10.4122/1.1000000351
Cheng, Hwai-Ping
Hwai-Ping
Cheng
hwai-ping.cheng@erdc.usace.army.mil
England, Steven
Steven
England
steven.m.england@nap02.usace.army.mil
Brown, Christopher
Christopher
Brown
christopher.j.brown@saj02.usace.army.mil
England, Steven
Steven
England
steven.m.england@nap02.usace.army.mil
Significance of Mesh Density in Aquifer Storage and Recovery Modeling
XVI International Conference on Computational Methods in Water Resources
2006
2006
Aquifer Storage and Recovery (ASR) is means to store fresh water deep underground
in brackish water aquifers. This stored water can be recovered at a later date
during emergencies or times of water shortage. ASR is expected to provide a cost-
effective solution to many of the world’s water management needs. However, the
quality of the stored water may degrade over time due to mixing and buoyancy
stratification. Water quality may further be reduced during extraction due to
upconing of saline water underlying the ASR well. This water quality degradation
may reduce the volume of the available fresh water during recovery to the point
that the ASR well is no longer cost effective. Because flow and concentration
gradients may be high in the vicinity of an ASR well, the vertical and horizontal
resolution of the 3D mesh in the vicinity of the ASR well is important. Meshes
that do not have sufficient resolution in the area of interest may not accurately
simulate observed conditions in these high gradient areas. On the other hand,
meshes that contain too much resolution will require more computational resources,
resulting in extended simulation times. A preliminary study has shown that the ASR
simulation became more computationally stable as the vertical resolution was
increased in the geologic units above, below, and containing the ASR well when the
WASH123D model was used as the computational tool. In this paper, a sensitivity
study of how various vertical and horizontal mesh resolutions may impact
computational accuracy and efficiency is presented. For this paper a sensitivity
analysis will be performed to evaluate the effect of various mesh resolutions on
computational speed and accuracy.