Integrating Plug to WellScale Petrophysics with Detailed Sedimentology to Quantify Fracture, Vug, and Matrix Properties in Carbonate Reservoirs 
Kansas Geological Survey Openfile Report 9947 
To intetegrate PlugWhole CoreLogDST measures of permeability it is necessary to develop an upscaling model that is conceptually consistent with the nature of each measurement and the assumptions inherent in each. This requires revisiting basic assumptions.
Standard calculations of permeability from DST data assume:
Porosity Distribution
Porosity is normally distributed. Because porosity is a static, scalar property, an arithmetic average provides an accurate, scaleindependent, measure of porosity, assuming that the range of lithofacies were sampled appropriately 
Permeability Distribution
Permeability is lognormally distributed. Permeability is a vector and in heterogeneous media, permeability is scaledependent. An arithmetic average of logk provides an accurate, scaledependent measure of permeability, assuming that the range of lithofacies were sampled appropriately. As a vector, the data also have orientation. An average of the logk values provides a statistic of the values for the scale at which the values were measured which does not necessarily represent other scales. Application of this average to larger scales assumes a specific permeability architecture. 

email : webadmin@kgs.ku.edu
Last updated November 2002
http://www.kgs.ku.edu/PRS/Poster/1999/9947/P301.html