Scholarly
    Communications Project


Document Type:Dissertation
Name:Khawla AbdulMohsen Al-Shayji
Email address:khw95@vt.edu
URN:1998/00475
Title:Modeling, Simulation, and Optimization of large-Scale Commercial Desalination Plants
Degree:Doctor of Philosophy
Department:Chemical Engineering
Committee Chair: Y.A. Liu
Chair's email:design@vt.edu
Committee Members:Hanif D. Sherali
Hugh F. Vanlandingham
William H. Velander
William L. Conger
Keywords:Neural Network, Artificial Intelligence, Desalination, Multistage Flash, Reverse Osmosis, Modeling, Simulation, Optimization, Linearization, Interaction Analysis
Date of defense:April 17, 1998
Availability:Release the entire work for Virginia Tech access only.
After one year release worldwide only with written permission of the student and the advisory committee chair.

Abstract:

This dissertation introduces desalination processes in general and multistage flash (MSF) and reverse osmosis (RO) in particular. It presents the fundamental and practical aspects of neural networks and provides an overview of their structures, topology, strengths, and limitations. This study includes the neural network applications to prediction problems of large-scale commercial MSF and RO desalination plants in conjunction with statistical techniques to identify the major independent variables to optimize the process performance.

In contrast to several recent studies, this work utilizes actual operating data (not simulated) from a large-scale commercial MSF desalination plant (48 million gallons per day capacity, MGPD) and RO plant (15 MGPD) located in Kuwait and the Kingdom of Saudi Arabia, respectively. We apply Neural Works Professional II/Plus (NeuralWare, 1993) and SAS (SAS Institute Inc., 1996) software to accomplish this task.

This dissertation demonstrates how to apply modular and equation-solving approaches for steady-state and dynamic simulations of large-scale commercial MSF desalination plants using ASPEN PLUS (Advanced System for Process Engineering PLUS) and SPEEDUP (Simulation Program for Evaluation and Evolutionary Design of Unsteady Processes) marketed by Aspen Technology, Cambridge, MA.

This work illustrates the development of an optimal operating envelope for achieving a stable operation of a commercial MSF desalination plant using the SPEEDUP model. We then discuss model linearization around nominal operating conditions and arrive at pairing schemes for manipulated and controlled variables by interaction analysis. Finally, this dissertation describes our experience in applying a commercial software, DynaPLUS, for combined steady-state and dynamic simulations of a commercial MSF desalination plant.

This dissertation is unique and significant in that it reports the first comprehensive study of predictive modeling, simulation, and optimization of large-scale commercial desalination plants. It is the first detailed and comparative study of commercial desalination plants using both artificial intelligence and computer-aided design techniques. The resulting models are able to reproduce accurately the actual operating data and to predict the optimal operating conditions of commercial desalination plants.


List of Attached Files

ETD.PDF

At the author's request, all materials (PDF files, images, etc.) associated with this ETD are accessible from the Virginia Tech network only.


The author grants to Virginia Tech or its agents the right to archive and display their thesis or dissertation in whole or in part in the University Libraries in all forms of media, now or hereafter known. The author retains all proprietary rights, such as patent rights. The author also retains the right to use in future works (such as articles or books) all or part of this thesis or dissertation.