Published online by Cambridge University Press: 03 September 2012
A new method for estimating the mass transport by using the stochasticvalues (the arithmetic mean, the standard deviation and the skewness) ofpermeability is presented. Generally, detail of permeability distributioncannot be obtained except for moments of the distribution. Also, measurementresults of permeability for the rock matrix including cracks or fastflowpaths do not always follow the log-normal distribution frequentlyapplied. In such a situation, we must evaluate the characteristicpermeabilities for the whole or some regions of the disposal site includingthe accessible environment.
The authors have investigated the characteristic permeability on the basisof some probability density functions of permeability, applying the MonteCarlo method and FEM. It was found that its value does not depend on type ofprobability density function of permeability, but on the arithmetic mean,the standard deviation and the skewness of permeability [1].
This paper describes the use of the stochastic values of permeability forestimating the rate of radioactivity release to the accessible environment,applying the advection-dispersion model to two-dimensional, heterogeneousmedia. When a discrete probability density function (referred to as ‘theBernoulli trials’) and the lognormal distribution have common values for thearithmetic mean, the standard deviation and the skewness of permeability,the calculated transport rates (described as the pseudo impulse responses)show good agreements for Peclet number around 10 and thedimensionless standard deviation around 1. Further, it isfound that the transport rates apparently depends not only on the arithmeticmean and the standard deviation, but also on the skewness of permeability.When the value of skewness dose not follow the lognormal distribution whichhas only two independent parameters (the mean and the standard deviation),we can replicate the three moments estimated from an observed distributionof permeability, by using the Bernoulli trials having three independentparameters.