DESIGN-ORIENTED REVERSE OSMOSIS MODEL
John F. Metzger, Graduate Student
Donald Angelbeck, Associate Professor
Civil Engineering Department
University of Toledo
Toledo, Ohio 43606
Over the past decade, effluent discharge standards have become increasingly more stringent. In some industrial categories, even zero discharge has been proposed as best available
technology. As such, zero discharge can consist of either a true reduction of hydraulically
discharged water to zero through total recirculation systems or a discharge with suspended
and dissolved contaminants (contaminant mass) reduced to levels equal to those in intake
waters. In the case of total recirculation, dissolved solids must be continuously removed to
control high dissolved solids levels that are caused by evaporative hydraulic losses within the
system; otherwise, scaling or corrosive conditions will result.
There are several methods of demineralization available: ion exchange, electrodialysis,
evaporation and solvent extraction. Reverse osmosis may provide a significant advantage
over the others in that the process can remove a wide range of chemical species including
many dissolved organics that may have escaped removal by classical, upstream processes.
Design of reliable, low cost, reverse osmosis systems present many difficult engineering
problems. Some of these problems include: (1) fragile membrane support design to sustain
differential pressures of 300 to 1500 psi, (2) separation of the high pressure feed and brine
streams from low pressure product water, (3) development of high density packaging to
minimize pressure vessel cost, (4) minimizing concentration polarization and fouling effects
in feed channels, (5) overcoming parasitic pressure drops in feed, brine and product streams,
and (6) development of low cost replacement membranes.
Much of the research and development related to reverse osmosis in the past five to ten
years has been done in the interest of desalting water supplies. These efforts have resulted in
some innovative reverse osmosis system designs which have reduced many problems associated with widespread application of the process.
The predominant problems related to wastewater treatment are, most probably, related
to membrane stability and membrane fouling through concentration polarization and/or
physical plugging. As in any field of rapidly increasing technological improvement, it may be
safe to assume that reverse osmosis system improvements will be made, and as discharge
standards require increasingly exotic treatment solutions, the practical design engineer will
need accurate and unencumbered methods or models to represent the reverse osmosis system.
Much information detailing theoretical aspects of reverse osmosis systems can be found in
the literature. Lakshminarayanaiah [1], Friedlander and Rickles [2] and Rickles [3] have
studied the process in some detail. However, little information is given offering design techniques and equations that are both reliable and easily applied. For example, a model devised
by Tweddle et al. [4], involves eight equations for eight dimensionless quantities. Weber [5]
offers a somewhat simpler approach, but this model requires an iterative solution of the
primary design equations.
The remainder of this chapter proposes a direct model that can potentially be used to
design and operate reverse osmosis systems as well as to define operational field data. In its
present form, this model successfully describes reverse osmosis systems. With future refinements, it should be an even more effective and simple tool for the theoretically complicated
process of reverse osmosis.
785