-------
A LABORATORY TECHNIQUE FOR INVESTIGATING THE RELATIONSHIP BETWEEN GAS
TRANSFER AND FLUID TURBULENCE
T. Dickey, B. Hartman, D. Hammond, E. Hurst
University of Southern California
A technique has been developed to simultaneously measure fluid turbulence
and the exchange rate of gases across the air-water interface of a tank under
controlled laboratory conditions. Turbulence is generated with a vertically
oscillating grid whose stroke length, frequency, and distance from the
interface may be varied. Instantaneous fluid velocity vectors are determined
using a streak photography method. Tracks of neutrally buoyant particles are
recorded photographically, digitized, and the displacement is determined.
Exposure times are controlled by a light chopping wheel and measured with a
photo-diode and timer system. Ensemble average turbulent velocities and
turbulent integral length scales are then computed as functions of distance
from the grid.
The gas transfer rates of five gases (0 , N , CH , CO , and Rn) are
2242
measured concurrently with the turbulence measurements described above.
Before each run, tank water concentrations of 0 and N are reduced to about
2 2
50% of saturation by stripping the tank water using helium. Concentrations of
CO , CH , and Rn are elevated above saturation by bubbling CO and CH through
24 24
the water and by adding water with high Rn activity. The water is stirred to
insure homogeneous conditions before initiation of the experiment. Small
samples of water are taken at various time intervals to determine the rate of
change in gas concentrations. The mass transfer coefficient for each gas is
computed from this data, assuming the rate of exchange is proportional to the
deviation from atmospheric equilibrium. Experiments have been conducted
utilizing three different locations of the grid and two oscillation
frequencies. The turbulent velocity decreased nearly exponentially with
distance from the grid while the turbulent integral length scale increased
-------
Dickey et al.
Page 2
linearly with distance from the grid. The mass transfer coefficients appear
to be proportional to the square root of the molecular diffusivities and to
the square root of the ratio of the turbulent velocity to the turbulent
integral length scale. This is in accordance with the surface renewal concept
of gas exchange.
-------
Duran and Hemond
Duran,A.P.*, and H.F. Hemond.**
Dichlorodifluoromethane (Freon-12) as a Tracer for Nitrous Oxide Release from
a Polluted River
A Freon-12 tracer technique has been developed to measure gas exchange
coefficients for nitrous oxide in a reach of the Assabet River receiving
the effluent of the Westborough,Mass. sewage treatment plant. Stirred-tank
experiments were performed to measure the relative exchange coefficients of
nitrous oxide and Freon-12. Nitrous oxide is produced during chlorination of
sewage and by microbial metabolism and is highly oversaturated in rivers
receiving treated sewage effluents. Release of N20 to the atmosphere can
contribute to significant destruction of stratospheric ozone and to "greenhouse"
warming. Magnitudes of the nitrous oxide exchange coefficient have been determined
with greater precision than is possible through use of predictive equations.
Preliminary results also suggest a strong wind effect not accounted for by
predictions based on depth and velocity. The Freon-12 tracer technique facilitates
the determination of the respective biogeochemical roles of sediment and water
column with respect to ^0 production. The required sampling for this tracer
technique is logistically simple and the sample analysis is rapid.
* Division of Applied Science, Harvard University
** Department of Civil Engineering, Massachusetts Institute of Technology
-------
Fabre et a1.
TURBULENCE STRUCTURE
OF WAVY STRATIFIED AIR-WATER FLOW
FABRE J. , MARODON D. , MASBERNAT L. , SUZANNE C. (*)
This work deals with experimental studies of near hrozontal
stratified gas-liquid flows. Experimental results on the local turbulent
structure presented here concern two phase flow modelization in pipes
and the prediction of interfacial exchanges in environmental air-water
flows.
Indeed, two sorts of flow conditions were studied: one flow
is an internal fully developpedstratified flow where the turbulence is
controlled by fluid-wall and fluid-fluid interactions; the second is a
developping stratified flow where interfacial interactions have the main
role in turbulence production. These flows are realized in rectangular
channels with the following dimensions:
-internal flow in a channel 12 m. long, O.lm.deep, 0.2m.large
-developping flow in a channel 18m.long,0.6 m.large, 1. m. deep,
where the mean water depth is near 0.2 m.
The kinematic analysis concerns the monophasic regions of
the flow and the interfacial two phase region.
In the monophasic regions, longitudinal and vertical mean
velocity components, turbulence intensities and Reynolds shear stress
were measured ,by laser doppler anemometry in gas phase. In the two
phase region, a new measurement method, using LDA, was perfected to
obtain 1st and 2nd moments of the longitudinal velocity in liquid.
Capacitive probes were used to obtain instantaneous interfacial height
(*) Institut de Mecanique des Fluides / 2,rue Camichel / 31 IOULDDSE
-------
Fabre et al.
Page 2
from which liquid time fraction and statistical wave characteristics were
determined.
Main features of air and water flow are displayed by the
examination of dimensionless 1st and 2nd moments profiles. In the mono-
phasic regions near the walls and in the gas phase near the interface
the classical shear flow behaviour is shown off. On the other hand, the
liquid region under the waves exhibits particular features: turbulence
intensities and shear stress profiles are showing the important role
of turbulence transport from the wavy region; a consequence of this fact
is the failing of Boussinesq concept for shear stress in this region,
and the existence of secondary flow (which could be measured in internal
flow).
A very crude scaling of Reynolds tensor component equations
allows to confirm the main role of transport under the wavy region and
suggests a. functional formulation for shear stress, linked with the
turbulent fluxes issued from the waves.
REFERENCES for ABSTRACT
(1) FABRE J. , MASBERNAT L. , SUZANNE C.
New results on the structure of stratified gas liquid flow.
Two-phase flow workshop . August 1982. Spitzingsee Germany
(2) AISA L. , CAUSSADE B. , GEORGE J. , MASBERNAT L.
Echanges de gaz dissous en ecoulements stratifies de gaz et de liquide
Int. J. of Heat and Mass transfer. Vol 24. N 6. 1981
(3) GAYRAL B. , MASBERNAT L. , SUZANNE C.
Mean velocities and Reynolds stresses in co-current gas-liquid
stratified channel flow.
Two-phase Momentum Heat and Mass Transfer in Chemical Process and
Energy Engineering Systems. Vol 2. Edited by F. DURST, G.V. TSIKLAURI
and N.H. AFGAN Hemisphere Publishing Corporation.
-------
Fain and Vigander
EFFECTS OF DIFFUSER SPACING AND
SUBMERGENCE DEPTH ON OXYGEN TRANSFER
EFFICIENCY USING THE SMALL OXYGEN BUBBLE
REOXYGENATION TECHNIQUE
by
T. G. Fain* & S. Vigandert
Both Research Engineers, TVA Engineering Laboratory, Norris, Tennessee
* Member, ASME
t Member, ASCE; Member IAHR
-------
Fain and Vigander
ABSTRACT
The turbine^of dams with deep reservoirs and low level intakes may
discharge water with unacceptably low dissolved oxygen concentration (DO)
during the summer and fall months of the year. This is caused by strati-
fication of the reservoir and the subsequent natural depletion of DO in its
hypolimnion. Hence reoxygenation of the turbine discharges may be necessary
to maintain a healthy environment for fish and other acquatic organisms in
the tail race. One reoxygenation technique is to inject small bubbles of
gaseous oxygen into the reservoir with porous diffusers. The diffusers are
located on the bottom of the reservoir, just upstream from the turbine intake.
Oxygen in the bubbles rising from the diffusers dissolves and enriches the
water entering the turbine.
A pilot scale study of the injection technique was performed by TVA to
demonstrate feasability of the method and to develop design criteria. An oxygen
transfer efficiency was defined as the ratio of mass of oxygen dissolved to the
mass of gaseous oxygen injection. The study included field tests to determine
the effects on oxygen transfer efficiency of diffuser pore size, diffuser spacing,
diffuser submergence depth, and oxygen flux. This paper presents results of the
field tests.
The test procedure consisted of collecting samples of the bubble plume
gas at the water surface and analyzing them for oxygen content. Bubbles
leaving the diffusers contained pure oxygen, but when they reached the water
surface, they contained both oxygen and other gasses picked up from the
water--mostly nitrogen. The fraction of oxygen not dissolved was used as an
indicator of oxygen transfer efficiency because an accurate oxygen balance
could not be obtained in the field tests. The oxygen fraction in the escaping
-------
2 Fain and Vigander
bubbles was correlated with the oxygen transfer efficiency on the basis of
laboratory test data for which a complete oxygen balance was possible.
In these tests, diffuser pore sizes ranged from 1.2 to 20 microns. The
diffuser spacing ranged from a single, isolated diffuser (hollow cylinder,
18 inches long by 3 inches OD) to a battery of 136 of the same, mounted about
5 inches apart. Submergence depths ranged to 43 feet. Oxygen fluxes ranges
to 0.8 acfm/ft2.
Test results indicated that transfer efficiency decreased with increasing
pore size, increased with increasing spacing, and decreased with increasing
oxygen flux. The transfer efficiency increased with increasing submergence
depth, but became independent of submergance depth at a point which depended
on oxygen flux and diffuser spacing.
-------
Faller
ABSTRACT
International Symposium on Gas Transfer at Water Surfaces
Cornell University, Ithaca , N.Y., 13-15 June 1983
Laboratory experiments with small-amplitude O(0.1 cm)
monochromatic plane waves on water (wave length O(20 cm))
and weak shear , O(0.5 s ), clearly demonstrate the generation
of convective longitudinal vortices as predicted by the Craik-
Leibovich (Craik, 1977; Leibovich and Paolucci, 1981) mechanism for
the generation of Langmuir circulations. The shear may be produced
by wind, a forced surface film motion, or wave-driven shear as
predicted by Craik (1982).
The initial occurrence of longitudinal rolls is rapidly followed
by deepening of the shear layer, the generation of larger-scale
rolls, and transition to turbulence. The experiments and theory taken
together show that with almost any observable surface waves
there will be mechanically generated convective overturning with
longitudinal .rolls (the Langmuir circulations) and turbulence. Gas
transfer will then be limited only by the thinnest layer at the surface
in which diffusion may be pres; roed to dominate. But even there
the Langmuir circulations will be important through their tendency
to concentrate surface films into lines parallel to the shear ,
thus presenting a significantly non-uniform surface layer .
Alan J. Faller
Research Professor
Institute for Physical Science and
Technology
University of Maryland, College Park
College Park , Maryland, 20742
-------
Giorgetti and Giansanti
EVALUATION OP TURBULENCE LEVEL IN FLOWING WATEH
NEAR THE AIR-WA-JSR INTERFACE
Marcitis P. Giorgetti, Professor of Hydraulics & Sanitary Engineering
Antonio E. Giansanti, Graduate Student
A general research program is being carried with the objective
of developing materials and methods for the inexpressive gross evaluation of
turbulence intensity near air-water interface in flowing waters.
It is thought that with one such additional parameter formulas
for the prediction of reaeration coefficients could be significantly improved.
The method is based on the dependence of the solubility of a solid immersed
in a liquid upon the level of turbulence of the liquid. A number of floating
probes of known weight dumped in a stream of water could be (partially)
recaptured downstream, dryed and weighted, thus giving a gross information on
the average rate of dissolution of the solid in water.
A report is presented of a series of tests carried out with
floating probes of salts and sugar compunds in a tank of water subjected to
convective turbulence. Turbulence is induced tj submerged jets; the geometry
of the apparatus and the flow were ;^sdG such that near the surface the
observed pattern duplicated that which is visualy observed in flovrinc streams
c.r -.T.tc.r. Probes made out of sugar compounds revealed sood sensibility to
changes of -turbulence levels. Turther tests will be conducted in a lone-
channel with bottom shear generated turbulence.
-------
Gulliver and Stefan
ABSTRACT for
International Symposium on
Gas Transfer at Water Surfaces
PREDICTION OF NON-REACTIVE WATER SURFACE
GAS EXCHANGE IN STREAMS AND LAKES
by John S. Gulliver and Heinz G. Stefan
Assistant Professor and Professor and Associate Director,
respectively
St. Anthony Falls Hydraulic Laboratory
Department of Civil and Mineral Engineering
University of Minnesota
A model is formulated for the water surface exchange of a non-reactive
or weakly reacting gas such as oxygen in streams and lakes. The model in-
corporates a molecular layer at the water surface, -vertical turbulent
diffusion due to bottom shear and wind shear, and the enhancement of surface
exchange by wind waves.
The mathematical model gives an equation for the gas exchange coefficient,
/s) which includes molecular transport in th
lent transport below. The predictive equation is
K (m/s) which includes molecular transport in the surface layer and turbu-
5
Ks=
I m
where 6 = thickness of molecular surface layer,
D = molecular diffusivity of dissolved gas in water,
m
K = Von Karman's constant,
U * = sum of bottom and water surface shear,
c
Pe = shear Peclet number = U * h/D ,
c m
and h = water depth.
K /h = K_ is- the more commonly used "reaeration coefficient."
S ™
-------
Gulliver and Stefan
A comparison of the predictive equation with flume observations reported
in the literature indicates that the molecular surface layer does exist, con-
trols surface exchange in most cases, and may be described by the equation
6 = A. v/U* (2)
where v = kinematic viscosity of water,
D* = bottom shear velocity in flume, and
A. - a constant, - 10.
Wind-wave flume measurements of K available in the literature are
s
used to expand Eq. 2 to incorporate wind shear and the effect of wind waves.
The surface layer thickness is then described by the equation
-1/2
(3)
where U * = wind shear velocity
&
$ = energy spectrum of wind waves,
f = wave frequency,
A_ = a constant - 14, and
-8-1
A = a constant = 2.3 x 10 sec
The three terms in Eq. 3 represent the contributions of bottom shear, wind
shear, and the circulation under wind waves, respectively. Under high winds,
i.e. 12 m/sec, the wave spectrum term in Eq. 3 will contribute only about
5 percent to K . If wind velocity is less than about 6 m/sec, however, the
S
wave spectrum term can assume an importance equal to that of the wind shear
term in computing K . Separate consideration of wind effects on the surface
molecular layer and on waves in addition to bed induced shear improves the
predictive capabilities beyond available theories.
The superposition form given in Eq. 3 is chosen because the theory is
adding the impact of various sources of turbulence, and turbulence character-
istics sum with the square of shear velocity or velocity components.
-------
Hartman et a I.
PREDICTION OF GAS EXCHANGE RATES IN WIND-DOMINATED NATURAL SYSTEMS
By B. Hartman, D. Hammond, T. Dickey
University of Southern California
The results from laboratory and field investigations on the processes
controlling gas exchange across the air-water interface are presented • and used
to develop a predictive relationship for gas exchange rates in wind-dominated
natural systems. Recent laboratory measurements of the exchange rates of
several gases and of parameters characteristic of the fluid turbulence suggest
that gas transport can be adequately described by a surface renewal model
based on fluid turbulent velocities and length scales.
In the field, the rate of gas exchange in south San Francisco Bay is
primarily controlled by wind speed. However, application of published models,
which characterize fluid turbulence based on the wind friction velocity and
water depth, predicts gas exchange rates which are 2-3 times lower than
observed rates. Furthermore, they do not predict the observed functional
dependence of gas exchange on wind speed.
The discrepancy between the success of the turbulence approach in the
laboratory and its failure in the field is probably due to the absence of
adequate means to measure or predict from wind speed the necessary parameters
of fluid turbulence. Attempts to use salt dissolution rates as a technique
to characterize fluid turbulence have proved unsuccessful to date.
Compilation of our field data with previous measurements of exchange
rates in wind-dominated systems suggests that average exchange rates can be
predicted to within 20* based on wind speed alone. Until measurements of the
necessary turbulence parameters can be accomplished these empirical
relationships based upon wind speed must suffice.
-------
Holley and Yotsukura
(Invited Keynote Paper)
ABSTRACT
Field Experimental Techniques for Reaeration Studies
by Edward R. Holley and Nobuhiro Yotsukura*5*"
This paper is concerned primarily with gas exchange measurements in rivers
with secondary attention given to estuaries, lakes, and reservoirs. Also, the
focus is on reaeration problems, with attention being given first to traditional
DO measurements and then to tracer gas measurements.
One of the first approaches was based on the mass balance of oxygen with
surface reaeration being evaluated as the residual after measuring DO distri-
butions and then accounting for other sources and sinks of DO. The accuracy
of reaeration determination with this method is limited by the accuracy of eva-
luation of the other sources and sinks, some of which are unsteady and therefore
prevent testing during steady-state periods. A disturbed equilibrium method has
also been used with DO measurements being taken first under natural conditions
and then with a steady-state injection of catalized sodium sulfite, which lowers
the DO concentration. The change in DO concentration between the two conditions
is used to determine the reaeration, assuming that all sources and sinks except
the surface transfer are unaffected by the sodium sulfite. The method has been
used only for small streams where the sodium sulfite became mixed relatively
quickly across the full stream width.
Most studies have assumed that DO concentrations are uniform across the
width of a river and have used ID representations. The data of Churchill et al.
shows that the deficit may vary as much as 15% across the width under natural
conditions (no BOD loads). The importance of natural transverse variations in
a given river can be estimated using Holley's result that the amount of variation
of deficit across the width is related to the ratio of a characteristic transverse
mixing time to a characteristic reaeration time.
There have been relatively few reaeration measurements in wide estuaries,
lakes, and reservoirs where the flow pattern is highly two-dimensional and/or where
a significant part of the reaeration is associated with wind induced turbulence.
One method that has been used is a floating dome with surface gas transfer being
determined by the change of pressure in the gas trapped between the dome and the
water surface. Problems with this method include heating of the dome by inci-
dent radiation, keeping the gas space under the dome sealed from the atmosphere
when surface waves are present, and possibly disturbing the natural turbulence
structure at the surface where the gas transfer is taking place.
Tracer gas techniques were developed in an effort to measure gas transfer
without interferences from the various natural sources and sinks of DO in rivers.
In Tsivoglu's method, a bottle containing dissolved radioactive krypton, tritium,
and Rhodamine WT is broken below the water surface. Using the Rhodamine as a
location tracer, the difference in measured concentrations of krypton and
tritium, a conservative tracer, can be related to the surface desorption
of krypton. In the USGS method, hydrocarbon gases are bubbled through the
depth of the river and Rhodamine WT is used as a conservative tracer. The
injection time may be short (15-30 minutes) to give a moving cloud of tracer
gas. The method is then similar to the krypton method. In another technique,
* University of Texas, Austin, TX
** U.S. Geological Survey, Reston, VA
-------
-2- Holley and Yotsukura
the hydrocarbon tracer gas is injected long enough to give steady-state concen-
trations. In small streams with complete transverse mixing, the difference in
tracer gas concentrations between cross sections can be related to the surface
exchange.
The krypton method suffers from public concern about the introduction of
radioactive materials into the environment. The hydrocarbon-gas methods
suffer from the inconvenience of setting up the bubbling apparatus in the river
and from needing to correct for loss of Rhodamine WT dye. Both the krypton
and the hydrocarbon gas methods have been used primarily in small streams where
transverse mixing could be achieved in the study reach. Investigations are
currently underway on using the steady-state hydrocarbon gas method with an
unmixed plume in larger rivers. The tracer gas methods appear to hold much promise
for future measurements.
(For each of the methods, the practical aspects of actually executing the
tests under field conditions will be addressed.)
-------
Hovis et al
ABSTRACT
GAS TRANSFER RATE COEFFICIENT MEASUREMENT
BY A STABLE ISOTOPE KRYPTON TECHNIQUE
I. TREATMENT PLANT STUDIES
J.S. Hovis , J.J. McKeown ,
2 't
D. Krause,Jr. , and B.B. Benson"
1) National Council of the Paper Industry for Air and Stream
Improvement, Inc., Northeast Regional Center, Tufts
University, Medford, Massachusetts 02155
2) Department of Physics, Amherst College, Amherst,
Massachusetts 01002
-------
Hovis et al.
Page 2
A new pair of tracers for gas transfer measurement have
been developed for use in turbulent, open systems. Stable
isotope krypton (Kr), with atmospheric composition, and lithium
g c O
(Li) salts replace the radioactive tracers Kr and H, respectively,
as the gas and dispersion tracers in the direct tracer method
developed by Tsivoglou, et. a_l (1965, 1967, 1968) . The primary
application of this technique has been in the precise determination
of oxygen transfer rate coefficients (a kTa ) in respiring wastewater
LI
treatment systems and natural water bodies.
A complete system of dosing, sampling and analysis has
been developed to make this new, non-radioactive tracer method
a viable field testing procedure. The field equipment and techniques,
and the calculations, will be discussed in this paper. The
analytical methods will be described in Paper II.
It is proposed that this new pair of tracers can be used
under operating conditions in a wastewater treatment plant employing
any common form of oxygen transfer equipment: coarse or fine
bubble diffused air, turbine or mechanical surface aeration.
Further, with the addition of a fluorescent dye to track the
tracer cloud, the method may be applied to large, surface water
bodies: rivers, lakes and impoundments. The technique is also
applicable to the measurement of transfer rate coefficients
for other gases, once appropriate relative transfer rates have
been established.
Two field tests of the stable krypton/lithium tracer method
have been completed to date. These tests have measured: the
-------
Hovis et al.
Page 3
aKTa of (1) a coarse bubble, diffused air system in an operating,
L
municipal activated sludge plant and (2) a fine bubble, dome
diffused air system in a manufacturer's, clean water, testing
tank. A third test is scheduled in an operating, municipal/paper
indusry, activated sludge plant, with mechanical surface aeration.
The precision of the rate coefficient measurements has
been determined. The precision of the Kr measurement by isotope
dillution mass spectrometry (+/- 0.05%, s.d.) and the precision
of the Li measurements by flame atomic absorption spectrophotometry
(+/- 3%, s.d.) are discussed in Paper II. Based on the clean
water test data, where dispersion was not a factor, the precision
of the krypton transfer rate coefficient (K a,.-Kr) was better
Xj I
than +/- 0.5%, s.d. The overall precision of the aKTa.,-Kr
Li I
measurement for the wastewater treatment plant was better than
+/- 3%, s.d. In calculating the oxygen transfer rate coefficient
(aK-a,) from these data, the precision proved to be limited
by the precision (+/- 5%, s.d.) of the literature value for
the krypton to oxygen transfer rate ratio (0.83 +/- 0.04, s.d.;
Tsivoglou, 1967). Refinement of this ratio is required to improve
the precision of the direct tracer measurement for oxygen.
-------
Hsu et al.
BUBBLES PRODUCED BY BREAKING WAVES
by
Y.-H. L. Hsu, Paul Anche Hwang, and Jin Wu
College of Marine Studies, University of Delaware
Newark, Delaware 19711
ABSTRACT
Bubbles produced by breaking waves are involved in many processes at the
air-water interface. In particular, bubbles are suggested to play an important
role in gas transfer across the interface, and in producing spray droplets.
However, this sequence of events — wave breaking, air entrainment, and
droplet production — has not been systematically studied. In the present
study, simultaneous measurements of breaking waves, bubbles, and spray at
different wind velocities were conducted.
The size and velocity of bubbles and spray were measured with optical
sensors. Their size spectra are found to be closely related. This provides
further evidence of the bubble bursting production mechanism.
Bubble population is found to increase drastically with wind stress; the
critical stress for bubble inception is also deduced. A simple model including
mechanism of entrainment is proposed to predict the bubble distribution as a
function of depth.
Time series of bubble flux shows intermittent property. The intermittency
is closely related to the frequency of wave breaking; the latter in turn is
related to wind stress and wave groupiness. Statistical and spectral analysis
on the bubble fluxes and waves are performed. Cross spectral analysis shows
higher coherence between wave and bubble flux at higher wind velocities (>_ 13
m/s). For lower wind velocities the bubble flux is intermittent due to only
occasional wave breaking.
-------
Hsu et al
Page 2
The population distribution per breaking interval (between breaking
waves) shows gradual change from relatively uniform distribution for strong
breaking to a skewed distribution for weak breaking. The spreading of the
distribution is also wider for lower wind velocities, indicating that wave
grouping plays a significant role in the breaking processes.
-------
Hunt
(Invited Keynote Paper)
Turbulent Structure and Mixing Near Gas Liquid Interfaces
J.C.R. Hunt
Department of Applied Mathematics and Theoretical Physics,
University of Cambridge, Silver Street, Cambridge CB3 9EW.
Summary
1. Turbulence structure
In boundary-layer flows of a fluid over another fluid much
denser (or more viscous) than itself, or over a solid surface, the
production and dissipation of turbulent energy occurs very close to
the surface where the mean velocity gradients are greatest;
turbulent energy is transported away from the surface by the
turbulence itself (.one can think of pairs of vortices inducing
such motions, or think in terms of 'bursting1 motions). Turbulence
can also be generated by many other means, by 'thermal convection,
by bubbling motions, by mechanical agitation or wave action on the
surface, or by breaking of internal waves. In most of these cases
the production and dissipation of turbulence is more evenly
distributed through the fluid. The differences and similarities
between these classes of turbulent flows will be discussed.
Now at the liquid side of a gas-liquid interface, in the
absence of breaking waves or a downward heat flux, no turbulent
energy is produced because the mean velocity gradient is smaller.
However the turbulence which is produced by the flow field (e.g.
by shear), well below the surface is distorted near the surface
because the velocity fluctuations normal to the surface are zero,
and the scale of the fluctuations decreases toward the surface.
Horizontal fluctuations within a thin viscous sublayer may be
-------
Hunt
Page 2
affected by the resistance to surface straining motions by
surface tension. The effect of wave and surface heat flux on
the turbulence structure will also be discussed.
2. Turbulent mixing
Mixing in a turbulent flow between a fluid and one or more
dispersed substances firstly depends on the scale and the intensity
of the turbulence, and secondly on the mean or slowly fluctuating
velocity gradients, (e.g. caused by shear or waves). The
mixing also must depend on the mean and fluctuating distribution
of the dispersed substances (e.g. the air dissolved in water).
The presence of a second phase (e.g. in the form of spray in air
or gas bubbles in water) can also have significant effect on the
turbulence and mixing. Depending on how interfaces affect the
turbulence and the mean flow, their effects on mixing vary.
These will be discussed Cas will the first part of this talk) in
largely qualitative terms but with some reference to mathematical
arguments and to experiments.
-------
Ilmberger et al,
Parametriration of air/lake gas exchange
J.Ilmberger, P.Liebner, B. Jaehne, K.H. Fischer
Institut fur Umueltphysik, Heidelberg University
W. Weiss. Bundesanstalt fur Zivilschutz - Messstelle
Freiburg
D. Imboden, U. Lemnin, EAWAG, Swiss Federal Instituts
of Technology, Duebendorf
Previous field data on air/lake gas exchange clearly show a strong
wind speed dependence, despite the large scatter of data. But
nevertheless these data are not appropriate to parametrise the
exchange rates, mainly due to the following reasons. Firstly, the
sampling periods were too long, so that the gas exchange rates
obtained were a mean value over various conditions. Secondly, and more
seriously, the parameters controlling gas exchange (windspeed,
stability, waves) were measured only imperfectly. The only parameter
known in some experiments was the windspeed, mostly only measured on
stations ashore.
Therefore in autumn 1982 we started an investigation on Lake Sempach,
Switzerland, in order to try a parametrization of air/lake gas
exchange. The gas exchange rate is measured with the Tritium/He3
method, furthermore oxygen profiles can be measured rapidly with a
jojosonde. In the middle of the lake a boje is anchored, measuring
automatically all important meteorological and limnological
parameters, and additionally the waves with three different optical
wave measuring systems.
-------
Ilmberger et al.
Page 2
We do not believe, that it will be possible to obtain a
parametrization of air/lake gas exchange during all the time with high
time resolution (a few days). Therefore the aim of our investigation
is twice: Firstly, to measure gas exchange under extreme conditions
(long period of low wind speeds, resp. short stormy period). Secondly,
to measure continuosly, all important parameters controlling the gas
exchange to compare them (espcially the waves) with wind tunnel
results and to test parametrizations predicted from lab results under
natural conditions.
-------
Iverfeldt and Lindguist
The Transfer of Mercury at the air/water interface
Ake Iverfeldt and Oliver Lindgvist
Department of Inorganic Chemistry, Chalmers University of
Technology and University of Goteborg, S-412 96 Goteborg,
Sweden.
The geochemical cycle of mercury have been of major interest
during many years. However, until recently only few basic data
governing the exchange of mercury at the air/water interface
has been known.
A large portion of the mercury in the atmosphere is elemental.
The Henry's law constant for elemental mercury (Sanemasa, 1975)
was used by Fogg and Fitzgerald (1979) to conclude that the
rainwater concentration of mercury must consist of other species
than Hg .
In the last few years, we have performed a series of determina-
tions of the Henry's law constants for different Hg species:
Compound
CH3HgCl
CH3HgCl
CH3HgCl
HgCl2
Hgci2
Hg(OH)2
Hg(OH)2
(Hg°
H =
: [HgX(g)] /[HgX(aq)] Temp
1.9-0.2-10"5
1 .6-0.2-10"5
0.9-
2.9-
1 .2-
3.2-
1.6-
0.3
TO'5
10'8
io-8
TO"6
-6
10
25°C
15
10
25
10
25
10
20
[Cl ]
0.7
1 .0
0.2-
0.2-
0.2-
0.2-
0.2-
0
M
ID'3
TO'3
TO'3
TO'3
-5
10'3
Ref
1
1
1
2
2
2
2
3 )
1. Iverfeldt and Lindqvist (1982); 2. Iverfeldt and Lindqvist (1980)
-------
2.
Iverfeldt and Lindquist
The table show that listed mercury compounds are effectively
scavenged from air by rain. From these constants and other
considerations, Brosset (1981) proposed a mercury cycle where
Hg° is emitted from ocean water to the atmosphere. He further
proposed that the elementary mercury is subjected to atmospheric
2 +
oxidation processes and transformed to water soluble Hg
species.
In a preliminary investigation, we have shown that Hg may be
oxidized by ozone. (To be published). The laboratory studies
show that the ozone induced oxidation is slow in air, but
that the rate is increased by several orders of magnitude having
a water phase present, cf „ cloud and raindrops in the atmosphere,.
To get a more complete picture of the mercury cycle we have
also performed in situ experiments using benthic chambers to
investigate sediment/water exchanges in an unpolluted coastal
water (Iverfeldt et al.,to be published). A rather large part
of the total mercury in water above the sediment was found to
be volatile, probably Hg°. The pore water contained relatively
less volatile mercury. We do not know, if the volatile Hg is
produced only in the sediment or/and at the sediment/water
interface by a change in redoxpotential or by -bacterial
degradation (cf. Spangler et al, 1973).
The sediment/water interface could be compared to the interface
between oxic surface water and anoxic subsurface water in a
permanent anoxic fiord (Framvaren, Norway). A very active
bacterial layer is found in the redox boundary (Skei, 1981).
-------
Iverfeldt and Lindquist 3.
Extremely high amounts of volatile mercury was also found just
above this boundary. (To be published).
Knowing that the ocean is covered with a surface layer with
high bacterial activity, a hypothesis may be stated, that
Hg emitted from the ocean to the atmosphere is partly
produced in the surface microlayer.
REFERENCES
Brosset C.(1981) Water, Air and Soil Pollut. 16, 253-255.
Fogg T.R. and Fitzgerald W.F.(1979) J. Geophys. Res. §4, 6987-6989
Iverfeldt A. and Lindqvist O.(1982) Atm.Environ. 1§ (in print).
Iverfeldt A. and Lindqvist O.(1980) The Coal, Health, Environment
Project, Report no 41 5 ( in Swedish with a summary in English),
The Swedish State Power Board, S-162 87 Vallingby, Sweden.
Iverfeldt A.,Hall P-,Rutgers van der Loeff M.,Westerlund S.,
Sundby B. and Andersson L.( To be published )
Sanemasa I. (1975) Bull.Chem.Soc.Jpn. 4J, 1795-1798.
Skei J.(1981) NIVA-Report F-80400(ISBN 82-577-0447-4) Norsk
institutt for vannforskning, Oslo, Norway.
Spangler W.J.,Spigarelli J.L.,Rose J.M. and Miller H.M.(1973)
Science 180, 192-193.
-------
Jaehne
Optical water waves measuring techniques
B. Jaehne, Institut fuer Umweltphysik
Heidelberg University, Western Germany
We focused our interest on optical wave measuring techniques
because of the importances of capillary waves for gas exchange
and knowing the difficulties to measure them with wire gauges.
Any optical method obvisiously has the advantage not to be in contact
with the water surface and therefore not to interact with the waves.
We developed a refraction technique similary to the ones already used
by other workers. The system measures the slope of the waves. A
He-Ne-laser beam pierces the water surface vertically from above. The
optical reciever consists of a Fresnel lens and a focusing screen, one
focal length distant, so that the displacement on the screen only
depends on the slope and not the height of the waves. Finally the spot
on the screen is imaged on a dual axis position sensitive photo diode.
The resulting electric signal is slightly nonlinear due to the optical
geometry and non linearities of the photo diode. The cutoff frequency
of the detector together with the electronics is about 100 kHz, Thus
the frequency response of the system is limited only by the diameter
of the laser beam. Tests with different diameters show, that all wave
frecencies occuring can be measured without any damping.
The fast response of the detector enables us to measure correlations
with only one wave measuring unit in the following way: The laser beam
oscillates horizontally (0.5 to 30 mm) over the water surface at the
-------
Jaehne
Page 2
sampling frequency (about 1 kHz) of the time series. Two time series,
0.5 to 30 mm distant, are put on the computer simultaneously. Phase
velocities of capillary waves are thus determined.
Finally, the possiblities of visualisation techniques based on the
same optical geometry as the spot method with the laser are discussed.
The first step done in this direction are photographs of a 20 cm
diameter area showing the slopes of waves indicated by different
colours. With this method it will perhaps be possible to calculate two
dimensional wave number spectras.
-------
Jaehne et al..
Wind/wave-tunnel experiments nn + h» Schmidtnumber- and
wave field dependence of air/water gas exchange
B. Jaehne, W. Huber, J, Ilmberger, T. Wais
Institut fur Umweltphysik, Heidelberg University
Western Germany
Gas exchange experiments have been carried out in order to study the
influence of parameters controlling gas exchange without breaking
waves: friction velocity u,, Schmidtnumber Sc=v/D for the gas
studied, and the wave field.
We use 3 different wind/wave tunnels for our experiments: Two circular
ones of different sizes and the the large wind/wave facility of
IsM.S.T Marseille, where the experiments have been carried out in
cooperation with L. Merlivat and L. Memery from CEN Saclay. The
circular tunnels have the advantage of a homogenous wave field of
unlimited fetch in contrast to the short fetch waves in all linear
tunnels.
Different tracers are used in order to study the dependence of the
exchange rate on the Schmidtnumber: C02 and heat in the small
circular tunnel, C02, He, CH«, Kr and Xe in the large circular
facility and Rn and He in the I.M.S.T. tunnel. Wave slope spectras are
obtained with an optical refraction method using a He-Ne-Laser and a
position sensitive photodiode.
With a smooth water surface the experimental results agree well with
theories treating the water surface as a rigid wall: The Sc'273
dependence and correct absolute values of the exchange rate are
obtained. When waves occure a strong increase in the transfer velocity
is observed as well as a change of the Schmidtnumber dependence to
Sc"1'2.
-------
Jaehne et al.
Page 2
Both results cannot be explained with existing theories (surface increase
and periodic dilation of the viscous sublayer by waves). The increase
of the gas exchange rates through the waves is different in the 3
tunnels due to their different wave fields. Several wave parameters
are discussed in order to parameterize the influence of waves on the
exchange process. Moreover, the windtunnel results are compared with
field data.
-------
Jirka and Brutsaert
MEASUREMENTS OF WIND EFFECTS ON WATER-SIDE
CONTROLLED GAS EXCHANGE IN RIVERINE SYSTEMS
by
Gerhard H. Jirka and Wilfried Brutsaert
School of Civil and Environmental Engineering
Cornell University
Ithaca, New York 14853
ABSTRACT
The traditional approach toward the analysis of gas transfer at the sur-
face of streams or rivers has been limited to consideration of turbulence gen-
erated by bottom shear in the flowing water. The effect of wind has been sim-
ply neglected. However, for wider open rivers with moderate slopes and more
tranquil flows, including inflow into lakes or reservoirs and approaches to
estuarine conditions the wind effect must be expected to become increasingly
important and a transition to purely wind-controlled conditions (as considered
in oceanographic work) will take place.
A simple criterion on the relative importance of streamflow versus wind
generated gas exchange is developed on the basis of the respective near-sur-
face turbulent energy dissipation rates e in the water. Open channel experi-
mental data by Nakagawa et al. (1975) are used for e under pure streamflow
conditions, while e for wind conditions is computed on the basis of similarity
in the wall layer of the wind driven turbulent boundary layer in the water
with estimates of the roughness height from Kondo (1975).
Data from two riverine gas exchange experiments with concurrent detailed
wind measurements are analysed. Correlations between wind shear velocities
and observed exchange velocities (KL) appear to be consistent with the dissi-
pation based criterion. The data also demonstrate the strong micro-meteoro-
logical sheltering effect of stream banks and vegetation on gas exchange in
small streams.
-------
Johnson
PREDICTION OF DISSOLVED GAS TRANSFER IN SPILLWAY
AND OUTLET WORKS STILLING BASIN FLOWS
by
Perry L. Johnson, Hydraulic Engineer, U.S. Bureau of Reclamation
An empirical model is presented that predicts the oxygen and nitrogen
transfer to and from the flows through hydraulic structure stilling basins.
The model may be applied to a wide variety of structures ranging from open
chute hydraulic jump basins to flip buckets and plunge pools and from hollow-
jet valve to fixed cone valve to slide gate controlled outlet works basins.
The model may be used to evaluate both the structure's reaeration potential
and the structure's potential for nitrogen supersaturation development.
The present model is a refined and expanded version of a model first developed
by the author in the mid-70's. The present model is better suited to handle
low Froude number stilling basins and is analytically more direct. The
model is based on data collected at over 24 different field structures.
Parameters considered include dissolved gas concentrations in the flow enter-
ing the structure; water temperature; barometric pressure; the velocity, cross
sectional shape, and orientation of the flow entering the stilling basin;
stilling basin length, width, depth, and shape; and tailwater depth. Included
is an example application.
-------
Jones and Hannan
A New Way to Study CO /Water Equilibria
-------
Kerman I
DISTRIBUTION OF BUBBLES NEAR THE OCEAN SURFACE
Bryan R. Kerman
Boundary Layer Research Division
Atmospheric Environment Service, Canada
Extended Abstract
Crowther has recently offered an interesting hypothesis that the
number of bubbles near the air-sea interface is only a function of the
rate of work of the atmosphere on the surface, the surface tension and
the bubble radius. Interestly, no consideration was given to the dynam-
ical significance of g, the acceleration due to gravity, which according
to Phillip's model of an equilibrium wind-wave field, including breaking
waves and extensive bubbling, uniquely defines the wave height spectrum..
A simple model, not explicitly derived in terms of the statistical
transport equations utilized by Garrettson and Crowther, is formulated
for the generation and 'spectral' transport of bubble mass. A subrange
in which there is no mass convergence due to coalescence or dissolution
is considered. From mass continuity arguments, the change of radius due
to pressure, and the hydrostatic balance, it is readily deduced that the
mass (and volume) flux per radius increment is both distributed as a neg-
ative exponential with depth and as r~3 with radial size. However the
model of Crowther, as demonstated, is not unique but is only one of a set
of possible similarity solutions.
Based on arguments for the radial forces on a cavity, there exists
an outer scale, for bubble production by shattering comparable to the
wavelength associated with a minimum phase speed of gravity-capillary
waves. A revised dimensional analysis including, g, and this outer
shattering scale, leads to a result, identical to Crowther's indicating
the similarity function, associated with Crowther's model, is a constant
in the problem variables.
-------
- 2 - Kerman I
An analysis of various data sets for bubble densities and distri-
butions, or deduced properties from acoustic resonance, establishes that
the wind dependence, vertical distribution, spectral shape and tenta-
tively, the influence of surfactants agree with the model's predictions.
However, for agreement in order of magnitude, it is necessary to consider
the relatively sparse nature of active bubble production in breaking
waves compared to the total surface area, as well as the relatively
sparse nature of large shatterable cavities within a whitecap. Intuitive
arguments for these characteristics result in agreement within a factor
of 10, but are considered weak due to lack of detailed experimentation on
the structure and life cycle of breaking waves.
-------
,S Kerman II
A MODEL OF INTERFACIAL GAS TRANSFER FOR A WELL-ROUGHENED SEA
Bryan R. Kerman
Boundary Layer Research Division
Atmospheric Environment Service, Canada
Extended Abstract
Deacon has recently proposed a model for wind-driven gas transfer
across a sea surface if the airflow is aerodynamically smooth. In this
case of transitional or fully rough flow, it is well-known, at least
from laboratory measurements, that the gas flux significantly exceeds
that predicted from an extension of smooth wall considerations. At the
same time, by analogy or direct comparison with the limited extent of
bubbling in breaking waves or peaked small waves, the flow separation
associated with rough wall flow is hypothesized to occur over a small
fraction of the total area. A key question is whether moderate local
turbulence near sparse regions of flow separation can transfer gas more
efficiently than weak, but global, turbulence for smooth wall flux.
A model is presented for the flux of a non-reactive, low solubility,
low diffusivity gas, like 02» through this air-sea interface in the
presence of patches of waves over which the flow is separating. The
development is based in the work of Yaglom and Kader for rough wall trans-
fer of heat and mass, but is applied here to a compliant rather than
solid surface. In order to estimate the surface (roughness) Reynolds
number, it is necessary to quantify the rms height of the roughness
elements. An extension of the Charnock relationship for surface roughness
is provided which accounts for the contributions from surface root mean
square elevation and slope for waves whose phase speed exceeds the minimum
phase speed of gravity-capillary waves. The areal extent of the transfer
is modelled in a manner similar to Cardone and Wu, but is normalized in
terms of the threshold stress for flow separation.
• • • • • C.
-------
p
Kerman II
Comparison of the model with various data sets, including the .stress
dependence of Charnock's constant, the areal coverage of whitecaps and
the gas flux in laboratory and recently reported field experiments, con-
firms the genera] validity of the formulation. However, in all cases
there is considerable experimental scatter or limited representativeness
so that it is uncertain whether the model is correct in detail or only in
its general trends.
-------
Khudenko
Stationary and Self-Propelled
Weir Aerators
Boris M. Khudenko, Associate Professor
Georgia Institute of Technology
Atlanta, GA 30332
Several weir types, sharp-crested, broad-crested, tooth-shaped, and tooth-
shaped orifice have experimentally been tested in a hydraulic flume. Sizes of
weirs, the flume length and depth, and the water flow rate have been varied over
a wide range. The maximum capacity of the test model is 10,000 m /d. The
efficiency of oxygen transfer has been related to the geometric and hydraulic
characteristics of the studied systems. The efficiency of oxygen transfer by
the use of weirs reaches 3 kg/kW-L.
A self-propelled modification of weir aerators has been developed. This
aerator consists of a weir box with a weir or orifice, and a propeller type
pump located on floats. The floating structure is connected by means of a
hinged arm to a central pile. When the pump is activated a jet of water is
discharged through the weir or orifice. The jet aerates and mixes the water and
propels the floating structure around the central pile. Rotation o.f the floating
structure around the pile increases the service area of the aerator. The major
parameters of this aerator type, oxygen transfer, velocity of rotation, and
mixing capacity have been evaluated.
Weir aerators can be used for water and wastewater treatment, aeration of
stagnant and stratified lakes, rivers, etc. The advantages of stationary weirs
are best utilized in high rate processes such as activated sludge process,
aerobic stabilization of sludge, postaeration, and other. An important feature
of self-propelled aerators is their ability to mix large volumes of water.
Such aerators can be used for biological ponds, lakes, etc.
-------
Kitaigorodskii and Donelam
(Invited Keynote Paper)
ABSTRACT
The Structure of Turbulence on Both
Sides of the Air-Water Interface
S. A. Kitaigorodskii
Department of Earth,and Planetary Sciences
The Johns Hopkins University
Baltimore,Maryland 21218
M. A. Donelan
National Water Research Institute
Canada Centre for Inland Waters
Burlington, Ontario
In this paper we discuss the structure of turbulence in the
vicinity of an air-water interface as it relates to gas transfer.
Because the molecular diffusion coefficient of gas in the liquid phase
is much lower than the gaseous coefficient, the transfer of gases
across the air-water interface is dominated by the resistance of the
boundary layer in the water. Even for hydrodynamically rough flow this
remains true because of the large Prandtl number in the water. That
is, the erosion of the viscous boundary layer implied by rough flow
does not necessarily mean any significant commensurate thinning of the
diffusive boundary layer.
However, in the presence of active wind-wave breaking the analogy
between transfer across the air-water interface and heat transfer at
solid srufaces is no longer valid and the structure of the turbulent
boundary layers on both sides of the interface becomes important.
Therefore in the first part of this paper we discuss the processes
responsible for the variability of the molecular diffusion sub-layer in
the presence of shear-generated turbulence near solid roughness
elements. In the second part of the paper we turn our attention to gas
transfer processes when there is active wind-wave breaking.
Frequent use is made of data from laboratories and natural
air-water interfaces to illustrate the variability of the structure of
turbulence as it pertains to gas transfer.
-------
Krause et al,
ABSTRACT
GAS TRANSFER RATE COEFFICIENT MEASUREMENT
BY A STABLE ISOTOPE KRYPTON TECHNIQUE.
II. ANALYTICAL METHODS
D. Krause, Jr. , B.B. Benson ,
2 2
J.S. Hovis , J.J. McKeown ,
3 3
R.P. Fisher , and M.N. Stryker
1) Department of Physics, Amherst College, Amherst,
Massachusetts 01002
2) National Council of the Paper Industry for Air and Stream
Improvement, Inc., Northeast Regional Center, Tufts
University, Medford, Massachusetts 02155
3) National Council of the Paper Industry for Air and
Stream Improvement, Inc., Southern Regional Center,
P.O. Box 14483, Gainesville, Florida 32604
-------
Krause et al.
Page 2
We have developed equipment and techniques for measuring
the transfer rate of krypton gas in large bodies of water or
wastewater treatment plant mixed liquor. The procedure is criti-
cally dependent on the ability to measure precisely the small
krypton content of liquid samples, which is accomplished by
a somewhat unusual isotope dilution technique. Because the
test is conducted under dynamic conditions, it is equally critical
to account for gas dispersion by using a lithium salt as '.a .conserva-
tive tracer added with the krypton.
78
The isotope Kr is used as the isotope dilution spike because
78
Kr constitutes only 0.35% of normal krypton gas,, The new
technique involves spiking the liquid sample with an amount
78
of nearly pure Kr which is approximately equal to the amount
to total krypton in the sample, and then measuring the abundance
ratio of 78Kr to "total krypton" (8°Kr, 82Kr, 83Kr, 84Kr, 86Kr)
with a specially designed dual collector mass spectrometer. This
78
ratio, combined with the known amount of Kr spike, yields
a very precise measure of the krypton content.
Five to thirty STP liters of krypton dissolved in twenty
to forty liters of lithium chloride brine (40% w/w) are quickly
injected into the system to be studied from two equilibrators,
each with a volume of ^85 liters and operable at pressures up
to 7 atm. Subsequent to this dump, samples of the liquid are
collected at appropriate time in -^250 cc glass bottles and sealed
with Poly-Seal caps. Exact liquid content is determined by differ-
ential weighing. An unsealing device has been designed and construct-
ed which allows the sample bottle to be opened inside a thoroughly
-------
Krause et al.
Page 3
78
evacuated chamber. The Kr spike is introduced into the chamber
before opening the bottle. The liquid is degased by violent
stirring, and the evolved gases are dried in dry ice - acetone
and liquid nitrogen traps before they are cycled over hot (950C)
titanium to remove all but the noble gases. The gas is then
put through a process gas chromatograph using helium as carrier
gas. The sample stream gas is directed into a mechanical vacuum
pump except during the time that the krypton is being eluted,
when the flow is switched to pass the gas over a molecular sieve
trap operated at 77K, which collects the krypton. After the
helium is pumped away, the sieve is warmed and the released krypton
is collected for the mass spectrometric analysis. Because the
spiking material is added at the time that the bottle is opened,
total quantitative recovery is not necessary because it is only
the isotopic ratio which must be measured.
The concervative tracer (Li) was determined by flame atomic
absorption spectrophotometry (AAS). Experimental results have
shown than among the standard procedures for sample preparation
and AAS analysis of lithium in mixed liquor, only one protocol
gives satisfactory results.
The precision and accuracy of the gas and conservative
tracer measurements are discussed.
-------
Kuo and Pilotte
Estimation of Volatilization of Toxics for Multimedia Modeling
Jan-Tai Kuo, Visiting Associate Professor, Civil
Engineering Dept., National Taiwan University, Taipei,
Taiwan, Republic of China and James 0. Pilotte. General
Software Corporation, Landover, Maryland, U.S.A.
Abstract
Multimedia modeling for estimating the concentration of
toxic substances considers the transport, transformation and
reactions of chemicals in air, water and soil, and the rate of
chemical exchange between any two of these three geospheres. The
methods used to estimate the changes in concentrations due to the
volatilization processes between water and air as well as between
soil and air can be of critical importance in a multimedia model.
Many multimedia models utilize simplified approaches (e.g.
fugacity, partitioning) in which simple and easy-to-use governing
equations are used to estimate the exchange rates of
volatilization processes (e.g. Henry's law or two-film theory for
the volatilization between water and air and Pick's law of
diffusion between soil and water).
This paper discusses and compares the governing equations
for modeling volatilization processes in existing toxic substance
multimedia models such as UTM (Unified Transport Model developed
by Oak Ridge National Laboratory), ENPART (Environmental
Partitioning Model developed by U.S. Environmental Protection
Agency and General Software Corporation), TOX-SCREEN (a
multimedia screening model developed by Oak Ridge National
Laboratory) and others (e.g. Mackey's and Neely's Partitioning
Models).
Chemical, physical and environmental factors which affect
the volatilization/evaporation processes of toxic substances are
discussed. Appropriate models for estimating volatilization/
evaporation of toxics at the water surface between water and air
and the interface (containing water, soil and air) between soil
and air are suggested and their complexity, accuracy and useful-
ness are discussed.
-------
Ledwell
Abstract for a paper to be presented to the International Symposium on
Gas Transfer at Water Surfaces at Cornell University, June 13-15, 1983:
The Schmidt Number Dependence of the Gas Transfer Coefficient
James Robert Ledwell
Lamont-Doherty Geological Observatory
Palisades NY 10964
Transfer coefficients for NpO, CH^, and He at the air water interface
,vere measured in an l8 meter wind wave tank at moderate wind speeds. The
transfer coefficient was found to vary with D , with the exponent 'a1 equal
to 0.4? + O.Ik (90% confidence interval). The diffusion coefficients used for
N_0, CH^, and He were 1.84, 2.32, and 6.12 X 10~5 cm2/s at 23.5°C. This
result supports models in which the transfer coefficient is proportional
to the square root of the Schmidt number (the ratio of the diffusion
coefficient to the kinematic viscosity of water). The result is contrasted
with that for transfer coefficients at solid boundaries, where the analog
of 'a' assumes a value of 2/3. The difference is attributed to the
boundary constraints which allow tangential convergence of fluid at the
air-water interface, but not at the solid boundary.
-------
Lincoff and Gossett
ABSTRACT
The Determination of Henry's Constants of Volatile Organics
by Equilibrium Partitioning in Closed Systems
by
Andrew H. Lincoff
Graduate Research. Assistant
Department of Environmental Engineering
Cornell "University
and
James J4, Gossett
Associate Professor
Department of Environmental Engineering
Cornell University
Ithaca, New York 14853
An accurate knowledge of Henry's constant is necessary for modeling gas
transfer in a variety of situations. Unfortunately, great accuracy is diffi-
cult to achieve using straightforward measurement of gas and aqueous phase con-
centrations from equilibrated, closed systems. Many of the compounds of current
interest are so sparingly soluble — and of sufficient volatility — that it is
difficult to construct accurate calibration curves with which to assay equilib-
rium concentrations. This is particularly true of aqueous phase assays. As a
result, current methods for determining Henry's constants usually involve the
use of a batch, diffused-air stripping tower, wherein it must be assumed that
equilibrium is achieved between the gas which exits the tower and the aqueous
phase through which it has bubbled (_e_.g_, , see Mackay et_ al, , Envir. Sci. S
Technol. 13, -333, 11979]}. In our paper, we demonstrate that for many compounds
of high volatility frequently found in contaminated surface and ground waters,
mass transfer limitations exist which invalidate the equilibrium assumption of
the batch air stripping method.
*
To whom correspondence should be addressed.
-------
Lincoff and Gossett
Page 2
ABSTRACT; Lincoff, A. a., and J. K, Gossett, "The Determination of Henry's
Constants of Volatile Organics by Equilibrium Partitioning in Closed
Systems."
A novel technique is presented •— Equilibrium Partitioning in Closed
Systems — which allows the calculation of Henry's constant from the mere
ratio of gas-phase concentrations resulting after equilibration of two closed
systems which have been prepared differently. The technique is extremely
simple and is free of mass transfer limitations which reduce the utility of
other techniques. Since this new method uses only the ratio between gas con-
centrations of two equilibrated systems, knowledge of the actual gas phase con-
centrations is unnecessary. This eliminates a major problem associated with
the research of volatile compounds in dilute solutions namely, the preparation
of calibration curves. Henry's constants can be determined with a mean co-
efficient of variation less than 5%, solely using peak height data obtained
through GC analyses of headspace samples from two equilibrated systems.
Knowledge of the actual mass of volatile substance added to the two systems is
also unnecessary.
Using both our proposed technique — and that of Mackay et al. •*•— Henry's
constants for five compounds were measured over a range of temperature CIO -
30 C) and ionic strength (0 - 1 N KC1) typical of natural bodies of water. .Mix-
tures of the five compounds in water were also investigated. The compounds,
tetrachloroethylene, 1,1,1-trichloroethane, trichloroethylene, chloroform, and
methylene chloride, encompass a wide range of volatility and are all EPA prior-
ity pollutants. Results demonstrate that for the more volatile of the compounds,
the batch air stripping method significantly underestimates Henry's constants
due to apparent mass transfer limitations preventing equilibration of exit
bubbles.
-------
Ljubisavljevic
Carbondioxid Pesorbtion from the Activated.
Sludge in the Waste Water Treatment Plants
bys Dejan Ljubisayljevid
Faculty of Civil Engineering, Belgrade, Yugoslavia
Abstracts
It is necessary to study more in detail the desorbtion of
carbondioxid from the activated sludge, in waste water
treatment plants where biological nitrification occurs
(sensitive on the decrease of the pH) , and where pure
oxygen is applied as an oxidant.
Experimental results obtained from the laboratory models of
the surface and diffused air aerator are presented in this
article. Desorbtion of carbondioxid was measured with the
help of the pH-meter. Experiments were done both with tap
water only, and with tap water and surface active agents.
Results from the models were compared with the existing
gas transfer theories, and the validity of certain theories
were estimated.
From the carbondioxid desorbtion point of view, the conside-
rations on the results obtained from the field measurements
made at the pilot waste water treatment plant in Buesnaw
(Sttutgart, Federal Republic of Germany) are presented. The
field measurements on the mentioned plant in which biological
-------
Ljubisavljevic
- 2 -
nitrification - denitrification occurs were performed by
the author in June 1982.
Among else, alkalinity and pH were analysed on the end points
of the nitrification and of the denitrification parts.
(Different from the carbon oxidation, besides the increase
of the carbondioxid content, nitrification decreases the
alkalinity of the activated sludge).
On the base of the theoretical and experimental considerations
the conclusions were drawen about the efficiency of certain
aeration systems on the desorbtion of the carbondioxid from
the activated sludge.
Some recommendations are propoused about the choise of the
aeration systems for the waste water treatment plants, were
special care is to be taken of the problem of the carbondioxid
desorbtion from the activated sludge.
-------
Maclntyre
CURRENT FLUCTUATIONS IN THE SURFACE WATERS OF SMALL LAKES
Sally Maclntyre, Department of Mechanical and Environmental
Engineering and Marine Science Institute, University of
California, Santa Barbara, California 93106
The transfer of momentum from the wind to the surface of a lake
creates fluctuations in velocity which are an expression of turbulent
eddies, surface and internal waves. To know the scaling from the wind
stress to the stress at the surface of a lake is important for budgets
of energy transfer and models of mixed layer dynamics. An environment
in which these processes have not been examined is shallow, naturally
eutrophic lakes with small fetches. I used an omnidirectional warm-
bead thermistor-flowmeter to measure fluctuations in current speeds
throughout the upper mixed layer of three shallow lakes in equatorial
Africa with differing exposures to wind. The flow sensor resolved
speeds of less than 0.1 cm/sec and frequencies of ca. 5 Hz. The first
four statistical moments, probability density functions, autocorrelation
functions, and energy spectra have been computed from the current records,
the mean and rms speeds have been compared with wind speeds measured two
meters above the lakes' surfaces, and Richardson's numbers have been
calculated.
Current speeds within 1.5 cm of the surface fluctuated rapidly from
0.3 to 16 cm/sec. The mean speeds were ca. 5 cm/sec and the turbulent
intensities (ratio of rms speed to mean speed) were greater than 0.4.
The data records showed a graded shift from high frequency, high amplitude
fluctuations to lower frequency (ca. 2.3 to 0.3 Hz), smaller amplitude
fluctuations from the surface to 10 cm depth. This shift occurred in the
absence of thermal stratification. At 10 cm the turbulent intensity was
still high (0.25). In thermally stratified water, waves with a frequency
of 0.01 Hz were observed within 30cm of the surface. These data illustrate
the extreme activity and variety of responses in low energy lacustrine
environments.
-------
Mackay
(Invited Keynote Paper)
Abstract for
Physical-Chemical Phenomena and
Molecular Properties (Keynote Paper Topic 1)
D. Mackay, University of Toronto
The transfer characteristics of solutes between water and air are con-
trolled by the solute's equilibrium and transport properties.
The equilibrium properties of aqueous solubility, vapor pressure and
Henry's law constant determine the relative concentrations at equilibrium
and hence the direction of transfer in any given situation. Partition co-
efficients to solid or liquid materials in both air and water phases near
or at the interface and interactions with dissolved material may also play
an important role. Methods of measuring correlating and estimating these
properties for gaseous liquid and solid solutes are reviewed and discussed.
Mass transfer rates are conventionally expressed using a "two resistance"
approach in which mass transfer coefficients are invoked for each phase.
These coefficients are primarily determined by the local hydrodynamic regime
but are also influenced by the solute's diffusivity. It is suggested that
air and water phase Schmidt Numbers currently provide the best characterization.
of solute properties and of temperature, when raised to an appropriate power
reflecting the exchange conditions. Methods of obtaining and testing such
data are discussed.
In view of the large number of solutes of potential interest, it is
suggested that a strong case can be made for improving the capability of
correlating and estimating these physical chemical properties for compounds
which are members of similar chemical classes.
-------
Martinelli and Liss
THE EFFECT OF INSOLUBLE AND SOLUBLE MONOLAYERS ON GAS
EXCHANGE ACROSS AN AIR-WATER INTERFACE
F.N. Martinelli and P.S. Liss
School of Environmental Sciences,
University of East Anglia, Norwich, U.K.
ABSTRACT
Two series of laboratory tank experiments are described in this paper.
In the first series the simultaneous transfer of H2O, SO2, C»2 and O2 was
monitored across clean and monolayer covered water surfaces under conditions
of low hydrodynamic turbulence. It is known that under certain conditions
insoluble monolayers are able to present a specific or barrier resistance to
gas transfer, and results presented here for long chain fatty acids and
alcohols are in reasonable agreement with previous work for the transfer of
^O and SO-, which is under gas phase control in these experiments. For 02
and CO-, which are under liquid phase control, the specific resistance of
the monolayer cannot be seen against the intrinsically high resistance of
the water itself.
Soluble monolayers do not fulfil the stringent requirements necessary to
present a direct resistance to gas transfer, but can provide an indirect
effect under conditions of high hydrodynamic turbulence. In the second series
of experiments, the effect of the surfactant Manoxol O.T. (sodium dioctyl
sulphosuccinate) was measured on the transfer of 02 under high stirring regimes.
The results show that surfactant concentrations as low as O.O1 mg 1 are
capable of significantly reducing 02 transfer velocities, and that reduction of
nearly 6O% can be obtained when the concentration of the surfactant exceeds
1O mg 1
The ability of sea surface microlayer material, collected using a Garrett
screen, to retard gas transfer was tested in both series of tank experiments *
In the quiescent experiments no reduction was evident but significant reductions
were noted in the turbulent experiments and the consequences of this finding
for air-sea gas transfer are discussed.
-------
Matteson
ABSORPTION OF GASES AT CONDENSING
AND EVAPORATING WATER SURFACES
Michael J. Matteson, Georgia Institute of Technology
ABSTRACT
Observations of water vapor condensing on both large droplets (2 mm)
and NaCl nuclei (1 pm) in the presence of a trace gas Y, indicate an en-
hancement in the concentration of Y in the liquid phase during condensation
and a depletion of Y during evaporation. A model is proposed wherein
homogeneous clusters of gas molecules of the type Y- CH90) formed in the
£ XI
gas phase may explain these departures from expected mass transfer be-
havior. When Y=S02:
[SO-1 + n[H-01 * [S07-n(H70)l
" 6 " S> *• " 6
[S02-n(H20)] * [S02-n(H2OJ]£
[S02-n(H20)]£ * [S02]z + n[H20]£
[S02-nCH20)J£ * H
During condensation or evaporation, a water vapor pressure gradient
•establishes a gradient in the gas phase hydrate. Mass transfer coeffi-
cients, equilibrium constants and heats of solution were calculated for
the above mechanism. Further tests with oxygen and N02 reveal similar
patterns during the condensation- evaporation cycle, and a general model is
proposed for absorption water soluble gases.
-------
McCready and Hanratty
A COMPARISON OF TURBULENT MASS TRANSFER AT
GAS-LIQUID AND SOLID-LIQUID INTERFACES
by
Mark J. McCready and Thomas J. Hanratty
Department of 'Chemical Engineering
University of Illinois
Urbana, Illinois 61801
For a number of years we have been studying the details of a turbulent
flow close to a solid boundary and have been using this information to develop
a theory for turbulent mass transfer. This work has led to a picture of tur-
bulent mass transfer to a solid boundary which is radically different from
theories currently available in the literature. Only very low frequency velo-
city fluctuations and, therefore, only a small fraction of the turbulent ener-
gy is effective in transferring mass. The fraction of this energy decreases
with increasing Schmidt number. In recent years we have been applying in-
sights gained in this work to the problem of defining the mechanism of tur-
bulent mass transfer at a gas-liquid boundary. This paper summarizes our
progress.
The system in which our experiments are conducted is a 2.54cm x 30.48cm x 9m
enclosed rectangular channel. Liquid flows along the bottom of the channel
and air flows concurrently. The rate of oxygen absorption is determined by
measuring the change of the oxygen concentration of the liquid film at dif-
ferent distances downstream. The wave properties are measured by a parallel
wire conductance probe. The shear stress fluctuations at the wall are
measured by a flush mounted thermal probe. The film varies in height from .05
to .55cm. Air velocities from 4 m/s to 14 m/s have been used. The viscosity
of the liquid can be varied using water-glycerine solutions.
-------
MoCready and Hanratty
O
Re
Sc = 440
6900
ILLINOIS
11500
25000
35000
TOULOUSE
45000
55000
00° o
SOLID WALL
10
m
Figure 1.
-------
McCready and Hanratty
Measurements of the mass transfer coefficient obtained with water are
shown in Fig. 1. Here the friction velocity is defined in terms of the
interfacial stress, vi* = (t^/P>i)1 ^2 and the height of the liquid layer has
been made dimensionless using the friction velocity and the kinematic viscos-
ity, m+ = m v-*/v. It is noted that for m+ > 45 the dimensionless mass trans-
1/9 A
fer coefficient is approximately constant, k Sc1' /v-_ =0.1. For comparison,
measurements obtained in a 10cm x 20cm x 6m channel at the University of Tou-
louse (Tsacoyannis, 1976) with flowing liquid layers much thicker than ours
(1cm to 3.5cm) are also shown. It is of considerable interest that such good
agreement is obtained in widely different systems, provided the results are
plotted in the manner shown in Fig. 1. These results suggest that the velo-
city fluctuations controlling mass transfer are independent of the liquid
depth for thick films and are caused by interfacial shear. Three mechanisms
are being explored as being responsible for these velocity fluctuations close
to an interface: turbulence generated at the interface, turbulent velocity
fluctuations transmitted by gas flow shear stress fluctuations, and waves.
From extensive experiments we have found that turbulent mass transfer at
a solid boundary is given by the equation
k/v * = 0.8 Sc~'704 .
w
It is noted that this relation is similar to what is found for a gas-liquid
interface in that the mass transfer coefficient is related to flow variables
through the friction velocity. This suggests that the shear at a gas-liquid
interface is producing turbulence by processes similar to what occur at a
solid-liquid interface. Support for this is obtained from the measurements of
Suzanne (1977) which indicate a maximum in the turbulence intensity in the
liquid near a gas-liquid interface.
-------
McCready and Hanratty
However, as shown in Fig. 1, the magnitude of the mass transfer coeffi-
cient is smaller for the solid-liquid interface. A possible explanation for
this can be obtained from an analysis of the relation between the concen-
tration fluctuations and velocity fluctuations close to a boundary. This
relationship for a clean gas-liquid interface, where the normal velocity var-
ies linearly with distance from the interface, is quite different from the
case of a solid-liquid boundary where the normal velocity varies quadratically
in y. For a gas-liquid interface the concentration boundary layer does not
act as a filter; this means that mass transfer at a clean gas-liquid interface
is being affected by velocity fluctuations of all frequencies, and not just by
very low frequencies, as is true for a solid-liquid interface.
The sharp drop off in k Sc ' /v^ noted for m < 40 can be explained if
for small film heights, turbulence generation at the interface can be inhibit-
ed by the nearness of the solid boundary. In this region the inflexible
boundary distorts the flow field and influences the scales of the turbulence
1/2 *
produced at the interface. Consequently, k Sc /v. is a function of the
gross flow parameters film height and gas Reynolds number for small m .
For very thin films measurements from our wall shear stress probes indi-
cate that fluctuations in the gas flow are causing fluctuations in the shear
stress transmitted to the liquid film by the gas flow. This provides another
source of velocity fluctuations in the liquid near the interface. By using
measurements on turbulent shear stress fluctuations of air at a solid bound-
ary, we have carried out an analysis to estimate their influence in gas-liquid
mass transfer. We find that the velocities scale with v. . This suggests
that these fluctuations should give a contribution to k which increases
linearly with v .
-------
McCready and Hanratty
5
A tentative explanation of the results in Fig. 1 is that mass transfer is
controlled by an additive effect of gas shear stress fluctuations being trans-
mitted to the liquid and of turbulence being created in the liquid by the
shear field interface. As m+ increases the creation of turbulence in the
liquid becomes relatively more important.
It is well known that surface waves on freely falling films can cause an
increase of mass transfer over predictions for laminar flow. A simple analy-
sis of the influence of waves on velocity fluctuations in gas-liquid flows
indicates a different scaling for k than what is given in Fig. 1. This, in
addition to the observation that the same results are obtained in systems with
widely different wave properties (our data and the data of Tsacoyannis), would
indicate that the results in Fig. 1 are not explained by a simple wave
mechanism.
Nevertheless, we have not ruled this out as a possibility^ Consequentlys
we are currently carrying out experiments which relate mass transfer rates to
wave properties. We hope to be able to give a report on these at the
Conference,
References:
Suzanne, Christian, These du Grade de Docteur-Ingenieur, L'Universite Paul-
Sabatier de Toulouse, 1977.
Tsacoyannis, Jean, These du Grade de Docteur-Ingdnieur, L'Universite Paul-
Sabatier de Toulouse, 1976.
10/19/82
-------
Memery and Merlivat
GAS EXCHANGE ACROSS AW AIR-WATER INTERFACE : RESULTS Of EXPERIMENTS
ANP MODELING Of BUBBLE CONTRIBUTION TO GAS TRANSFER
LouAent MEMERV and LLLLan 9 m/4). a dtidden
/amp 0;$ -the valued o^ fe. -cd ofa^eAved. Thx^d ob^eAvatton -cd -cnteApAeted ad bexng
the Aedatt ojj the ondet ojj faAeafccng waved wh-tch cAeate babbled thAough whxch
mad4 ttand^eA tafeed p^ace. No dt^eAence ^OA gad exchange extheA ^.n ^edh wateA
OA xLn dea wateA had been >cdent££ted ^Aom the who£e 4et o^ expeA-cmentd .
A ^Asit theoAetcca£ appAoach to gad t^and^eA fai/ babbled ^d andeAtafeen.
CeAtacn poAameteAd wh^,ch aAe neglected bt/ dmooth axA-wateA -oateA^ace mode£x.ngd
aAe dtadoed. It- -td iJoand that tAond^eA ve^occttf -cncAeaded when boiubJUiify decAeaded
FuAtheA, babble oveApAed^aAe £eadd to wateA dapeAdatoAatcoKi at eqa-ctcbAuun, thxd
dupeAdatuAotton bex.ng moAe d-cgnt^cant ^OA ^.ed-d io^ub£e gaded. 1^ the
veioccti/ AemaxLnd Aough£t/ condtant ^OA a vaAcafo£e concenttatton gAa
ec.tu£cbAtum, ^ctd Aange 0|J vaAcatton becomed -on^nxte neaA eqtuZcbAiam.
an e^oAt ^d made to xJtcoApoAate doA^actantd -cnto the modeling.
-------
- 2 -
Memery and Merlivat
Beotu&e the, notion o£ tsuw&fieA v&tocity tuAne.d cLuie.ctty to the. ££ux ^t&ii^ : the. ££ux JLt> a. tineast function
o£ firm conc.&ntnatLon giadiznt* At tzabt ^on. &ia.c,zni>, the, c.oe.6&Lci&ntt> o$ thL&
function OJUL e,ntuie£y de^otecf by the. phy&i,co-diemLc&t ptwp&LtizA o£ the. g&&
and by the. bubble, dL&tru,bution. Re^atts o& YUMUU.CJOJL apptlcjation& cme.
foti kuLLum, siadon, angon and cmbon dioxyde..
-------
Metcalf
Iii Situ pH Measurements as an Indicator of C07 Gas Transfer in
GTaciaT Melt Waters L
by
Richard C. Metcalf
Department of Geography
University of Manchester
MANCHESTER M13 9PL
England, United Kingdom
In situ pH measurements in glacial melt waters from glaciers in
Washington, Alaska, and Switzerland, during 1977-1981 have firmly
established that CC^ gas transfer across the air-^water interface is the
dominant process controlling hydrogen ion activity in these dilute waters.
Detailed understanding of this phenomenon is critical to : 1) evaluating
"acid-rain" pH observations from snow in the high Arctic, 2) understanding
the roles of regelation and subglacial precipitation in controlling the
basal sliding of glaciers, 3) correctly evaluating thermodynamic mineral
stability diagrams for rock weathering in glacier hydrologic systems, and
4) understanding the nature and kinetics of CO- cycling in the high alpine
environment in an effort to further knowledge of the global C09 budget.
^
During initial spring snow melt, melt waters in the stream exitting
the glacier terminus were often oversaturated with respect to PCQ , and
_ ^
rapidly increased pH during outgassing at 0 C and ambient atmospheric
pressure. This appears to be a result of CCL purification within the
glacier by a presently undetermined process. Not surprisingly, after years
of purification, summer melt waters from "old" glacier ice in the ablation
zone of a temperate glacier are usually undersaturated by 10 to 100 times
with respect to atmospheric PCQ , and rapidly lower their pH values to
achieve equilibrium upon encountering the atmosphere. Especially during
summer, proglacial stream waters sometimes show pH increases from rock
weathering with the rate limited by the transfer of 0)2 across the air-water
interface to drive the weathering, reactions.
-------
Metcalf L
dpH/dt experiments demonstrate that glacial waters achieve equilibrium
without agitation within 2 to 3 hours in 0.25 dm polythene beakers at
in situ pressures and temperatures. The change in pH is exponential with
time, driven by the Pm gradient between the sample and the atmosphere.
By comparison, melt waters in the shallow, wide proglacial streams generally
achieve pH equilibrium in 10 to 15 minutes, primarily hastened by enhanced
gas transfer across the often turbulent stream surface.
-------
Mines and Sherrard
THEORETICAL AND EXPERIMENTAL EVALUATION OF OXYGEN TRANSFER AND
TOTAL OXYGEN REQUIREMENTS IN BIOLOGICAL WASTEWATER TREATMENT
by
Richard 0. Mines, Jr.
and
Joseph H. Sherrard
Graduate Student and Professor
Dept. of Civil Engineering
VPI&SU
Blacksburg, VA 24061
A year long study was conducted to investigate the ability of the
two film theory to accurately predict oxygen transfer in the activated
sludge process and to evaluate the hypothesis that the efficiency of
aeration devices installed in activated sludge systems are not fixed but
increase with an increase in the oxygen uptake rate as purposed by
Albertson and DiGregorio [1]. Biochemical stoichiometric equations are
developed as a function of wastewater composition and the mean cell
residence time to theoretically predict oxygen requirements. The
theoretical valves are compared with actual valves obtained from an
extensive laboratory investigation.
Two laboratory scale activated sludge reactors were operated over
a 3 to 20 day mean cell residence time. Various analytical tests were
conducted on samples taken from the influent, effluent, and mixed liquor
so that complete materials balances could be written. The oxygenation
capacity and the overall oxygen transfer coefficient (K-,a) of the diffuser
stones (utilized for aeration) was determined in tap water and the waste-
water effluent from each of the reactors under steady state and nonsteady
state conditions. Presently, data are being obtained and evaluated.
Preliminary results indicate that the theoretical valves agree with actual
valves and that activated sludge systems operating under a dissolved
oxygen limitation yield an effluent with a high nitrite-nitrogen
concentration in the range of 200 to 300 mg/1 if the influent has a
total kjeldahl nitrogen of approximately 450 mg/1.
Albertson, 0. E. and DiGregorio, D., "Biological Mediated Inconsistences
in Aeration Equipment Performance," Jour. Water Poll. Control Fed.,
47_, 976-988 (1975).
-------
Monahan and Spillane
THE ROLE OF OCEANIC WHITECAPS IN AIR-SEA GAS EXCHANGE
BY
Edward C. Monahan & Michael C. Splllane
University College, Galway,
I re 1 and
The wind dependence of the air-sea gas exchange rate, as
expressed by the gas transfer coefficient or piston velocity,
k8, has variously been described as being linear (k'acU? Broecker,
1979), quadratic (k'
-------
Munz and Roberts
The Ratio of Gas-Phase to Liquid-Phase Mass Transfer Coefficients
in Gas-Liquid Contacting Processes
Christoph Munz and Paul Roberts,
Stanford University
The objective of the work reported herein uas to assess the relative
importance of gas- and liquid-phase mass transfer coefficients in
diffused aeration, surface aeration and packed columns. Recently, each
of these processes has been considered as a possible method to strip
volatile halogenated organic compounds from solution (1,2,3,4). When
modelling these systems, previous workers (1,3,5,8,14) consistently have
assumed a ratio of gas phase mass transfer coefficient to liquid phase
mass transfer coefficient (kg/ki) of 150 as the criterion to assess the
relative importance of gas and liquid side resistances. The value
kg/ki-150 was inferred from average estimates for kg(H20) and ki(C02) at
the air-sea interface by Liss and Slater(6) and thereafter applied to
volati1 nation/absorption at air-uater interfaces for treatment
processes(1,3,4,5,8,14) as well as natural processes(8,9,10,11).
However, it is likely that this ratio depends significantly on the type
of system or process and possibly also within a given system depending
on the flow conditions.
Laboratory experiments wore carried out modelling diffused aeration,
surface aeration, and packed columns in which the following compounds
were stripped out of solution while oxygen was absorbed (stripped in
packed column) simultaneously: CC12F2, CHC13, CH3-CC13, CClt,, CHC1=CC12
and CClj=CCl2- The surface aeration device consisted of a baffled
cylindrical glass vessel with a liquid volume of 7.3 liters; a ring-
guarded stirrer with three pitched blades uas used. The bubble aeration
-------
Munz and Roberts
Page 2
device consisted of a glass column 22.5cm in diameter; the diffuser was
a ceramic perforated plate and the liquid volume was 18 liters. Packed
column experiments were carried out in the same glass column. The
packing material was 1/2-inch ceramic berl saddles with a packed height
of 40cm. In all experiments non-adsorbing materials uere used: glass,
teflon, stainless steel and ceramic. Organic compounds were determined
by gas chromatography, and oxygen was monitored continuously with a
probe. The procedure for surface and diffused aeration consisted of
analyzing the change in organics concentration with time. In the packed
column experiments five replicate influent and effluent samples were
analyzed after reaching steady state. In each experiment an overall
mass transfer coefficient, Kn*a, for each of the compounds was obtained
by using an appropriate model (accounting for partial gas phase
saturation in bubble aeration( 12)). Linear regressions of 1/Kn»a vs.
1/He were carried out to estimate kg«a and ki*a according to the concept
of additivity of resistances:
1 1 1
(1)
Hc»kg»a
Although both kg and ki depend somewhat on diffusivity. the approach is
reasonable because the diffusivities of the halogenated organics differ
little (Table 1). Oxygen data were not included in the regression
analysis. Henry's Constants, Hc, were also measured(IS) except for
CC12F2 and 02 (Table 1).
Fourteen surface aeration experiments were carried out at varying
stirring rates from 131 to 376 rpm(13). The results of the regression
- 3 -
-------
Munz and. Roberts
Page 3
analysis are presented in Table 2. It is evident that the ratio kg/ki
decreases with increased stirring. At lou agitator speeds kia increases
faster than kga. At higher speeds both values increase approximately
proportionally! reaching what appears a limiting value of kg/ki « 20;
the correlation coefficients are in general >0.95 in the latter range.
Also, the validity of the regression can be judged by comparing the
measured Kia(CCl2Fz) values with the predicted kia'ss these should agree
closely> which indeed is the case.
Fourteen packed column experiments were carried out(7) varying the gas
flow rates from 2 to 16 lit/min. The results are presented in Table 3.
The ratio kg/ki increases with increasing gas flow rate, but is
substantially smaller than in surface aeration. While all correlation
coefficients exceed 0.92, the predicted k ia's differ substantially from
the measured KiaCCCl 2^2) values at low gas flows. This is because the
gas phase resistance is more important than the liquid phase resistance,
i.e. the slope of the regression (1/kga) is large and the Kia(CCl2F2>
contributes little to the regression. To account for this, the data
will be reanalyzed to obtain Kga values instead of Kia values.
Six preliminary bubble aeration experiments have been carried out at
varying gas flow rates from 1.7 to 7-7 lit/min. Results are shown in
Table 4, and in this case the ratio kg/ki appears to decrease with
increasing gas flows. Negative predicted kga values can be attributed to
insufficient mixing. An additional set of experiments will be conducted
to resolve this difficulty before the Symposium.
-------
Munz and Roberts
Page U
In summary. the data presented indicate strongly that kg/ki ratios vary
from system to system and also with flow conditions within a given
system. It is dangerous to assess the relative importance of gas and
liquid side resistances based on a generally accepted value of
kg/ki = 150.
REFERENCES
1. Matter-Mueller, C., W. Gujer, W. Giger, Transfer of Volatile
Substances from Mater to the Atmosphere, Uater Research, 15:1271,
1981.
2. Roberts, P. V., c. Munz, P.Daendl iker, Removal of Volatile Halogena_
ted Solutes by Gas Transfer in Surface and Bubble Aeration, Paper
presented at the 55th WPCF Annual Conference, St.Louis, 1982.
3. Mumford, R. L., J. L. Schnoor, Air Stripping of Volatile Organics in
Water, Paper presented at the AWMA Annual Conference, Miami, 1982.
4. Kavanaugh, M. C., R. R. Trussell, Design of Aeration Towers to Strip
Volatile Contaminants from Drinking Water, J. AWWA, 72(12) :G34,
1980.
5. Rathbun, R. E., D. Y. Tai, Technique for Determining the Volatiliza_
tion Coefficients of Priority Pollutants in Streams, Water Research,
15:243, 1981.
6. Liss, P. S., P. G. Slater, Flux of Gases Across the Air-Sea
Interface, Nature, 247:181, 1974.
7. Riojas, A. H., Trace Contaminant Volatilization in Packed Columns
with Two-Phase, Countercurrent Flow, Engineers Thesis, Civil Eng.
Dept., Stanford University, Stanford, CA., 1982.
8. Smith, J. H., D. C. Bomberger, D. L. Haynes, Prediction of the
Volatilization Rates of High-Volatility Chemicals from Natural Water
Bodies, Env. Sci. Tech., 14(11):1332, 1980.
9. Mackay, D., P. J. Leinonen, Rate of Evaporation of Low Solubility
Contaminants from Water Bodies to the Atmosphere, Env. Sci. Tech.,
9(13):1178, 1975.
10. Smith, J. H., D. C. Bomberger, Jr., D. L. Haynes, Volatilisation
Rates of Intermediate and Low Volatility Chemicals from Water,
Chemosphere, 10(35:281, 1981.
- 5 -
-------
Munz and Roberts
Page 5
11. Dilling, W. L., Interphase Transport Processes. II. Evaporation
Rates Rates of Chioromethanes, Ethanes, Ethylenes, Propanes, and
Propylenes from Dilute Aqueous Solutions. Comparisons with
Theoretical Predictions. Env. Sci. Tech., 1K4):405, 1977.
12. Munz, C., P. V. Roberts, Mass Transfer and Phase Equilibria in a
Bubble Column, Paper presented at the AUWA Annual Conference, Miami *
1982.
13. Roberts, P. V., P. Daendliker, C. Hatter, C. Munz, Volatilization of
Trace Organic Contaminants During Surface Aeration: Model Studies,
Technical Report No. 257, Civil Eng. Dept., Stanford University-
Stanford, CA., 1981.
14. Albin, G. W., G. C. Holdren, Removal of Organics from Water in an
Aeration Basin: A Mathematical Model, in press, 1982.
15. Munz, C., P. V. Roberts, Transfer of Volatile Organic Contaminants
into a Gas Phase During Bubble Aeration, Technical Report No. 262,
Civil Eng. Dept,, Stanford University, Stanford, CA., 1982.
- 6 -
-------
Neal
REAERATION MEASUREMENT IN SWAMP STREAMS
RADIOTRACER CASE STUDIES
Larry A. Neal
Law Engineering Testing Company
Reaeration in swamp-like streams is particularly difficult
to characterize due to the unusual variability of swamp stream
hydraulics and the difficult access for field survey crews. The
normal regulatory tendency is to use "conservative" (low) estimates
of reaeration rates when confronted with a swamp stream situation.
Probably due to the difficulty in studying swamp streams, there
is little published data on swamp stream reaeration.
This paper presents the results of radio-tracer (krypton - 85
and tritium) studies of reaeration in several different swamp
streams. These swamp streams are located in Georgia, Louisiana,
North Carolina, and South Carolina with flows ranging from a few
cubic-feet per second (cfs) up to about two-hundred cfs. Flow
velocities in the study segments range from less than 0.05 feet
per second (fps) up to about one fps with measured reaeration rates
from about 0.05 to over 1.0 (per day, base e, 20°C).
In addition to the tracer-measured reaeration rates, some
energy dissipation data are presented indicating that these swamp
streams are relatively efficient in utilizing available head for
mixing and resulting reaeration. Interim recommendations
are provided concerning reaeration predictions for swamp-like
streams and' further study needs are identified.
-------
Nguyen et al.
EXCHANGE RATES OF GASEOUS SULFUR COMPOUNDS BETWEEN OCEAN AND ATMOSPHERE
Ba Cuong NGUYEN, Christian RFRGFRFT and Gerard LAMBERT
Centre des Foibles Radioact-ivites
Laboratoire Mixte CNRS CEA
BP n°l
91190 - Gif-sur-Ivette,, France
Biological activity plays an important role in the gaseous sulfur
compound exchanges between ocean and atmosphere. Indeed, this activity
evolves organic sulfur compounds, one of which is dimethyl sulfide (DMS).
This gas is rather insoluble and volatile, it escapes from the water,
oxidizes in the atmosphere to give S02 and then sulfates.
Two cruises, in the Indian and Pacific Oceans, were undertaken
to measure DMS at the ocean-atmosphere interface. Measurements were taken
in the water and in the air at heights of 1, 10, and 17 m above the ocean
surface. The results obtained during these voyages show a good correlation
between surface water and atmospheric concentrations. These concentration
gradients from surface water layer to atmosphere enable us to determine the
diffusion coefficent in the first 17 meters altitude over ocean, and conse-
quently the ocean-to-atmsophere transfer coefficient (or piston velocity).
These parameters depend mostly on meteorological factors (wind scale and
sea state, etc,..).
From these parameters, gaseous sulfur compound oceanic production
in the atmosphere could be estimated to be about 30 x 106 tons of sulfur per
year.
-------
0'Connor
(Invited Keynote Jr'aper)
THE SIGNIFICANCE OF GAS-LIQUID EXCHANGE
IN WATER QUALITY MODELING & ASSESSMENT
Donald J. O'Connor
Manhattan College, New York
Environmental probleus, in which gas-liquid exchange plays a major role,
concern aesthetics, eutrophication and toxicity. These are not only of
environmental importance, but also are of economic significance, in view of the
costs of treatment and control methods which are invariably required to address
such problems.
Rational water quality management requires the application of mathematical
models which describe the spatial and temporal distribution of the relevant
water quality constituents. These models contain transport, transfer and
kinetic components. While the transport factors are fairly well established for
many natural water systems, little fundamental basis exists for the
quantification of the kinetic interactions and transfer routes. These
coefficients are either assigned empirically or calculated from observations in
the prototype systems. The one notable exception is the gas-liquid transfer
process, for which at least some theoretical basis exists for the determination
of the relevant coefficient. An objective element is-thereby provided in the
important process of the validity of the water quality model. Furthermore, the
fundamental nature of the gas transfer relationships permits extrapolation to
other environmental conditions, an element which is an integral part of the
planning process. The coefficient, which quantifies the process, is one which
may be determined without recourse to water quality measurements in the
prototype and thereby, provide for the validation of water quality models and
the subsequent application in water resources management.
The specific constituents considered are oxygen, carbon dioxide, ammonia,
sulfur dioxide and volatile organic chemicals, whose concentration in natural
waters is effected by the discharges of numicipal, agricultural and industrial
wastewaters, as well as been natural runoff and drainage. The first three gases
are important because of their roles in chemical and biological cycles. Ammonia
is toxic to aquatic organisms, as are many volatile synthetic chemicals. Carbon
dioxide, an integral element in the pH-acidity-alkalinity balance of a natural
water systems, is important in acid mine drainage and dry and wet deposition
from teh atmosphere.
A brief description of general elements of water quality models is first
presented. A review of the gas transfer relationships relating to dissolved
oxygen, carbon dioxide, ammonia, sulfur dioxide and volatile organic chemicals
follows. The various viological, chemical and physical factors which affect
these constituents are discussed, including the significance of the gas-liquid
exchange. The types of models, commonly employed in the analysis of dissolved
oxygen and reactive toxics, are reviewed. For the analyses of pH and inert
toxics, the general structure of the models are suggested. For each
constituent, a practical methodology of environmental analysis and examples of
practical applications are presented.
-------
O'Connor
ABSTRACT
Transfer Coefficients of Smooth Transitional and Rough Flows
D.J. O'Connor
Manhattan College
Many constituents, which are susceptible to transfer through the
air-water interface, are of environmental importance. Factors which
affect the exchange rate across this boundary are the air and water
velocities, the relative smoothness-roughness of the surface, and the
physio-chemical characteristics of the interface and of the transferable
constituents. Fluid motion exerts a shear at the interface, which
establishes the structure of the boundary layers in both media. The
dynamic features of these interfacial regions, defined the viscous
sublayer and roughness height, influence the transfer between air and
water. Velocity functions are developed for both smooth and rough
surfaces. Relationships between the transfer coefficient and these
hydrodynamic parameters are established for each regime. A transition
function is also proposed for the momentum transfer which is used as a
basis to formulate the mass transfer relations. The overall transfer
coefficient is expressed in the usual manner of resistances in series,
involving both the film and surface renewal concepts. The latter is
predominant for smooth flow and the former for rough flow. In the
transition, the viscous sublayer and thus the diffusional sublayer are
eroded with a simultaneous growth of roughness effects and cavities, the
latter imparting an additional resilience to transfer. A constituent is
characterized by the solubility and partial pressure (Henry's Constant)
and by its diffusivity and viscosity of air and water (Schmidt Numbers).
Both liquid and gas film control are included in the applications. The
proposed model yields reasonable and consistent correlation with
transfer data from a number of laboratory systems. Extrapolation of the
relationships to prototype conditions are considered.
-------
Pankow et al.
INTERNATIONAL SYMPOSIUM ON GAS TRANSFER AT WATER SURFACES
Cornell University
Ithaca, New York
June 13-15, 1983
James F. Pankow and William E. Asher
Department of Environmental Science
Oregon Graduate Center
19600 N.W. Walker Road
Beaverton, OR. 97006
and
E. John List
Department of Environmental Engineering Science
138-78
California Institute of Technology
Pasadena, CA. 91125
CARBON DIOXIDE GAS TRANSFER AT GAS/WATER AND OIL/WATER INTERFACES
AS A FUNCTION OF SYSTEM TURBULENCE
For processes which are not rate limited by chemical reactions,
interfacial mass transfer will be controlled by the fluid mixing. This
applies to transfer across both the gas/water and the oil/water (e.g., pe-
troleum spill) interfaces. As they involve environmental water bodies, such
processes are therefore often controlled by the degree and nature of the
turbulence underlying the interface. Efforts in our laboratories have been
directed towards the use of the pH-dependent laser-induced fluorescence (LIF)
of fluores.cein compounds to characterize the mass transfer of C02 (an acid)
-------
Pankow et al
-2-
across both the gas/water interface and the oil/water Interface. The photo-
diode monitoring of the LIF intensity in dilute fluorescein solutions
(vLCT8 - 10~7 M) as a function of depth allows the calculation of the C02
flux across the interface as well as across any internal solution surface.
Fine structure superimposed on the average signal has been found to be
caused by the mixing of turbulent eddies of fluid particles of dissimilar
net CC>2 content. Turbulence has been generated by means of an oscillating
grid. Mass transfer measurements have been made under conditions of varying
turbulence intensity, length scale, and distance to either the gas/water or
the oil/water interface. (In the latter case, the oil phase is maintained
saturated at a constant CC>2 level.) Experimental data will be presented and
conclusions will be drawn for gas transfer at the air/sea and oil slick/sea
interfaces.
-------
Papadimitrakis et al,
MEASUREMENTS OF THE FLUCTUATING PRESSURE IN THE TURBULENT BOUNDARY LAYER
OVER PROGRESSIVE, MECHANICALLY GENERATED WATER WAVES
By YIANNIS ALEX. PAPADIMITRAKIS, EN YUN HSU, and ROBERT STREET
Department of Civil Engineering
Stanford University, California 94305
The structure of the pressure and velocity fields in the air above mech-
anically generated water waves was investigated in order to evaluate their
contribution to the transfer of momentum and energy from wind to water waves.
Two wave-height gauges, an array of two X hot-film probes, a specially
designed high-sensitivity pressure instrument, and five piezocrystal pressure
transducers mounted on the roof of the Stanford wind-water wave facility were
used to monitor the water-wave height and the velocity and pressure fields in
the air.
The pressure and velocity measurements in the air boundary layer were
performed in a transformed Eulerian wave-following frame of reference at pre-
selected distances from the wave surface. Seven different wind speeds were
examined in the range 140.59-402.04 cm/sec, with 1 Hz, 2.54 cm amplitude,
mechanically generated water waves.
Measured acoustic and other pressure waves associated with the traveling
upstream-reflected water wave were found to alter the wave-induced pressure
behavior. These components are an order of magnitude greater than, and com-
parable to. the actual wave-induced pressure.
The nonlinearities of the propagating water wave and the drift current
seem to introduce pressure components at frequencies other than that of the
main wave. Wave-wave resonant interactions and wave-turbulence nonlinear
interactions are also responsible for the strong wave-induced pressure har-
monics.
-------
Papadimitrakis et al.
Page 2
The wave-induced composite pressure coefficient, at the fundamental mode,
shows in general an exponential decay behavior, but the rate of decay is dif-
ferent from that predicted by potential flow theory.
The relative wave-induced pressure phase remains fairly constant through-
»
out the boundary layer, except at c/Ufi = 0.780 and 0.677. This phase dif-
ference was found to be about 130° during active wave generation, pressure
lagging waves.
The momentum and energy transfer rates supported by the waves were found
to be dominated by the wave-induced pressure, but the transfer of the corres-
ponding total quantities to both waves and currents may or may not, depending
on c
°o
The contribution of the wave-induced Reynolds stresses to the transfer
processes is negligible.
-------
Peng
INVASION OF FOSSIL FUEL C02 TO THE SEA
T.-H. Peng
Oak Ridge National Laboratory1
Oak Ridge, Tennessee 37830
ABSTRACT
Our knowledge of gas transfer at water surfaces can be applied to
environmental problems such as the uptake of fossil fuel CO^ by the
ocean. CO- exchange rates across the sea-air interface have been
successfully estimated from radon profiles in the upper few hundred
meters of the ocean measured during 6EOSECS cruises. The amount of
anthropogenic C02 uptake is estimated by the one-dimensional box-
diffusion model of Oeschger et al. (1975). The zonal variations of
COp exchange rates and vertical mixing rates as derived from 6EOSECS
tritium measurements are included in these model calculations. The
results of these calculations and the amount of fossil fuel C0?
reaching major active carbon reservoirs will be discussed.
^Research sponsored jointly by the National Science Foundation's
Ecosystem Studies Program under Interagency Agreement No. DEB 8115316
and the Carbon Dioxide Research Division, Office of Energy Research,
U.S. Department of Energy, under contract W-7405-eng-26 with Union
Carbide Corporation.
-------
n™ I? u?T?ne^ented at: iNTERNATIONAL SYMPOSIUM ON GAS TRANSFER
TRANSFER AT WATER SURFACES to be held from June 13-15. 1983
REAERATION OF TURBULENT WATER
E.J. Plate and R. Friedrich
The reaeration of oxygen depleted turbulent water through the water-
atmosphere interface has been studied extensively, but the phenomenon
has not been fully explained. The most logical qualitative explanation
is the surface renewal theory according to which the water layer very
nearest the interface is enriched by molecular processes and then
swept into the interior of the water by turbulent eddies which pene-
trate the surface and remove the enriched layer. In the interior of
the water body, the oxygen rich water is mixed with its surroundings.
The process of surface renewal is diffuclt to be quantified, and even
more difficult to investigate experimentally, because it is non-stationary
and involves measurements near the surface which cannot be conducted
by standard transducers, such as hot wire anemometers.
It has been our belief that in order to solve the reaeration problem one
has to rely on circumstantial evidence, in particular on the action of
turbulence. Therefore we set up a program of studying reaeration on inter-
faces at which the turbulence has been generated by many different mecha-
nisms. From these experiments we hope to amass a body of experimental data
against which theories can be checkedj_The following experimental set ups
have been used: a stirred tank, with turbulence generated by an arrangement
of parallel horizontal rods located near the bottom of the tank, which could
be rotated about their axis; a wind wave tank with standing water, in which
the surface turbulence was generated by the shear of the wind and the wind
waves, and a wind wave tank with flowing water, in which waves and currents
acted as turbulence generators. Quantities which were measured included velo-
cities, turbulence intensities and spectra, wind waves, and optically visible
turbulent cells.
Whereas the tank experiments lend support to the surface renewal theory
in the form given by Fortescue and Pearson (1967), one finds that the
wind wave results require a different explanation. Theoretical and experi-
mental arguments will be presented which we hope will help towards establish-
ing a unified surface renewal theory without conflict between experimental
data and theory. The importance that this theory may have in evaluating
reaeration in environmental and chemical engineering problems justifies
extension and continuing experimental and theoretical efforts.
-------
Rathbun and Tai
VOLATILIZATION OF CHLORINATED HYDROCARBONS FROM WATER
By
R. E. Rathbun and D. Y. Tai
U.S. Geological Survey, NSTL Station, MS 39529
ABSTRACT
The distribution and the fate of organic compounds in the
waters of our environment are determined by the interactions of
a number of complex chemical, physical, and biological processes<,
One of the most important physical processes for many compounds
is volatilization which is the transport of the compound from the
water across the water-air interface into the air.
This report considers two aspects of volatilization: (1) the
concentration dependence of the volatilization coefficient; and
(2) laboratory measurements that permit estimation of the
volatilization coefficients of organic compounds in streams without
actual introduction of the compounds into the stream,. The discussion
of these two aspects is based on laboratory measurements of the
volatilization characteristics of 1,1,1-trichloroethane and
1,2-dichloroethane. Both these compounds are on the U.S. Environmental
Protection Agency list of priority pollutants, and both are significant
contaminants in surface and ground waters.
-------
Rathbun and Tai
Page 2
Volatilization coefficients for the chlorinated hydrocarbons
and absorption coefficients for oxygen were measured in a stirred
tank in the laboratory. The general procedure consisted of stripping
the water with nitrogen gas to reduce the dissolved oxygen concentration,
adding water containing sufficient dissolved chlorinated hydrocarbon
to give the desired concentration, stirring the water at a constant rate,
and measuring concentration as a function of time. Hydrocarbon
concentrations were determined by a strip-and-trap procedure followed
by gas chromatographic analysis with a flame ionization detector.
Oxygen concentrations were determined using the Winkler technique.
Experiments for each of the compounds at constant mixing
conditions over a wide concentration range showed that the volatilization
coefficient was independent of concentration. This result confirms
the assumption that the volatilization process follows first-order
kinetics analogous to the kinetics for the absorption of gases such
as oxygen.
Simultaneous measurements of the volatilization coefficient
for each of the compounds and the oxygen absorption coefficient
over a wide range of mixing conditions showed that the ratios of
the chlorinated hydrocarbon and the oxygen coefficients were independent
of mixing conditions. The fact that these ratios are constant
makes it possible to estimate the volatilization coefficients for
these compounds for a given stream from estimates of the oxygen
absorption coefficient.
-------
Arturo Riojas and Paul Kruger
VOLATILIZATION OF FISSION PRODUCTS Civil Engineering Department
Stanford University
IN NUCLEAR REACTOR BUILDINGS Stanford, CA 94043
Fission products released from a nuclear reactor core are distributed
throughout the containment vessel: in the reactor building atmosphere, in
the cooling system, on building surfaces, and in pooled waters that may
form in the building. Of particular concern are the volatile radionuclides,
especially the short- and long-lived radioisotopes of iodine, which influ-
ence markedly the cost and methods of cleanup. The potential for volatili-
zation following a specific accident depends on the conditions under which
a contaminated pool of water formed and the thermo- and hydrodynamic condi-
tions for volatilization from the pool. The potential for volatilization
also is dependent on the competing processes, such as adsorption, pre-
cipitation, and rainout in sealed buildings. A study to identify and
model competing processes that influence volatilization of fission products
in pooled waters is underway. The model is designed to incorporate these
competing processes into a prediction of volatilization as a function of
water quality parameters, reactor building conditions, introduction
of abatement agents and time.
Input to the model is the likely distribution of fission products in
a reactor accident of given severity. The model compiles a data base of
pertinent literature data for the several processes included in the numerical
model. Laboratory experiments are underway to acquire necessary data not
in the literature and a scale model of the TMI-2 containment vessel has
been constructed for feedback and verification experiments to improve the
model.
Processes under study include: volatilization from liquid- and solid-gas
interfaces, rainfall in the saturated-humidity enclosed vessel, precipitation
of solids from the pooled water, adsorption from the liquid phase on
1
-------
Abstract (Continued) A. Riojas and P. Kruger
building surfaces, gas-phase adsorption on building surfaces, liquid-phase
adsorption onto suspended solids, gas and aerosol absorption by sprays and
condensation, and gas absorption by falling films.
-------
Roberts
DEPENDENCE OF MASS TRANSFER
ON ENERGY DISSIPATION DURING SURFACE AERATION
Paul V. Roberts
Department of Civil Engineering
Stanford University
Stanford, CA 94305
It has been reported previously that the oxygen mass transfer rate
constant (IL. a) is proportional approximately to the specific energy dis-
sipation (P/V(W/m ) in streams (1) , large-scale surface aeration as practiced
in wastewater treatment (2), and in laboratory studies of surface aeration (3).
The objectives of this paper are three-fold: to compare the proportionality
coefficients relating IL a to P/V in several kinds of systems observed; to
present experimental evidence that the values of !L a for volatile organic
solutes are approximately proportional to that of oxygen; and to assess
whether the values of the proportionality coefficients are consistent with
mass transfer models such as the surface renewal theory. Significant
relations between the value of IL a for oxygen and the value of P/V have
been reported for streams (1), large-scale surface aerators (2), and a
laboratory-scale surface aerator (3). The dependence of IL. a on (P/V) dem-
onstrates a near-linear proportionality for each of the several systems.
Differences between the systems can be explained qualitatively in terms of
the method of energy dissipation. In large-scale surface aerators the
energy dissipation is concentrated at the air-water interface; therefore,
the value of IL a is relatively large compared to the other contacting
systems, for a given specific energy input. Conversely, the value of K_a
for a given energy input is relatively low in streams because energy dis-
sipation occurs largely at or near the stream bed. The laboratory experi-
ments reported here represent an intermediate case because the agitator used
was conceived as a combination surface aerator and mixer, with energy input
to mix the bulk liquid as well as to promote air-water contact.
The value of ILa for volatile organic solutes (CCl^, CCl^, CCl^CCl™,
CHC1=CC12, CH-CCl-, and CHC1-) was found to be proportional to that of
oxygen over a wide range of power inputs. The coefficient of proportionality,
3i = (ILa)./(tL a) was in the range of 0.55 to 0.65 for the organic solutes
-------
Roberts
Page 2
studied. The proportionality coefficient 3 was found to depend on the
ratio of diffusivities to the 2/3 power, in accordance with the boundary
layer and film-penetration models. Nonetheless, the observation that K.a
is proportional to specific energy input seems inconsistent with mass
transfer theory as applied to a plane air-water interface. The surface
renewal theory predicts a significantly weaker dependence of IL. a on P/V,
03
e.g., ILa <* (P/V) * if the surface renewal rate is estimated as suggested
by Davies and Khan (4). Similarly, boundary layer analogy approaches
predict 1C a.proportional to (P/V) raised to an exponent in the range
0.2 to 0.3, when the characteristic average velocity is evaluated using
either the Blasius Relation or Kolmogoroff-Turbulence Theory (5).
It appears that specific energy input is a useful predictor of mass
transfer at the air-water interface in natural systems as well as engineered
surface aeration contactors.
References
1. Tsivoglou, E. C., and L. A. Neal (1976). J. Water Pollution Control
Federation, 48, 2669-2689.
2. Tchobanoglous, G. (1979). Wastewater Engineering: Treatment, Disposal,
Reuse, 2nd edition, Mc-Graw-Hill Book Co, p. 497
3. Roberts, P., P. Dandliker, C, Matter, and C. Munz (1981). Volatilization
of Trace Organic Contaminants during Surface Aeration: Model Studies.
Technical Report No. 257, Stanford University, Department of Civil
Engineering, Stanford, CA. 92 pp.
4. Davies, J. T., and W. Khan (1965). Chem. Eng. Sci., 20, 713-715.
5. Davies, J. T. (1972). Turbulence Phenomena, Academic Press, New York
and London, p. 64
-------
Roether and Kromer
(Submitted to "International Symposium on Gas Transfer at
Water Surfaces" Cornell University, June 1983)
OPTIMUM APPLICATION OF THE RADON DEFICIT METHOD TO OBTAIN
AIR-SEA GAS EXCHANGE RATES
Wolfgang Roether and Bernd Kromer
Institut fiir Umweltphysik der UniversitMt Heidelberg,
Heidelberg, W. Germany
72?
The deficit of Rn relative to radioactive equilibrium
226
with the dissolved Ra in the ocean's mixed layer is caused
by 2 Rn escape into the atmosphere and allows to determine
gas exchange rates at sea. The extensive spot observations of
radon deficits by W. S. Broecker and coworkers (see Peng et al.,
J. Geophys. Res. 8!4_ (1979) 2471) give averages of this rate.
We persue a different approach, in which momentary gas exchange
rates are obtained in intensive and carefully controlled field
observations and are used to adapt gas exchange parameteriza-
tions resulting from laboratory and/or theoretical investigations
to field conditions.
We made 222Rn-deficit observations during JASIN 1978 (^59°N,
12.5°W) and FGGE 1979 (^2°S, 22°W) , employing a precise and fast-
acquisition, automatic radon measuring system (B. Kromer and
W. Roether, "Meteor" Forsch.-Ergebnisse A, in press) . The FGGE
gas transfer velocities average considerably lower than those
for JASIN, despite little difference in average wind velocity,
while the range is similar to that observed in laboratory experi-
ments. The JASIN/FGGE difference is ascribed to considerably
steadier winds met during FGGE. On the other hand, our 222Rn-
deficit time series showed a surprising degree of variability,
OO O
which we believe to be caused by " Rn redistribution in the
mixed layer during the period between generation and escape
into the atmosphere, i.e., a few days.
-------
Roether and Kromer
- 2 -
At present we are preparing the next generation of field
observations. The most important requirements are anticipated
O 9 ^
to be (i) careful bookkeeping of the mixed-layer Rn deficit;
and (ii) adequate monitoring of the external variables.
Ingredients for (i) are
• v
- Obtaining repeated 222Rn deficit observations MO min
repetition) over periods of days or longer, the short-term
Ra variability being monitored by salinity and silica
concentration, with, continuous control of system performance
and" with on-line evaluation.
- The observations are preferably to be carried out while
drifting with the horizontal mixed-layer flow. The local
structure of the mixed layer and thernocline is to be moni-
tored by temperature and salinity profiling and vertical
current shear is to be assessed. Verification of the
absence of mixed-layer fronts in the observation area is
desirable.
- A suitable observation area is to be selected.
As for (ii), meteorological variables have to be measured, but
furthermore also waves, and perhaps surface slicks, which play
an added role; improved understanding of the effects of
waves would help in outlaying the necessary measurements.
On the basis of our JASIN and FGGE results, we estimate
the achievable precision in gas transfer velocities in such an
observational program as ±20% under favourable conditions, and
the temporal resolution as 1Oh. The necessary measurements and
their precisions can be outlined, and criteria can be given for
the selection of an observation area.
-------
Seaward, ex
EFFECTS OF EVAPORATION AND CONDENSATION
ON AN ABSORPTION PROCESS
by
David O. Seaward, PJE.
Dr. Burton A. Segall
Dr., Alfred A. Donatelll
Dr. Charles R. Ott
ABSTRACT
Research was conducted to investigate the effects of evaporation and condensation on
a gas phase controlled absorption process. Individual droplets of distilled water of known
weight and temperature were dropped through a counter-flowing gas mixture that contained
a known concentration of ammonia and water vapor. The rates of evaporation and
condensation of water vapor were varied by modifying droplet temperature and the water
vapor content of the air/ammonia gas mixture. The droplets were collected after passing
through the gas mixture for each set of conditions and the collected solution was analyzed
for ammonia concentration. The resulting concentrations were then compared. Theoretical
Diffusion Prediction Equations were also derived to predict the effects of increased
evaporation and condensation rates on absorption.
The experimental results indicate that absorption rates increase with increases in
condensation rate and decrease with increases in evaporation rate. The experimental trends
showed good similarity with the trends projected by the derived Diffusion Prediction
Equations.
The intent of this paper is to present an overview of the test methods used, the results
obtained, and the conclusions drawn. Potential applications and implications of the results
are also discussed.
The theoretical Diffusion Prediction Equations are presented along with the simil-
arities between the experimental absorption trends and those projected by the Prediction
Equations discussed. Due to assumed time constraints, detailed derivation of the Prediction
Equations is presently considered outside the scope of this presentation.
-------
Simpson
ON THE EXCHANGE OF OXYGEN AND CARBON DIOXIDE BETWEEN OCEAN AND ATMOSPHERE
IN AN EASTERN BOUNDARY CURRENT. J. J. Simpson, Marine Life Research Group,
Scripps Institution of Oceanography, La Jolla, California 92093.
Simultaneous measurements of temperature, salinity, dissolved oxygen, pH, nitrate,
phosphate, silicate and chlorophyll-a were made continuously while underway from R.V.
New Horizon during July 1979. The observations were made over a region of the northern
California shelf bounded by Pt. Arena to the north and Pt. Reyes to the south, off the
southern California coast near Pt. Conception and enroute between these two areas. In
addition, discrete titration alkalinity and total carbon dioxide determinations were
made with the GEOSECS titrators approximately every 15 minutes during the underway
survey. Water for these measurements was drawn from a depth of 3 meters via the un-
contaminated seawater system of R.V. New Horizon. From these data the spatial distribu-
tions of dissolved oxygen, percent saturation of dissolved oxygen, partial pressure of
carbon dioxide and the PCQ? departure from atmospheric equilibrium were calculated.
These measurements show that the oceanic concentrations of dissolved oxygen and the
partial pressure of carbon dioxide in a coastal regime depart radically from equilibrium
values. The spatial coherence and phase between these distributions cannot be explained
solely by physical dynamics. The data support the hypothesis that the dynamics of gas
exchange in an eastern boundary current are largely controlled by biological processes.
The latter are clearly influenced by the physical dynamics associated with such features
as offshore jets, upwelling fronts and intrusions of warm Central Pacific water onto the
shelf. The spatial distributions of chlorophyll-a and nutrients independently support .
this conclusion. The global distribution of biological productivity and its significant
impact on the air-sea exchange of oxygen and carbon dioxide demonstrate the need to
develop reliable remote sensing techniques to monitor synoptically global gas exchange
processes and their possible effects on climate.
-------
Sivakumar
REAERATICN AND WIND INDUCED TURBULENCE SHEAR
IN A CONTAINED WATER BODY
M. Sivakumar, B.Sc.(Eng), M.Eng., Ph.D.,
Department of Civil and Systems Engineering,
James Cook University of North Queensland,
Townsville, Queensland, Australia, 4811.
Reaeration in contained bodies of water (such as lakes and ponds) is mainly
achieved by the oxygen transfer at the air-water interface and its subsequent
transport into bulk fluid. The action of wind induces shear and causes
turbulence at the air-water interface which can significantly enhance the
reaeration process.
The process of atmospheric reaeration (it is, in fact, reoxygenation) can be
simply defined as the physical absorption of atmospheric oxygen in water.
However, the primary resistance to oxygen absorption in water takes place
at the water side of the air-water interface. Detailed experiments were
conducted to study the relationship between reaeration and wind induced
turbulence in a laboratory size wind-water tank facility- In this paper,
the results of the turbulence measurements and their correlation to reaeration
coefficient will be presented.
The reaeration and turbulence measurements were made simultaneously- Hot-
film anemometer and surface floats were used to measure fluctuating and mean
components of the water drift. Spectral analyses were made on the fluctuating
components. These results indicate that there is a definite similarity
existing- between the wind induced water drift below the water surface and
flow past 'solid walls'. This is indeed interesting in so far as the
transfer theories developed for the 'solid walls' may be applied to mass
transfer occurring at the water side of the air-water interface.
-------
2. Sivakumar
Using the observed similarity and starting from the eddy cell theory of
mass transfer proposed by Lament (1970) for gas transfer in pipe flows,
an expression is derived relating reaeration coefficient and wind shear
velocity. Laboratory data seems to agree well with the derived expression.
-------
St. John et al.
THE SIGNIFICANCE OF GAS TRANSFER
ON THE DISSOLVED OXYGEN BALANCE
By: John P. St. John, Thomas W. Gallagher, and Donald J.
O'Connor
Dissolved oxygen is widely used as both a specific and
general measure of water quality conditions in the streams,
rivers, lakes and estuaries of the country. Sufficient levels of
dissolved oxygen are necessary to support fish life and
reproduction, and to maintain a balanced biological community.
Each state is obligated by the law to develop and enforce water
quality standards, and all such standards contain specific
numerical requirements for dissolved oxygen. The vast majority
of municipal and industrial wastewater treatment plants in the
United States were built primarily to protect the dissolved
oxygen resources of the nation's waterways.
In many critical cases, calculations are performed to relate
the effect of specific wastewater discharges to dissolved oxygen
conditions in the receiving waters. The purpose of such
calculations is to determine the specific level of wastewater
treatment which is necessary to achieve or maintain the local
disolved oxygen standard. The calculation procedure consists of
the application of a water quality model which incorporates
relevant transport, transfer, and kinetic terms associated with
the dissolved oxygen balance. Several of the kinetic factors are
associated with the utilization of oxygen for stabilization of
carbonaceous and nitrogenous materials present in the receiving
waters. A particularly important factor, however, characteristic
of such models is the atmospheric reaeration coefficient which
defines the rate at which atmospheric oxygen is transferred
through the air-water interface to replenish that which is
utilized in the stabilizaton of waste materials.
Many of the transport and kinetic factors incorporated in the
mathematical modeling framework are measured or empirically
determined from field water quality data. The atmospheric
reaeration coefficient, however, is generally estimated from one
of a number of formulations which have been developed over the
years by a variety of theoretical and/or empirical means. Many
of the more empirically based equations were developed from data
collected for specific ranges of those hydraulic and physical
characteristics which have a bearing on the gas transfer process.
Such formulations often give widely different estimates of the
atmospheric reaeration coefficient when applied in a problem
context. Their differences can seriously affect the calculation
of the oxygen balance, lead to a misinterpretation of the effect
of the wastewater discharge on the oxygen resources of the
waterway, and" result in an improper determination (too great or
-------
small) of the required degree of waste treatment. Hence,
accurate knowledge of the reaeration coefficient as represent-
ative of the gas transfer process is required for proper
environmental and economic decision making.
The proposed paper will illustrate the importance and
significance of the gas transfer process, as represented by the
atmospheric reaeration coefficient, on the oxygen balance and on
the decision making process, that is, on the proper determination
of the degree of required waste treatment. An example stream
situation will be presented and described. Wastewater inputs to
the stream will be described, and characteristic water quality
profiles, particularly dissolved oxygen will be shown. The
mathematical water quality model used for analysis of data will
be described as well as procedures for evaluation of kinetic
coefficients. It will be demonstrated that it is possible to
correlate a particular set of water quality (dissolved oxygen)
data with different estimates of the reaeration coefficient. It
will then be shown that the different reaeration coefficients,
when modified for and used to project water quality impacts at
critically low stream flows, yield markedly different results.
The degree of waste treatment necessary to maintain the
equivalent level of dissolved oxygen in the stream as calculated
with the different reaeration estimates will be indicated. The
practical significance of gas transfer and the importance of
atmospheric reaeration in the specifications and design of
wastewater treatment facilities will be discussed.
-------
Su and Green
Experimental Studies of
Surface Wave Breaking and Air Entrainment
by M.Y. Su and A.W. Green
Naval Ocean Research and Development Activity
NSTL Station, MS 39529
Tel. (601)688-4733 or 5241
Physical processes of air entrainment and subsequent air bubble generation
from breaking of steep surface gravity waves under controlled situations will
be described and discussed in this paper.
It is generally known from actual field measurements that wind-wave
breaking in lakes and oceans contributes by far the most significant portion
of air/gas/particle transfers cross the water surface; and yet, the underlying
physio-chemical processes are only poorly understood. To a large extent, this
slow progress made so far might be attributed to the highly nonlinear complex
wave breaking process itself, which has defied full understanding of physicists
and applied mathematicians despite a long history of both theoretical and
experimental investigations.
Recently, a close interplay between experimental and theoretical studies has
lead to an important breakthrough on the physics of deep-water wave breaking
as a consequence of a new type of three-dimensional wave instability and
subsequent bifurcation (for details see Su (1982), Su et al. (1982), McLean (1982),
Saffman & Yuen (1980, 1982)). This breakthrough now provides, in turn, a sounder
foundation for us to study the gas transfer resulted from such wave breaking
processes.
In this paper, we shall present some preliminary experimental results conducted
in a large-scale wave tank, which exhibits clearly the three-dimensional wave
plunging and spilling that cause air entrainment and bubble generation/transport
in the deep water. The rapid and complex physical processes involved are recorded
-------
Su and Green
.Eage_2 _
by photographic techniques. These experimental results are then used to delineate
general characteristics of the physical mechanisms involved.
Finally, we discuss the relationship of our results with previously published
results and oceanic observations.
References:
1. Su, M.Y., (1982), J.F.M. 124, 73-108.
2. Su, M.Y., et al., (1982), J.F.M., 124, 45-72.
3. McLean, J., (1982), J.F.M., 114; 315-330.
4. Saffman, P.G. and H.C. Yuen, (1980), J.F.M., 101, 797-808.
5. Saffman, P.G. and H.C. Yuen, (1982), J.F.M., 124, 109-121.
-------
Theofanous
(Invited Keynote Paper)
*
Conceptual Models on Gas Exchange
The characteristically high Sc number process of mass transfer at
free, turbulent gas/liquid interfaces implies complicating as well as
simplifying factors in the development of general predictive methods.
The simplifications arise due to the thinness of the concentration bound-
ary layer that often occupies only a minute fraction of even the smallest
turbulent eddies present. The convective field can therefore be con-
siderably simplified into elementary flows, hence convective diffusion
solutions become readily available. The difficulties arise in relating
the characteristic parameters of these flows to the turbulence state in
the bulk (i.e., to measurable turbulence properties). Since these flows
within the concentration boundary layer are not directly observable their
nature must be deduced by the mass transfer response of the interface
(i.e., through conceptual mass transfer models). This process is often
further complicated by the need to crudely estimate the bulk turbulence
(lack of simultaneous turbulence and mass transfer measurements).
Although considerable progress has been achieved along these lines during
the past fifteen years or so certain controversial aspects remain, primarily
due to the incomplete definition of mass transfer "regimes" within the
broad context in which such turbulent interfaces are encountered. A
critical, comparative evaluation of available conceptual models will be
presented in an effort to identify strengths, weaknesses, and possible
additional research needs.
*T.G. Theofanous, School of Nuclear Engineering, Purdue University,
West Lafayette, IN
-------
Thomas et al
Abstract for International Symposium on Gas Transfer at Water
Surfaces, Cornell, June 1983
ENTRAPMENT AND TEANSPOHT OF BUBBLES BY PLUNGING WATER
N.H. Thomas, T.E. Anton, K. Sene, J.C.E. Hunt
DAMTP, University of Cambridge
Free surface penetration by plunging vater gives rise to
gas entrainment and transport of bubbles in the submerged flows.
The phenomenon is found in a variety of practical circumstances.
In chemical engineering it is exploited as a convenient method of
dissolving gases in liquids. In environmental engineering,weir
flows provide for reoxygenation of rivers and waste-waters. In
oceanography there is much interest in the transport of bubbles
from breaking waves (which appear as spilling or plunging flows
depending on wind conditions and water depth) . Factors affecting
the efficiency of gas transfer into the liquid include entrainment
flux, sizes of bubbles, their ntsid«/ice, times and dispersion
throughout the liquid. Our studies, outlined below, e^m. particular/y
relevant to the last of these.
Using still photographs and high speed movie film, we
have observed that submerged plunging flovs transport bubbles in
discrete clusters. We attribute this behaviour to bubble
entrapment by large transient vortices travelling in ihe free
shear layer between the submerged flow and surrounding bubbly
The existence of such vortices is HfeM-documented —
they are often referred to as large eddies or 'coherent structures'.
their inhibjtlnr 1'nfluSWe on bubble dispersion has^not
-------
Thomas et al.
Page 2
been videly appreciated and our paper will help to fill this gap
in understanding.
We shall describe elements of a recently developed
general theory of bubble motion in unsteady, non-uniform^high
Reynolds number flow. The theory is based on inviscid analyses
of the forces arising from pressure gradient, acceleration and
vorticity-lift, together with a simple drag law. Errors in
previous formulae for interfacial forces will be identified.
Having introduced "the general theory, we shall concentrate on the
practically relevant special case of bubble motion in line vortex
flow. A laboratoratory experiment will be described which
quantitatively confirms the theoretical predictions of bubble
entrapment:.
We shall report on recent progress towards a computer
simulation of the bubbly free shear layer. Our simulation
makes use of a discrete vortex method to model the fluctuating
large eddy structure; bubble trajectories are calculated using the
new general theory. Exploratory tests to date show indications
of bubble entrapment by transient vorticity concentrations. We
wo^/ld Kofe. to present results of more detailed calculations now in
progress. Also measurements of local gas and liquid flows in a
controlled experiment on the bubbly downflowing free shear layer.
-------
Top and Clarke
ANOMALOUS NEON-HELIUM RATIOS
IN THE ARCTIC OCEAN
Zafer Top, Tritium Laboratory, University of Miami, Miami, Florida
W. Brian Clarke, Physics Department, McMaster University, Hamiltoni Ontario
Abstract; Dissolved rare-gas studies in the ocean have become increasingly
important in the last decade. The subjects of the present study, oceanic
helium and neo^have the atmosphere as their major source. The heavier
and more abundant helium isotope "He, is sometimes augmented at the
sea-floor due to the injection of a sedimentary-crustal (radiogenic)
component. Both helium and neon are often found in the ocean in excess
of their solubility equilibrium values due to the forced dissolution of
air bubbles in wave action (air injection) in the upper layers. Because
there is apparently no mechanism to alter the neon concentrations in the
ocean, neon can be used to strip off the air injected helium component. In
a recent study of the Baffin Bay (Top et al., 1980) however, this procedure
gave rise to negative helium excesses! What seemed to be a suspicion then,
of the constancy of the neon concentration became a certainty in the later
studies at two Arctic locations. The data confirming the anomalous He/Ne
ratios are presented. The source of the anomaly appears to be at the
surface. The ice formation is thought to be primarily responsible for
the relative enrichment of neon. As well, differential diffusion of neon
and helium in ice, and ice melting are discussed as potential contributors
to the observed effect. The great potential of Ne/He ratios in the
Arctic Regions (and probably Antarctic) as a tracer for the newly formed
bottom water is proposed.
Ref: Z. Top, W.B. Clarke, W. Eismont and E.P. Jones. (1980) Radiogenic
Helium in Baffin Bay bottom water. J. Mar. Res., 38, 435-452.
-------
Topalian et al
On the Exchange Rate of Organic Gases
Between Air and Falling Water Drops
J. H. Topalian, S. Mitra, and H. R. Pruppacher
Cloud Physics Laboratory, Dept. of Atmospheric Sciences
University of California, Los Angeles
Abstract
The exchange rate of organic gases between falling water drops and
air has been studied as a function of drop size and concentration in the
gas phase. The study was carried out using the UCLA Rain-Shaft, a unique
facility consisting of a 35 meter high and 50 cm diameter shaft which
allows a quantitative determination of the amount of gas scavenged by
water drops after reaching their terminal velocity. Following a success-
ful study on the exchange rate of SO- between air and falling water drops
(Walcek et al., 1981: J. Atmos. Sci., 38, 871-876) using the same facility,
we determined the rate of absorption of the organic gases: acetaldehyde,
formaldehyde, methylene chloride and chloroform by water drops of 300 to
2500 ym equivalent radius, falling at terminal velocity in air as a function
of gas concentration. The concentration of the organic gas in the gas phase
and in the water were determined using gas chromatographic techniques. The
results of our experimental study are compared with and discussed in
light of the results derived from our theoretical model for forced convec-
tive gas absorption by falling water drops which exhibit internal circulation,
-------
A New Optical Bubble Measuring Device
Wais et al,
T. Wais, M. Barabas, B. Jaehne
Institut fur Umweltphysik, Heidelberg University
Western Germany
Encouraged by the good experience with optical wave measuring systems
we developed an optical method to measure bubbles in the water.
t
A He-Ne-Laser beam pierces the water surface vertically from below.
The light scattered by bubbles crossing the beam is collected in 4
detectors at different depths ranging from 0.1 - 0.5 m. Two sets of
this arrangement are used with differnt laser beam diameters (1 and 5
mm) for the following reasons. Firstly, the dynamics in bubble size
and even more in intensity detected beeing proportional to the square
of the bubble radius is too large for only one detector. Secondly, the
large bubbles are rare if compared with small ones, and a larger
sampling volume is therefore necessary to obtain good statistics.
The first measurements with this device are carried out in the large
wind/wave facility of the I.M.S.T Marseille, simultaneously with He-
and Rn gas exchange experiments. We obtained bubble spectra at four
water dephts in the wind speed range from 11 to 13.8 m/s. First
bubbles are observed at 12 m/s. The spectras are compared with other
wind tunnel results. The total number of bubbles is about a factor 3
smaller than in the wind tunnel of the Sonderforschungsbereich
Meeresforschung, Hamburg.
With a simple model we estimated that even at the highest possible
wind speed of the large facility (13.8 m/s) no noticeable enhancement
« 10t) even of the He gas exchange is possible.
-------
Waldichuk
LABORATORY OBSERVATIONS ON TRANSFER OF ATMOSPHERIC OXYGEN
INTO STRATIFIED SEA WATER
By Michael Waldichuk
West Vancouver Laboratory
Department of Fisheries and Oceans
4160 Marine Drive
West Vancouver, B.C., Canada, V7V 1N6
ABSTRACT
Laboratory sea later las stripped of most of its dissolved oxygen by
bubbling nitrogen through it in a column filled vath marbles. The sea vater,
starting lath dissolved oxygen concentrations of 0.5 to 1.0 mg/L, las held in
a round plastic container, 56 cm high and 34 cm in diameter at the top, in a
later table vith sea later 14 cm deep at a temperature of 12~18°C. Water
samples rare taken at the surface, 10, 25, 40 cm and at the bottom of the
container periodically for dissolved oxygen determinations. Within 4 days,
dissolved oxygen concentrations reached 7 mg/L at all depths, tfien only sea
rater las present; but then a 2-cm layer of fresh later covered the sea
later, dissolved oxygen in the bottom half of the container, starting at 1.0
mg/L, ranged from 0.3 to 0.6 mg/L after 5 days. In deoxygenated sea later
(0.5 mg/L), covered lith a 1-mm layer of South Louisiana crude oil, dissolved
oxygen las near 1 mg/L at the bottom of the container, 3 mg/L at 25 cm depth
and 5 mg/L at 10 cm and at the surface after 5 days. Results suggest that
stratification ilth fresh later can be at least as effective as a thin layer
of oil in retarding the transfer of atmospheric oxygen dowivard in sea
later. They are in line also lith observations of vertical dissolved oxygen
distributions found in deep, partially-restricted marine inlets receiving
substantial fresh later runoff.
-------
Warhaft and Bolgiano
Moisture and Heat Transport in a Stably
Stratified Boundary Layer Over a Water Surface
Z. Warhaft and R. Bolgiano Jr.
Cornell University
Ithaca, NY 14850
A laboratory experiment in which a stably stratified boundary layer was
formed by blowing warm, dry air over a water surface is described. The exper-
iment was carried out in the air-sea interaction facility of d'Institute de
Mechanique Statistique de la Turbulence (I.M.S.T.), Marseille. Velocity, tem-
perature and humidity fluctuations were measured by means of hot (and cold)
wires and a microwave refractometer. In order to obtain high bulk Richardson
numbers, Ri'b, [ = (g/PQ) (ApL/U2) where Ap is the density difference
across the depth L of the boundary layer, U is the free stream velocity, g is
the gravitational acceleration and p is a reference density] low wind
speeds (£ 0.7 m/s) were used. Under these conditions, the water surface re-
mained smooth. Particular attention was given to a situation in which the
Ri. was so strong(~0.4) that the boundary layer turbulence, tripped by
means of vortex generators at the entrance to the tunnel, was rapidly suppres-
sed. However as the flow evolved downstream and the boundary layer thickened
(and hence Ri. diminished), bursts of turbulence occurred close to the
water surface, causing rippling of the water. Most of the momentum, heat and
moisture transport occurred during these bursts. Classified averaging of the
momentum, heat and moisture flux formed from the time series of the velocity,
temperature and humidity fluctuations showed remarkable similarity to the
intermittency observed in the neutral boundary layer (at a much smaller scale)
by previous workers. The results also have similarities to bursting observed
in river estuaries and in the sea.
-------
Whittemore
An Assessment of Tracer Techniques for
Measuring Reaeration Rates in
Large River Systems
By: Raymond C. Whittemore, PhD
National Council of the Paper
Industry for Air and Stream Improvement, Inc.
Tufts University, Medford, MA 01890
NCASI had the opportunity to investigate tracer techniques
for the estimation of stream reaeration rates during its study
of mathematical water quality modeling on the Quachita River
basin. Data was first collected which allowed for the exam-
ination of the variablity of measurements of stream reaeration
rates which were subsequently used to assess uncertainty
in the calibration and verification of four water quality
models. This study was followed by a comparison of hydrocarbon
and radiotracer techniques on the same river reach„
The major conclusions from these studies were two fold:
(a) The precision of the radiotracer measurement was
directly related to the amount of gas lost during the experi-
ment. Thus, the precision of the measurement is ultimately
limited by an ability to follow the dye and conservative
tracer, the precision of the krypton to oxygen transfer ratio,
and the temperature correction factor.
(b) The hydrocarbon tracer techniques produced reaeration
rates higher than that obtained by the radiotracer technique
and selected empirical equations. This difference was two
large to be accounted for by variability in the radiotracer
method.
-------
Wilcock
METHYL CHLORIDE AS A GAS-TRACER IN REAERATION
STUDIES ON SOME NEW ZEALAND RIVERS
Robert J. Wilcock
Water and Soil Science Centre,
Ministry of Works and Development,
Private Bag, Hamilton, New Zealand.
Many New Zealand rivers receive waste inputs with significant oxygen demands
from sources such as dairy factories and wood-pulp mills. In order to have
effective models of dissolved oxygen profiles in these rivers it is necessary
to have accurate estimates of the reaeration coefficient, K_. This paper
describes a gas-tracer technique using methyl chloride, that has been
developed and successfully tested in several rivers and streams. Methyl
chloride is a stable substance not found in significant background concentrations
-12
in New Zealand waters, that can be detected down to 1 x 10 g by a gas-
chromatographic technique. In this study, values of K« and an analagous
gas transfer coefficient for methyl chloride, K „ _.., were measured simultaneously
in laboratory experiments over the temperature range 5 - 35 C. The results
indicated a temperature dependence for the transfer coefficient ratio, K _/K_,
Cn_ul /
which was fitted empirically to an expression based on the absolute rate model
for diffusion, giving
KCH Cl
— == — = 4.323 exp(-530.97/T)
K2
where T is the absolute temperature.
-------
Wilcock
Page 2
Results are given of applications of the methyl chloride technique in several
rivers, in which rhodamine-WT was used as a conservative tracer to correct
for dispersion and dilution. The measured reaeration coefficients range from
0.2 to 16.6 day" (base e) and are compared with values calculated using the
O'Connor-Dobbins formula. The experimental results correlate well with the
calculated values but are generally about 50% greater than the predicted
— 1
values of K . It is concluded that for K values greater than about 1 day
the methyl chloride method gives values having a much better precision than
those calculated from empirical expressions, and provide estimates of K«
with an uncertainty of about + 10%. The major source of error in the tracer
technique appears to be associated with estimating peak concentrations of
the dissolved methyl chloride. Empirical expressions can give rise to
errors of + 50% in their estimates of K_ in streams having irregular cross-
sections, or where estimating mean depths is difficult. Rivers in which K
was 1 day or less had comparable uncertainties between measured and
calculated values of the reaeration coefficient, because of the small
change in the ratio of methyl chloride : rhodamine-WT peak concentrations,
with distance downstream.
-------
Yu
WIND EFFECT ON AIR-WATER OXYGEN TRANSFER
IN A LAKE
by
Shaw L. Yu
University of Virginia
Charlottesville, Virginia
The paper will report results obtained from field experiments conducted
in a lake in Northern New Jersey to investigate the effect of wind on the
air-water oxygen transfer process. Mass transfer and boundary layer theories
were utilized to formulate theoretical prediction models for the reaeration
coefficient, K_.
Parameters considered in model formulation include wind velocity, wind
shear velocity, effective mass diffusivity, and water depth. Equations of
linear, nonlinear, and dimensionless forms were examined. Statistical methods
such as analysis of variance and residual analysis were employed to determine
model adequacy-
Results indicate that wind speed is a limiting factor in affecting lake
reaeration. Below a "critical" wind speed (4.2 m/s at 10 m) , there is no
significant wind effect on K_. Above another "critical" wind speed (6.5 m/s
at 10 m), K9 varies nonlinearly with wind speed. Between these two limits, K
assumes a fairly linear relationship with wind speed, as shown in the
following dimensionless equation:
K9H2 . /UH \ 1.136
— 1.63 x 10 f —y
v v '
where K_ = reaeration coefficient in sec , H = water depth in meters, v =
2
water viscosity in m /s, and U = wind velocity at 10 meters in m/s.
An investigation was made later to test some of the proposed reaeration
equations using data collected for the Occoquan Reservoir in Northern
-------
Yu
Page 2
Virginia. Two years of water temperature and dissolved oxygen data taken at
various depths and at six stations throughout the lake were available for the
analysis. Wind recorders were obtained from two locations in the vicinity of
the lake.
The Occoquan data indicate that the wind is responsible for keeping
dissolved oxygen levels near or above saturation for the upper 1 to 3 meters
of the lake even though the lake is strongly stratified and dissolved oxygen
levels dip to zero near the bottom of the lake. The proposed lake reaeration
equation was also found to be adequate in predicting surface reaeration
coefficient.
-------