Effects of turbulence on hydraulic heads and parameter sensitivities in preferential groundwater flow layers

Abstract

[1] A conduit flow process (CFP) for the Modular Finite Difference Ground-Water Flow model, MODFLOW-2005, has been created by the U.S. Geological Survey. An application of the CFP on a carbonate aquifer in southern Florida is described; this application examines (1) the potential for turbulent groundwater flow and (2) the effects of turbulent flow on hydraulic heads and parameter sensitivities. Turbulent flow components were spatially extensive in preferential groundwater flow layers, with horizontal hydraulic conductivities of about 5,000,000 m d À1 , mean void diameters equal to about 3.5 cm, groundwater temperature equal to about 25°C, and critical Reynolds numbers less than or equal to 400. Turbulence either increased or decreased simulated heads from their laminar elevations. Specifically, head differences from laminar elevations ranged from about À18 to +27 cm and were explained by the magnitude of net flow to the finite difference model cell. Turbulence also affected the sensitivities of model parameters. Specifically, the composite-scaled sensitivities of horizontal hydraulic conductivities decreased by as much as 70% when turbulence was essentially removed. These hydraulic head and sensitivity differences due to turbulent groundwater flow highlight potential errors in models based on the equivalent porous media assumption, which assumes laminar flow in uniformly distributed void spaces.

Geostatistical Borehole Image‐Based Mapping of Karst‐Carbonate Aquifer Pores

“…The hydrodynamics of this aquifer are likely to be non‐Darcian under head gradients as small as 10 −5 based on simulations of flow in vuggy megaporous rock (Sukop et al ) and, therefore, sometimes outside the realm of traditional groundwater flow models, which are based on the assumption of a constant hydraulic conductivity and linearity of Darcy's Law. Two alternatives to traditional models are models that account for changes in apparent hydraulic conductivity with increasing flow rates (e.g., Shoemaker et al and Reimann et al ) or direct modeling of the flow in the megapores. This paper assesses strategies for using optical borehole logs to develop input data for the direct modeling of flow.…”

Modifications to the Conduit Flow Process Mode 2 for MODFLOW‐2005

“…Frequently‐used numerical groundwater flow models such as MODFLOW‐2005 (Harbaugh 2005) consider laminar flow only. However, rapid turbulent flow can occur through highly conductive underground structures like preferential flow layers with vuggy porosity (Kuniansky et al 2008; Shoemaker et al 2008a, 2008b) and the impact of these turbulent flow processes can be considerable, for example when spring flow needs to be simulated reliably to monitor adherence to minimum flow goals by water resource managers (Jeannin 2001; Scanlon et al 2003).…”

“…A promising new approach is the CFPM2, a continuum model, which can be used to compute laminar or turbulent flow in large pore aquifers where the onset of turbulence is at low Reynolds numbers (1 to 100) (Kuniansky et al 2008; Shoemaker et al 2008a, 2008b;). Because the solvers for MODFLOW sometimes use only head closure and not flow closure, and the turbulent conductance is now a function of head, the solution may not converge on a single time step.…”

Lattice Boltzmann Models for Flow and Transport in Saturated Karst

“…The Colebrook-White equation is used for estimation of the friction factor. Shoemaker et al (2008) added a module called Conduit Flow Process Mode 2 to MODFLOW to simulate inertial flow in karst aquifers. Birk et al (2005) integrated the equilibrium and nonequilibrium (two-region) advection-dispersion equation (ADE) with CAVE and compared field spring discharge and tracer breakthrough data with model predictions.…”