INTRODUCTION AND BACKGROUND
Thermal conductivity measurements have been performed for many years. Clark (1966) and Robertson (1988) give summaries of results obtained for geological materials. Paitnik and Singh (1978) and Ashworth et al (1985) give overviews of the various techniques that have been used. Even though many methods have been developed and much data exists, there is need for additional work. Each method has specific attributes.
A good design must consider a number of factors; magnitude of the thermal conductivity being measured, appropriate temperature gradient within or applied pressure to the specimen, moisture in the specimen, ease of specimen fabrication, homogeneity/inhomogeneity and isotropy/ anisotropy of the rock, and a system symmetry which allows the development of known isotherms. Steady-state techniques give thermal resistance and hence conductivity directly. Transient techniques usually measure thermal diffusivity and the thermal conductivity calculated. Unfortunately, geological materials present many problems. Their thermal conductivity is low, requiring sensitive measuring systems or high heat input giving a high temperature gradient. The latter can cause relative expansion of parts of the specimen modifying thermal contact unpredictably. Rocks are difficult to machine this makes accurate location of temperature sensors in small grooves difficult.
There is large variability from one location to another even within the same rock type, requiring many specimens to be measured to best quantify thermal properties for a specific site. [Site kinowledge is highly recommended by Touloukian (1981)]. Anisotropy can be significant requiring the measurement of conductivity in a defined direction. The design of the laboratory system should not ignore the expected in situ conditions such as cracks, joints, moisture, pressure.