68    ENERGY EFFECTS IN
DISSOLVED GAS FLOTATION
Melbourne L. Jackson, Professor Emeritus
Department of Chemical Engineering
University of Idaho
Moscow, Idaho 83843
INTRODUCTION
A previous study' of the formation of bubbles by desorption of gas from saturated water
demonstrated that the discharge or transfer pressure, as separate from the saturation pressure, had
a defining function: the bubble size decreased and the number of bubbles increased as the transfer pressure increased. The excess pressure over that lost to friction was shown to provide the energy for bubble nucleation and surface formation. In very clean water only, large bubbles formed
but the addition of a small amount of a surface active agent provided low energy nucleation sites
such that very minute bubbles filled the column with a cloud-like appearance.
The current state of knowledge of flotation, for both dispersed and dissolved gas processes,
has been considered.2-3 The recovery of biological cells by flotation has been recommended.9
The use of residual gases from treatment processes under high hydrostatic heads, without the
need for additional pressurization, has been employed5-6 for solids separation by flotation.
Experimental Procedures
The use of a two-column system for saturation followed by transfer to a flotation column is
shown in Figure 1 and is described in detail.' Pressures employed are P (saturation) and Pd
(transfer), both as kPa gauge, and Pb (barometric). The time of rise of the bottom of the flotation
layer was observed visually and as recorded by a video camera. The percent input solids, S(, was
determined by filtration and drying. The percent solids in the float, S, at any time of flotation,
was determined by the relation:
S = Sj F/(F - D - V/45.6)
where F is the filling level, D is the level of the bottom of the float, V is the volume of gas at Pb
associated with the suspended solids, and 45.6 cm2 is the cross-sectional area of the column. V is
determined using Henrys' law at barometric pressure for 95% retention of the dissolved gas in
the float. The filling level was maintained at about 82 cm and measured. Flotation was for suspended solids saturated directly before transfer; a few runs were made for comparison with use
of recycle where clear effluent is saturated and mixed with the unsaturated suspension. Most determinations were at room temperature of 23°C.
The time to attain a saturation approaching 100% was evaluated to assure that the gas affecting
flotation was accurately known. An oxygen electrode was used t observe the unsteady-state
transfer of oxygen from air to water at barometric pressure (Pb = 0.92 kPa) using a ceramic diffuser. Clean water required 3.7 minutes to attain 98% saturation and wastewater liquid needed
5.3 min, which gives an alpha factor ratio of 0.7. Use of a coarse bubble inlet required much
longer times of 14 min for clean water and 20 min for the wastewater. Volumetric transfer factors
determined for elevated pressures using the ball diffuser showed the need for 8 min saturation of
wastewater at P = 103 and 11 minutes at 207. Coarse bubble inlets increased the required time by
factors of four. These values are also of interest in evaluating the results of other laboratory procedures.
51st Purdue Industrial Waste Conference Proceedings, 1996, Ann Arbor Press, Inc., Chelsea, Michigan 48118.
Printed in U.S.A.
663