Determination of Hydraulic Properties With Point Source Experiments

Abstract

A finite-element model was developed to simulate the water flow underneath a surface point source using the Galerkin method of weighted residuals and the mixed formulation of the Richards' equation. The saturated radius usually formed during a point source test was estimated by the numerical model. Comparisons with published experimental and analytical results indicated that the model is accurate and reliable. The numerical model aids in understanding the water regime under different point source scenarios. The finite element model was coded in C++ by using an object oriented approach. This allows flexibility for future updating. Finite and infinite elements were combined to obtain a numerical solution for unbounded 1-D flow domains. The results showed that the use of infinite elements may reduce the computational time when solving the flow equations for deep profiles. Although the transient point-source problem can be solved numerically, generalized solutions were developed and their coefficients calculated by best fitting numerical results of dimensionless saturated radius and time. The solutions allow calculation of the transient saturated radius given values of the van Genuchten hydraulic function parameter "m" and dimensionless point-source application rate. The resulting generalized solutions have the advantage of reducing large volumes of data. The estimation of soil hydraulic parameters from field data by using existing flow theory is analyzed. The results obtained from these studies verified that soil type and conditions can limit field estimation of steady state data for some soils. As a consequence, estimation of steady data is highly recommended by fitting field data to empirical equations. An empirical equation for transient saturated radius as a function of time was proposed. This equation is robust to predict steady-state saturated radius. To test the applicability of point source tests, hydraulic properties were estimated from field disc and point source tests. The parameters obtained from the disc and point-source methods matched closely. An overall methodology is also given for conducting point-source tests.