Numerical Determination of Rock-Mass Thermal Conductivity

Abstract

Temperature distributions from the Single Heater Test of the Yucca Mountain Project were used to determine rock-mass thermal conductivity. The Single Heater Test, located in a densely welded tuff in Alcove 5 of the Exploratory Studies Facility at Yucca Mountain, is nominally 13-m wide, 10-m deep and 5.5-m high. A centrally located, 5-m long, 4 kW electrical heater was activated for 9 months. During the heating phase and subsequent cooling phase of a similar duration, temperatures were measured hourly from more than 300 thermocouples emplaced in boreholes strategically drilled into the test block. An inverse method, that assumes a linearized system, was applied. This method minimized the sum of residuals between temperature measurements and simulations. The simulations were based on temporal and spatial superposition of a series of point sources that represented a linear heat source akin to the line-source heater in the Single Heater Test. Also the method accounted for fluctuations in the power of the central heater through the use of convolution methods. Subsequently, the derived value for rock-mass thermal conductivity was compared to values determined from several laboratory and field techniques that accounted for both matrix and lithophysal porosity. In general, agreement between the various methods was good.