This thesis investigates the technical problems associated with large-scale stand-alone
wind powered desalination employing a short-term energy store, particularly the
complexities associated with the intermittent operation of the desalination plant.
To achieve this, a non-linear, time domain system model of an existing wind powered
desalination plant has been developed using the propriety code Simulink.
Two desalination techniques have been considered: reverse osmosis and electrodialysis,
due firstly to their relatively low specific energy consumption, and secondly, their
efficient coupling to a wind turbine generator.
As a way of reducing power mismatch, optimising water production, and above all
reducing the switching rates of the desalination units, operation of the reverse osmosis
and electrodialysis units under variable power conditions is suggested.
Little information is available on plant performance under such conditions. A
mathematical model has therefore been developed to ascertain the performance of
reverse osmosis and electrodialysis processes under transient power conditions. The
model consists of the set of partial differential equations (PDEs) describing the
conservation of mass, momentum and chemical species coupled with the appropriate
boundary conditions. A numerical solution based on the finite volume method has been
employed to solve for the system of PDEs, as no analytical solution is available for the
particular set of model equations derived.
Sensitivity of plant performance to key design parameters (such as operating pressure
and energy storage capacity) and operational strategies is predicted from simulation
results.
This technology is economically attractive for islands where wind energy density is high
and water resources are scarce.
History
School
Mechanical, Electrical and Manufacturing Engineering