experiments have shown that D is much higher at low temperatures, and much lower at high temperatures than previous methods suggested. The temperature-dependence of D is therefore very strong (typically a factor of 2 to 5 over the range 20 to 500˚C), and D of common crustal minerals and rocks can vary by a factor of 4 or more at ambient temperatures. Thermal diffusivity of common crustal rock types can vary by a factor of 4 or more at ambient temperatures, and even by a factor of 2 for a given rock type. Variability between rocks of the same category may result from differences in modal abundances of rock-forming minerals, or contrasts in textures such as pore fraction. The combination of mineralogical, textural and temperature controls make predicting rock D by broadly-defined rock types (e.g. granite, basalt, marble) a potentially large source of error. For k, variability is even more complex because it is the product of D, heat capacity, and density, all of which are controlled by several properties. Accurate prediction of thermal transport properties requires petrographic (mineralogical and textural) investigation, and the effects of each property. The strong temperature-dependence of thermal transport properties, especially between ambient and ~500˚C, introduces the potential for feedbacks between heating and thermal insulation in the crust, an effect which needs to be incorporated in models when prospecting geothermal resources. |