Pumping Optimization of Coastal Aquifers using 3-D ...

Object Details


XVI International Conference on Computational Methods in Water Resources (CMWR-XVI) Ingeniørhuset

Pumping Optimization of Coastal Aquifers using 3-D Density Models and Approximations with Neural Networks
Author:George Kourakos <giorgk@gmail.com> (National Technical University of Athens)
Aristotelis Mantoglou <mantog@central.ntua.gr> (National Technical University of Athens)
Presenter:George Kourakos <giorgk@gmail.com> (National Technical University of Athens)
Date: 2006-06-18     Track: Special Sessions     Session: Modeling and managing coastal aquifers

A method for pumping optimization of coastal aquifers using a 3-D Density Model as well as model approximations with Neural Networks is developed. The numerical code SEAWAT was utilized to evaluate the distribution of salts in the brackish zone for various pumping scenarios. Since pumping optimization using SEAWAT is CPU time intensive, a model approximation using neural networks was developed. Then the optimization utilized simple neural network approximations rather than the complex partial differential equations of flow and transport. The objective is to maximize the total pumping rates subject to constraints that protect the aquifer from saltwater intrusion. The formulation of the constraints is based on the neural network model approximations. The results based on the model approximations are compared to the "exact" results based on the SEAWAT solutions and a good correlation was found. The optimization methodology is applied effectively for determining the assessment of optimal pumping rates in an unconfined coastal aquifer in a Greek island.