Evaluation of Variation in Permeability Measurements when Using the MARK VI Standpipe
Using the MARK VI standpipe in a substrate-filled flume under constant flow conditions, we found significant differences in permeability readings made by different people at four of five sites. Coefficients of variation at each site ranged from 27 to 79%. Readings usually varied greatly for each person at each site. To detect a 10% change in the mean permeability at single sites (α = 0.05) would require from 34 to 90 samples; for a 30% change, the range was 4 – 10. Repeated sampling at a site did not produce directional changes in permeability estimates, nor did the performance of individuals change during the test. We propose a sampling strategy based on repeated readings taken by one person.