Bench-scale experimental determination of the thermal diffusivity of crushed tuff

A bench-scale experiment was designed and constructed to determine the effective thermal diffusivity of crushed tuff. Crushed tuff particles ranging from 12.5 mm to 37.5 mm (0.5 in. to 1.5 in.) were used to fill a cylindrical volume of 1.58 m3 at an effective porosity of 0.48. Two iterations of the experiment were completed; the first spanning approximately 502 hours and the second 237 hours. Temperatures near the axial heater reached 700 degrees C, with a significant volume of the test bed exceeding 100 degrees C. Three post-test analysis techniques were used to estimate the thermal diffusivity of the crushed tuff. The first approach used nonlinear parameter estimation linked to a one dimensional radial conduction model to estimate thermal diffusivity from the first 6 hours of test data. The second method used the multiphase TOUGH2 code in conjunction with the first 20 hours of test data not only to estimate the crushed tuffs thermal diffusivity, but also to explore convective behavior within the test bed. Finally, the nonlinear conduction code COYOTE-II was used to determine thermal properties based on 111 hours of cool-down data. The post-test thermal diffusivity estimates of 5.0 x 10-7 m2/s to 6.6 x 10-7 m2/s were converted to effective thermal conductivities and compared to estimates obtained from published porosity-based relationships. No obvious match between the experimental data and published relationships was found to exist; however, additional data for other particle sizes and porosities are needed