Validity range and reliability of the United States Bureau of Reclamation (USBR) method in hydrogeological investigations

Abstract

The United States Bureau of Reclamation (USBR) formula was derived from a table presenting values of hydraulic conductivity as a function of grain size, K = f(d20). The original table was empirically designed as a sequence of variation of different permeability coefficients of deposits and was intended for the design of earth dams, for the purpose of assessing leakage where percolation tests are not available. The USBR formula has since been used for predicting the hydraulic conductivity of water-bearing uniform sand deposits but systematically derives values of hydraulic conductivity several times lower than realistic values for materials. In this article, the optimal analytical formulation of the series of original data for K = f(d20) from Justin et al. (1945) is presented. Additionally, through calibration using results of hydrogeological research in Croatia, Germany, China and Nigeria, a formula (named USCRO) for predicting the permeability of sediments over a wide range of uniformity and d20 grain size was derived. The validity of this function for expressing permeability and the utilization of relative nondimensional coefficients is examined through a graphical correlation of the permeability of uniform and especially well-graded materials. Samples of poorly graded sand (63) and well-graded sandy gravel (131) were included in the calibration procedure. Data for mechanical analyses were taken from published articles. The numerical correlation of the USCRO formula for uniform sand samples resulted in a Pearson correlation coefficient of R2 = 0.902; for the well-graded sandy gravel, R2 = 0.838. Justin JD, Hinds J, Creager WP (1945) Engineering for Dams (Vol III), John Wiley & Sons.