Air Bubble Screens as a Tool for Water Quality Control

HELMUT E. KOBUS
fur
Privatdozent, Institut Hydromechanik und Sonderforschungsbereich 80
Universitat Karlsruhe, Kaiserstr. 12, 75 Karlsruhe, Germany.

SYNOPSIS
The flow field induced by the release of compressed air in a water body has
been studied analytically and experimentally. The quantitative knowledge of the
flow pattern in standing water and in a cross flow serves as a basis for the design
of air bubble installations, which can· be used advantageously for a number of
purposes in water quality control, such as pneumatic oil barriers, barriers against
density currents due to differences in salinity, silt or sediment concentration or
temperature, installations for combating the formation of ice, artificial des/ratifi-
cation devices or installations for oxygenation purposes.

1. Introduction 2. Basic Analysis

The conservation and wise use of our natural water
resources calls for an ever-increasing effort in water
quality control. This contribution discusses an engi-
neering device which holds considerable potential as a
useful tool for a variety of water quality conservation
tasks.
An air-bubble screen-generated by releasing com-
pressed air in a body of water-produces a flow field in
the receiving water body which can be used for engi-
neering purposes. Unfortunately, its first application as
a pneumatic breakwater more than 60 years ago,
disputed for over 40 years, turned out to be a failure.
The situation is quite different, however, with appli-
cations in water quality projects, where air-bubble
screens have been used successfully in recent years
(see Chapter 4).
A prerequisite for a successful and economic opera-
tion of air-bubble screens is a sound understanding of
the quantitative relationships between the flow field
generated by the bubbles and the parameters describing
the air · installation and the receiving water. In the
following, the results of analytical and experimental
research on this problem are presented and, wherever
possible, compared to. prototype observations. · This
leads to quantitative design information for air-bubble
screens in natural waters.
The air is supplied to the water through narrow
orifices from pressurized pipes or hoses. Under the
action of inertia and gravity, the discharging air jets
rapidly disintegrate into bubbles and are decelerated
within decimetres to their buoyant free rising velocity.
The main part of the flow field can hence be described
as a freely rising bubble stream(1). For discharges in a
sideways unlimited body of water otherwise at rest
(Figure 1), integral equations for continuity of the air-
and the waterflow and for vertical momentum flux
(assuming similarity profiles and accounting for the
change in bubble volume with pressure) can be formu-
lated(4-7). These yield a description of the flow
field with three empirical coefficients (entrainment
coefficient, turbulent Schmidt number, bubble slip
velocity). The same solution can be obtained by
treating the bubble swarm as a modified turbulent
plume, taking proper account of the bubble slip velo-
city and of the compressibility of the air.
The empirical coefficients . were obtained from
measurements of the velocity field and air concentration
distribution in a laboratory flume of two metres in
depth for the entire range of practically feasible air
discharges(10). With these results, the flow field can
now be predicted as a function of the air discharge and
the water depth.
At the free surface, the vertical plume is deflected

163
164 KOBt:S

Surface current Region of

surface
L effect:,;
rully
developed
flow

' ~ ~-
Exchange Auumption:
2
U ( X1 Z·l =e-(X/b)
U (Z) .
0

Return flow

-
.77.7.777.~~~~????7?7?~~~00~777/~~~~~~~~~~----i_j_Jl~ln~i~tj~ol
region
ow of orifices - x
analytical origin
FIGURE 1 : Flow field of air bubble screen.

and produces a horizontal surface current. The maxi- that the resulting surface currents in a cross flow can
mum horizontal velocity at the free surface can-as pro- now be predicted.
posed by G.I. Taylor( 1) and checked experimentally-
be taken as being approximately equal to the hypotheti- In a cross flow, the flow towards the barrier in the
cal · velocity on the plume axis that would be -attained lower layers is augmented, and the surface current in
at the elevation of the free surface if the latter were the downstream direction increases, whereas the region
not present. With this assumption, the plume analysis of return flow on the upstream side (which determines
data can be used to predict resulting surface velocities. the barrier action) decreases both in size and intensity,
A comparison of such predictions from various empiri- until, for very strong cross currents, it finally disap-
cal formulas and the authors analysis with all available . pears, so that no more barrier action is possible.
field data is given in Figure 2. The agreement between From experiments over a wide range of depths, air
analysis and observed data is satisfactory over the discharges and cross current velocities (Figure 4), an
entire range of conditions. Thus a tool is available for empirical relation of the form
predicting surface velocities for given air discharges v* v,.(UH=0) 2 UH
and water depths in standing bodies of water. . .. (I)
(gq0 )1J3 (gqo)lf3 - T (gqo)l/3
3~ Ambient Flow Conditions can be derived for the relative velocity v•. The acting
absolute "barrier" velocity Vm in the upstream direction
The ideal conditions assumed in the basic analysis follows as the difference between v* and UH to
are seldom met in nature : here one encounters fre-
quently cross flow currents or density gradients either Vm Vm(UH=O) 5 Uy
... (2)
of which may have a pronounced effect on the' perfor- (gq0 )1f3 (gq0 )1/3 3 . {gq,.)L/3
mance of the air-bubble screen. This relation gives an estimate of the influence of a
cross flow upon the surface velocity and can be used
3.1 Effect of a Cross Flow for design purposes until an analytical solution for this
A cross flow causes a deflection of the rising bubbles complicated flow configuration becomes available.
in the downstream direction and corresponding 3.2 Effect of a Density Gradient
changes of the flow pattern (Figure 3). These effects
have been studied experimentally, and analytically, so In stably stratified water bodies (such as reservoirs
 AIR BUBBLE SCREENS AS A TOOL FOR WATER QUALITY CONTROL 165

1.0 ,-----:-j--~,----------J:O""~------ .........--.....~

~
Buts on {
l(o/H 1.0
0.5
•
4)

x0 itf .. 0.4 O.S + D.

0.5 -r0.6 Q
Delft
{ 0.7 0
10 <>
() 2.0
.. Hensen x0 /H :- 0.3+0.6 X
..Q
(i) f<obus x 0 /H"' 0.2+0. 3 0
>
Cl>
u
t''onj ()
.E ~L-
0
L.
J

"'E .i :;:(
H
-.- _ _o •
--·----1::_-:::-- •L...,.._ -=-=.::.=..-:.=_---------
1:1 ___ .J_~- - - ~-------.-• ---------
:::l
_g 1.0 )./ -~--,_,____
~ _____ id_ __ ---------------------:.--
--;~f-
t- .
-.-
-- - - - - - - - - - - - - - - - - : - --::_-

~~go
X
0
;:

·o
0 0.25 0.50 c.: 5
H
Relative depth ( .- - )
H 0 +H
FIGURE 2 : Maximum surface velocity as a function of air discharge and depth.

with a summer temperature stratification, or ice-covered
freshwater lakes in winter) the upward transport of
water due to the bubble swarm is impeded by the
negative buoyancy of the entrained water. At pro-
nounced density interfaces (or when the total upward
momentum of bubbles and entrained water becomes
negative), some temporary decoupling between bubbles
and entrained water may occur (Figure 5). However,
after some time of operation the air bubbler will
achieve local destratification.
4. Design Information for Applications in Water
Quality Control
The results of the analysis and the experience collect-
ed with existing installations (both failures and
successes) yield a sound basis for designing air-bubble
screens and for predicting their performance and
evaluating the economics of operation.
4.1 Pneumatic Oil Barriers
The operating principle of a pneumatic oil barrier is
that the spreading tendency of the oil is counteracted
by the surface current induced by the bubble stream. A
simplified momentum equation (neglecting friction
losses, etc., and thus being on the "safe" side) can be
formulated for the configuration sketched in Figure 6
as
~v, 2 (D-8)+ Pwg(D-8) 2 = p., g D 2
2 2 2 (2) This value ~as to be augmented (safety factor
-(aML. + C1MW)
From this, the barrier velocity v, required to [retain the
oil film is given by
v,= / gD
\j
[l- Pm _2(aMw+~ML)J
Pw Pm gD ... (4 )
This relation is plotted in Figure 6. Since the influence
of surface tension is secondary, the required barrier
velocity is determined by the layer thickness and the
density of the mineral oil.
With this _all ~ecessa~y i~form_ation for the design
of a pneumatic 01! barner IS available. For a given
mineral oil of a given layer thickness, Equation {4)
yields the required barrier velocity at the surface to
which proper additions for safety have to be m~de.
For a given water depth, the bubble plume analysis now
yields the air discharge required to produce this barrier
velocity. All these design components are combined
in the nomogram given in Figure 7.
The various design steps are best illustrated by giving
a numerical example (see Figure 7). Assume an oil
harbour, which is to be protected against oil spills by a
pneumatic barrier across the harbour entrance. The
barrier is located at a depth of 8 m and is supposed to
retain a layer of gasoline (pm= 0.73 · t/m3 ) of 8 cm
thickness.
(1) By Equation (4), a surface velocity v, of 45
cm/s would be required.
... (3) c:= 1.5 e.g.) m order to safeguard against possible
166 KO BUS

....... _____ --------

Air discharge tqg Q,OOSrrfls

UH \"Upstream surface current

--=OA7
~

I --
H=l,80m_

UH S Upstream surface current

~=0,70 ... ,
------~------~-------.r-------------~r-----~----------
-
0
' \ ----- ' .,..
~
- - - ...,-
/

....
.....
---
-
·-
H=l,80m_

uH=0,30m/s-----·- - - t - - - - - - - -

-
Air discharge f q =0,008 m /s
0
2

FIGURE 3 :Flow pattern of an air-bubble screen in a cross flow.

 AIR BUBBLE SCREI!NS AS A TOOL FOR WATER QUALITY CONTROL . 16'i

2 .0

E
c::l
;
Ill
a.
:;:) l.O

-
-a
41
u
~
fxp.points
for H
H0 +H
Curves from eq..( f )
"0 D. 0.148
H
0 0.104 f<H

~~~ 0.5 0 0.055 H0 +H

0. 11.8
0.10 I.
0.055
>.
.~

~
41
>
-a 0.2
<11
u
~
-a
.=

Relative cross curren\

FIGURE 4: Maximum surface velocity in the upstream direction.

-+-- . _)QWD
:.:' (
c_ ...•
Density
) ::.....
..

1. Par\iQl decoupling at interface 2 .Complete decoupling of entrained

water and air bubbles at surfac•

FIGURE 5 : Possible How pattern at a density interface after Cederwall and Ditmars.

disturbances like plugged orifices, fluctua-
tions of the barrier flow, etc., and provided
with an addition for wind effects.
(3) Thus the air installation is to be designed for a
maximum surface current of 80 cm/s, which
requires, at a depth of 8 m, an atmospheric air
discharge of 0.017 m3/s.m or 1.05 m 3/min. m.
{4) For the compressor data and the orifice rating
curves, given in the third diagram, there results
a layout of the air installation of 2 mm orifices
along the barrier spaced 20 cm apart, operating
at a pipe pressure p; of 5.1 kp/cm1 absolute.
A more refined calculation would have to account for
the pipe losses and variations along the barrier, of
cours~. as is discussed in detail in Ref. (1°).
4.2 Pneumatic Barriers against Silt or Salt Water
Intrusion
Since air bubble screens produce strong vertical
mixing locally, they can be used for destratification
168 KOBVS
purposes (see para·s 4.3 and 4.4) and as barriers against
density currents (Figure 8). Air bubble installations
have been used successfully to reduce salt water intru-
sion into navigation locks joining freshwater canals
and the sea( 4). They can also be installed in estuaries
and canals or in harbours to prevent intrusion of silt
or suspended sediments. Furthermore, they may be
useful to prevent recirculation between cooling water
outlet structure and intake of power· plants.
If the air bubble stream is sufficiently strong, it
produces a flowfield as sketched in Figure 8, and the
density-induced exchange flow is greatly reduced by
the barrier action. The minimum air flow rate required
to establish this flow pattern depends upon the differen-
tial pressure acting upon the barrier (Figure 8). From
an extensive study(10), design information for predicting
the performance of an air bubble barrier against density
currents is available.
4.3 Pneumatic Installations for Preventing 1ce Forma-
tion
There are a number of hydraulic structures like
sluices, weirs, gates or piers which must be kept free of
ice in winter. Here also, air-bubble screens offer a
simple but efficient means of achieving this purpose.
Oil layer thickness D cm
In fresh water lakes and reservoirs, the warm water is
transported from the deep layers to the surface by the
FIGURE 6 : Surface velocity requirements for pneumatic oil action of the air bubbler, which thus makes use of the
barriers. thermal reserve of the entire water body (Figure 9).
1.50
f:. ?:-
V
u .2

.
0
<I>
<I> >
;>
100
u'"'
.2 E
~ u
., -::
'0
tl
L

::J
0'
<I>
a:

~----~----~~------~--~~0
0 0,01 0,02 0,03 0,04
Oil layer thickness 0 cm Atmospheric air discharge q 0 m3/s·m
•
'S
..
0
.D
0
Design do to: Example:

Max. oil volume } D a cm

basin surface
Oil den:;ily f,.. 73 o, t ;m3
Safely foetor £ 1,5
Wind correction
Water depth
0(
H(obs.pa)
12
8
cm/s
-m t 1·8 kp/cm2 1 ..•..
::>
Compressor dolo} b kp 7cm2 .,
losses in system

Layout:
to s.pi1mox 6
q 0 = I,05m3;min·m
.."
a.
0,001 0,002 0,003 0,004
ob,. P;• 51 1 kp/cm2; d • Z mmiAI: 20 cm .Potl'l\
Air discharge ( q O · - ) m;ts·m
3 (or c :0·7
PI 0
FIGURE 7: Nomograrn for the design of pneum~tic oil barriers.
 Flow pottern ofter Abrohom 141
~

----------
Air dischor~Je required for borrier:- oction
(l)Salt wat., (1) Fr~sh
wattr o.ooe
(2)Cotd wat~r (2) warm water 1-----------~:.......~:.:.:.:.::_:__ _~
Ill Silt -laden water [3)Cl~ar wat~r . >
~th H= OS 1.0 1.5 2.0m ;;
Cnti'eal discharge t:l:l
Optimal diKharge
•o · •o ... • 0 ....,/" c:
t:l:l
" t:l:l
/ r
Salt cont~nt 5o from 0.5 to 2.5"1. / l:l'l
0.005 /.0 en
/ ("')
/ ~
/ l:l'l
/ m
-==r zen
9"/
~"' / . >
E /"""-"P t 1mu m en
0 >
er 0.004 "'"' >-!
Salt water Frnh wat., 0
"' "'/ 0
V ..,01:""
/
•
Critical air d i st'harg~
~
/
"'· for barrirr action
/ ~
Salt wat~r intrusion / 'V. ~
m
0002 ~

. pres:>ure 0
___.... ,0 Max1mum )gH c:
diffrrPnc~(l'z-Y, ~ >
r
:::;
:'fi<&, . 92 ><
\ ("')
0
·~;.'\
$\ '\ z
\ >-!
00~--~----~----~~--~.~~~_j5
~" \ 2 3 ' ~
''
Ma1111num prts~ur~ diHt!rt>nct (t~ -~ )gH] P/c.rnl 0
r
' '- I 1 1
ws+:au e
j ', --- ~
c.wm»N.~W~&W~~'fit~~~&'*-',wwt'Yfll.. 'N1i.VifiWI»R-"9ff~\SI,~&.~,.urN. lmtrJ

Frto.sh wat~r escapt>

FIGURE 8 : Air bubble screen as a barrier against density currents.

$
-
 170 KOBUS

Heat losses due to ture stratification (Figure I 0) seems to be more promi-

radiation, evaporation
and convection sing for water quality improvement.
tieat losses through
ice and snow cover

~t/ 5. Conclusion
The flow field induced by an air-bubble screen in a
r r sideways unlimited body of water has been investigated
analytically and experimentally, and the effects of a
Heat potential Wr cross flow and of a density stratification have been
of water body
studied. The description of the flow field serves as a
basis for a quantitative analysis of air-bubble installa-
tions for various applications in water quality control :
lhmWTaw//!1/~aw pneumatic oil barriers, installations for combating
density currents due to temperature, salinity or
Heat influx
.
A•r bubble
screen
sediment concentration gradients, for preventing ice-
from bottom formation, for density-destratification of lakes and
FIGURE 9 : Air-bubble screen for ice protection. reservoirs, or for in-stream aeration. These applications
are briefly discussed ; for a detailed account, the reader
is referred to Ref. (1°), where a quantitative basis is given
4.4 Air~bubbleInstallations for Destratification and for assessing the feasibility of air-bubble installations
Oxygenation both technically and economically.

Freshwater lakes and reservoirs develop a thermal 6. Selected References

stratification during summer, in which the warm,
oxygen-rich surface waters (epilimnion) are separated
by a sharp thermocline from the cold, lower layers
(hypolimnion), which can deteriorate to a very poor
water quality and be depleted of oxygen. At the ther-
mocline, no natural exchange processes occur any more
between the two layers.
There are basically two approaches to the problem
of improving the water quality in the hypolimnion
(Figure 10). One can cause an artificial destratification,
so that natural exchange processes in the vertical can
take place again. It seems, however, that the vertical
mixing necessary in this procedure may also have
detrimental effects upon the water quality. Therefore,
the use of various devices for oxygenation of the
hypolimnion without disturbing the natural tempera-
(1) G .I. TAYLOR : "The Action of a Surface Current Used
as a Breakwater." Proc., Royal Society of London,
Series A, Vol. 231, 1955.
(2) E. STEHR : "Berechnungsgrundlagen fu·r PreBluft-O.lsp-
erren." Mitteilungen der Hannover'schen Versuchsans-
talt fur Grundbau und Wasserbau, Franzius-Institut der
TH Hannover, Heft 16, S. 275 ff, 1959.
(3) P.S. BULSON : "Currents Produced by an Air Curtain
in Deep Water." The Dock and Harbour Authority,
May 1961.
(4) G. ABRAHAM und P.v.d. BURGH : "Reduction of
Salt Water Intrusion through Locks by Pneumatic
Barriers.'' ASCE Journal of the Hydraulics Division,
Hy 1, Jan. 1964.
(5) A. SJOBERG ''Stromningshastigheter kring

s~tup of the umno·

Air bubble se run \lot lex rings Pumping Wahnbachtalsp«r.,.,.«band

~ttlods ~ Oftificial dutratificatioll H,YPolimnion- a~ration

FIGURE 10: Various methods of aerating lakes and reservoirs.

 AIR BUBBLE SCREENS AS A TOOL FOR WATER QUALITY CONTROL
luftbubbelrid; i tathetshomogent ach still;st;ende vatten."
Meddelande nr. 39, Institutionen for vattenbyggnad,
Chalmers Tekniska Hogskola, Goteborg 1967.
(6) H. KOBUS : "Analysis of the Flow Induced by Air-
Bubble Systems." Proc., Xlth International Conference
on Coastal Engineering, London, Sept. 1968.
(7) K. CEDERWALL und J.D. DITMARS : "Analysis of
Air-Bubble-Plumes." California Institute of Technology,
W.M. Keck Laboratory of Hydraulics and Water Resour-
ces, Report No. KH-R-24, Sept. 1970.
171
(8) J. GRACE und A. SOWYRDA :"The Development and
Evaluation of a Pneumatic Barrier for Restraining Sur-
face Oils in a River." Journal Water Pollution Control
Federation, Vol. 42, No. 12, pp . 2074-2093, December
1970.
(9) B. VERNER: " Pneumatic Oil Barriers." Atlas Copco,
tryckluft 1, 1972.
(10) H. KOBUS : "Bemessungsgrundlagen und Anwendungen
f~·r Luftschleier im Wasserbau." Heft 7, Wasser und
Abwasser in Forschung Proxis, Erich Schmidt Verlag,
Berlin, 1973.