Computing thermal conductivities of rocks from chips and conventional specimens

Suggestions are made for detecting minerals arranged in parallel in rock specimens of the size commonly used for thermal conductivity measurements. Methods are proposed for testing the validity of conductivities computed from the composition of a specimen for various simple models. For a system having minerals of only two conductivities, it is not necessary to know either conductivity; if there are minerals with three different conductivities, only one of the conductivities need be known. In both cases the unknown mineral conductivities are obtained from the tests, thus giving a quasi in situ conductivity value for these minerals. The tests have been applied to a set of 48 disks of a coarse-grained granite described in an earlier paper. It was found that the parallel model could definitely be discarded but that it was difficult to distinguish between the dispersive (Maxwell) model and the series model. The conductivity of the feldspar was significantly higher and the conductivity of the biotite significantly lower than was previously supposed. The application of these results to computing the conductivity of chips of rock from oil wells, or from friable cores, is discussed.