Conference PaperPDF Available

Evaporation reduction from farm dams using air-bubble plume destratification

Page 1 of 8
Evaporation reduction from farm dams using air-bubble plume
Fernanda Helfer1,2, Hong Zhang1, Charles Lemckert1
1Griffith School of Engineering, Griffith University, Gold Coast, Queensland, Australia
2Corresponding author,
Abstract: The objective of this paper is to assess the effectiveness of destratification by air-
bubble plumes in reducing evaporation from shallow and deep dams in Australia, based on
modelling. DYRESM was applied to model the thermodynamics and evaporation from a farm
dam in Queensland under different destratification conditions and two different depths: 6.5 and
13.5 metres. The results showed that destratification can reduce surface temperature and
evaporation from both shallow and deep waters, but higher effectiveness is expected in deep
lakes. This occurs because the bottom water of a deep lake is colder than the surface water during
almost the entire year, which is a condition necessary for reducing the surface water temperature
and, consequently evaporation rates, under destratification conditions. However, the reduction in
surface water temperature and evaporation will not be significant during hot weather periods (that
is, when the difference between the surface water temperature and the air temperature is high).
Under this condition, the cooled surface water will be rapidly heated and the destratification
system will only contribute to further increasing the heat stored in the water.
Keywords: evaporation, DYRESM, stratification, water resources
The evaporation of water from farm dams is
seen as one of the most significant
contributors to loss of water in the rural areas
of Australia. The volume of water stored in
farm dams accounts for 9% (around 7,000
GL) of the total water stored in the country;
however, it is estimated that 40% of this
volume is lost every year due to the high
rates of evaporation (Craig et al., 2005). In
fact, evaporation can often exceed 2000 mm
per year in most agricultural areas in
Australia (Department of Natural Resources
and Mines, 2005). As a consequence, the
availability of water for crops decreases,
affecting productivity and farmer income.
For a considerable period, Australia has been
developing and studying different
mechanisms for reducing evaporation from
farm dams. Most of the techniques however,
have been shown either not to be effective, as
in the example of windbreaks (Helfer et al.,
2009); to be excessively expensive, for
example, physical covers (Yao et al., 2010)
and modifying the dam’s shape to minimize
surface area (McJannet et al., 2008a); or to
impose potential risks to the water quality,
for example, the use of chemical and physical
covers (McJannet et al., 2008b, Yao et al.,
Destratification by air-bubble plumes is one
technique that deserves further investigation,
as it has been suggested in literature as a
potential mechanism for reducing
evaporation (Koberg and Ford, 1965 and
Helfer et al. 2011). The primary aim of
artificial destratification is to maintain or
improve the quality of the reservoir water.
The potential of the technique in controlling
evaporative losses is related to the change in
water temperature brought about by the
mixing device. The principle is that heavy
hypolimnion fluid is lifted by the air injected
at the bottom of the lake. At the surface, this
water mixes with lighter epilimnion water,
reducing the temperature, and consequently,
Page 2 of 8
evaporation rates. Based on this idea, the
technique may only be effective in lakes that
experience thermal stratification, since there
has to be a difference between the
temperatures of the bottom and the surface
waters. However, to date, few studies have
focused on the use of air-bubble plumes to
reduce evaporation from dams. In a previous
publication we have applied the one-
dimensional model DYRESM (Imberger and
Patterson, 1981) to simulate temperature
stratification and mixing dynamics of an
Australian shallow lake under artificial
destratification conditions (Helfer et al.
2011). The results indicated that the
technique, although not effective for shallow
lakes, may be effective for deep lakes due to
the higher difference between surface and
bottom water temperatures. The studied
shallow reservoir stratifies in summer if low
wind conditions dominate, but also mixes
rapidly under strong wind gusts. For the
destratification system to be effective in
reducing evaporation at a rate that the
aeration system can be feasibly maintained,
the period of natural thermal stratification has
to be longer and not interspersed with periods
of well mixed profile, a pattern that is usually
observed in deep lakes.
Therefore, in this study DYRESM is applied
to a deeper stratified lake located in
Queensland to assess the effectiveness of
destratification by air-bubble plumes in
reducing evaporation and to compare the
results with the outcomes from the previous
study (shallow lake).
DYRESM is a 1-D model used for the
prediction of the vertical distribution of
temperature, salinity and density in reservoirs
of medium size. The development of the
model dates back to 1978 with the first
successful application on Wellington
Reservoir in Australia (Imberger et al., 1978;
Imberger and Patterson, 1981). It has since
been applied partly with modified codes to
many other reservoirs and lakes (e.g.
Patterson et al., 1984; Ivey and Patterson,
1984; Hamilton and Schladow, 1997;
Moshfeghi et al., 2005; Hipsey, 2006; Helfer
et al., 2011).
The model requires little calibration due to
validation from numerous simulations and
the inclusion of physical limnetic parameters
derived from field and laboratory studies
(Hamilton and Schladow, 1997). The main
input data in DYRESM are the depth-area
relationship of the studied lake, daily or sub-
daily meteorological data (including incident
long and short wave radiation, rainfall, air
temperature, humidity and wind speed), daily
discharge, temperature and salinity of
inflows, daily discharge of outflows and
initial profiles of temperature and salinity.
The model is based on a Lagrangian layer
scheme in which the reservoir is represented
by a series of adjoining horizontal layers of
uniform property that vary in thickness and
number. This is opposed to the fixed grid
approach used in other lake models in which
differential equations are solved numerically
on the mesh points. In DYRESM, at each
time step, as inflows and outflows enter or
leave the reservoir, the affected layers
expand or contract and those above move up
or down to accommodate the volume change.
Mixing and surface layer deepening are
modelled by amalgamation of layers, based
on a criterion of available turbulent kinetic
energy and required potential energy.
Diffusion in the meta and hypolimnion is
simulated by transferring volume fractions
from one layer to the overlying layer, where
the fraction is determined by the Lake
When combining two layers, the
conservation laws for temperature, salt,
energy and momentum can be generalised as:
Page 3 of 8
where the subscripts refer to layer indices, M
is the mass of water and C is the property
being conserved (energy, salt or momentum).
For conservation of temperature, the above
assumes the specific heat to be constant.
The surface heat, mass and momentum
exchanges comprise the primary driving
mechanisms for DYRESM. The surface
exchanges include heating due to short-wave
radiation penetration into the lake and the
fluxes at the surface due to evaporation,
sensible heat, long-wave radiation and wind
stress. The sensible and latent heat fluxes are
described by bulk aerodynamic formulae.
The evaporative flux is calculated as:
( )
saEa eeUC
E= λρ
where E is the latent heat flux, P is the
atmospheric pressure, ρa is the air density, λ
is the latent heat of evaporation, CE is the
bulk aerodynamic transfer coefficient
(=1.3E-3, Fischer et al., 1979), U is the wind
speed, ea is the vapour pressure in the air and
es is the saturation vapour pressure, which is
a function of the surface water temperature.
Energy transferred according to this equation
is always added or subtracted to the
uppermost layer.
DYRESM also incorporates a sub-routine to
model water mixing by destratification
devices, such as impellers and air diffusers.
For air diffusers, the algorithm is based on a
simple, single core plume whose motion is
determined from three differential equations
of conservation of mass, momentum and
buoyancy (Patterson and Imberger, 1989).
This sub-routine requires a number of
destratification devices operating in the
reservoir, device type (that is, air diffuser or
impeller), draft tube length and diameter (for
impellers), height and number of ports (for
diffusers), volume flow rate of air (for
diffusers) and volume of water (for
The model DYRESM was applied to a large
farm dam located in Queensland, Australia to
study the effects of artificial destratification
on evaporation. Logan’s Dam (27o34’26’’S,
152o20’26’’E, Fig. 1) has a storage capacity
of 0.7GL, a full storage surface area of
approximately 17 hectares and a maximum
depth of 6.5m. All data sets, including
meteorological, morphological and outflow,
were provided by the South-East Queensland
Urban Water Security Research Alliance,
which has been monitoring the dam since
August 2009 for evaporation study purposes.
Mean daily values for all inputs were adopted
generating daily outputs at midday. The
period chosen for the simulations, based on
available and consistent data, was from
29/09/2009 to 11/04/2010 (195 days).
Fig. 1. Aerial view of the selected lake (Source:
Google Earth)
Scenarios: The shallow lake scenario
represented the real situation, in which the
maximum water depth is 6.5m, and for which
the model was calibrated (Helfer et al.,
2011). A hypothetical situation, with
maximum water depth equal to 13.5m
composed the deep lake scenario. The
baseline scenarios for both situations
(shallow and deep lake) consisted in the
absence of any destratification system
operating in the dam. Then, scenarios with a
destratification system operating
continuously throughout the 195 days of
Page 4 of 8
study were simulated. For both shallow and
deep lakes, the destratification system
consisted of one air diffuser laid at the
bottom of the lake. The air-flow rates
injected through this diffuser into the water
varied from 0.003 to 0.8m3/s, and the number
of ports on each diffuser varied from 10 to
960. A total of 128 simulations were
performed and evaporation and water
temperature were noted for analysis as
described below.
Baseline water temperature: The water
temperature profiles for the baseline
scenarios for both shallow and deep lake
cases are presented in Figure 2. For shallow
water, Logan’s Dam experienced thermal
stratification from the beginning of
November until the end of the period of
study, interspersed with some periods of
weak stratification (Fig. 2a). For deep water,
thermal stratification occurred during the
whole period with no periods of well mixed
water. For both shallow and deep lakes, the
average surface temperature was 26.2oC,
varying from a minimum of 18.6oC at the
beginning of the simulation period (end of
September), to a maximum of 30.8oC at the
end of January. The strongest thermal
gradients were observed at the beginning of
December, with drops of up to 2.3oC/m in the
metalimnion for shallow water, and at the
end of January, with drops of up to 2.7oC/m
in the metalimnion for deep water.
Baseline evaporation: The baseline
evaporation rates from both shallow and deep
lakes for the whole period of simulation (195
days) are presented in Figure 3. The daily
evaporation for both scenarios was very
similar over the period of simulation. The
daily rates varied from 1.0mm on 08/03/2010
to 11mm on 19/01/2010. Total evaporation
was 897mm for the shallow lake scenario and
901mm for the deep lake scenario. The
similarities in surface temperature and
evaporation between the two lakes are due to
same boundary conditions on both scenarios.
a)Water temperature (oC) for maximum water depth = 6.5m
Depth (m)
29-Sep-2009 02-Dec-2009 04-Feb-2010 11-Apr-2010
b)Water temperature (oC) for maximum water depth = 12.5m
Depth (m)
29-Sep-2009 02-Dec-2009 04-Feb-2010 11-Apr-2010
Fig. 2. Simulated water temperature profiles for
shallow and deep lakes (29/09/2009 - 11/04/2010)
Modelled evaporation from Logan’s Dam
Evaporation (mm)
Shallow Lake
Deep Lake
29-Sep-2009 02-Dec-2009 04-Feb-2010 11-Apr-2010
Fig. 3. Baseline daily evaporation from shallow and
deep lakes from 29/09/2009 to 11/04/2010 – modelled
Evaporation and temperature under
destratification conditions shallow lake
scenario: The total evaporation for the whole
period of simulation for the shallow lake
scenario is presented in Figure 4. The results
are for one diffuser operating in the dam with
number of ports varying from 10 to 960. The
total air-flow rate pumped through the
diffuser into the water (x axis) varied from
Page 5 of 8
0.003 to 0.8m3/s at the ambient pressure at
the level of the diffuser. The results show
that as the air-flow rate increases,
evaporation rate reduces, and that
evaporation also reduces as the number of
ports increases; however, for this shallow
lake scenario, the reduction in evaporation
reaches a maximum of only 1.0%, for 960
ports and for air-flow rate above 0.4m3/s.
Total evaporation under aeration conditions – shallow lake
Evaporation (mm)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0. 8
Evaporation, mm
10 ports
30 ports
60 ports
120 ports
240 ports
480 ports
960 ports
Air-flow rate (m3/s)
Fig. 4. Total evaporation from shallow lake from
29/09/2009 to 11/04/2010 under destratification
conditions – modelled data
Figure 5 shows the water temperature profile
of the shallow lake after destratification for
the whole period of simulation. The scenario
shown in Figure 5 is for the strongest
aeration case (960 ports and air-flow rate =
0.8m3/s). Although the surface mixed layer,
defined up to the level where the water
temperature is less than 95% of that of the
surface, deepens from an average value of
2.5m (baseline) to 4.4m, the temperature of
the surface remains practically unchanged.
For a lower number of ports and lower air-
flow rate per port, the mixed layer after
destratification is little affected (e.g. 2.5m for
10 ports and air-flow rate = 0.003m3/s, 2.7m
for 10 ports and air-flow rate = 0.8m3/s and
2.9m for 960 ports and air-flow rate =
0.003m3/s). The average buoyancy frequency
for an equivalent linear profile (Lemckert and
Imberger, 1993) for the whole period is
0.028s-1 in the baseline scenario and drops to
0.017s-1 after aeration, indicating that mixing
has taken place effectively. However, the
average surface temperature in the strong
aeration case reduces from 26.2oC (baseline
temperature) to only 26oC. For the other
scenarios, the surface temperature remains
unchanged after destratification. This means
that, although the destratification system
effectively brings the cold bottom water to
the surface, this is quickly heated by the sun.
As a consequence, the total heat stored in the
dam is increased during this process. This is
confirmed by analysing the average bottom
temperature, which increases from 23.7oC in
the baseline scenario to 25.3oC in the strong
aerated scenario.
Temperature (oC) under strong aeration conditions – shallow lake
Depth (m)
29-Sep-2009 02-Dec-2009 04-Feb-
Fig. 5. Simulated water temperature profile for
shallow lake (maximum depth = 6.5m) from
29/09/2009 to 11/04/2010 after strong aeration
Also, it is important to note that the reduction
in surface temperature is only observed in
those days when the bottom temperature is
much colder than the surface. Figure 6
illustrates that the highest reductions in
surface temperature after strong aeration
occur at the beginning of the simulation
period, in the middle to end of November and
at the beginning of March, exactly when the
highest reductions in evaporation were also
observed. This suggests that if the difference
in temperature between bottom and surface
waters was significant during the entire
period of study, for example,. more than 4oC,
the reduction in evaporation after
destratification would be higher. This
Page 6 of 8
hypothesis is analysed in the following
Fig. 6. Temperature difference between bottom and
surface waters for the shallow water lake
Evaporation and temperature under
destratification conditions deep lake
scenario: The evaporation rates under
destratification conditions for the deep lake
case are presented in Figure 7. The
evaporation trend is similar to that observed
in the shallow lake scenario. However, for
the deep lake, the reductions are higher. The
maximum drop is from 901mm (baseline
evaporation) to 852mm, representing a
reduction of 5.4% (compared to 1% for the
shallow lake).
Total evaporation under aeration conditions – deep lake
Evaporation (mm)
0 0.1 0.2 0.3 0.4 0. 5 0.6 0.7 0.8
10 ports
30 ports
60 ports
120 ports
240 ports
480 ports
960 ports
Air-flow rate (m3/s)
Fig. 7. Total evaporation from deep lake from
29/09/2009 to 11/04/2010 under destratification
conditions – modelled data
The water temperature profile for the deep
lake after strong aeration is shown in Figure
8. The average surface temperature reduces
from 26.2oC (baseline temperature) to
25.7oC, a slightly higher reduction than that
observed in the shallow lake scenario. The
surface mixed layer deepens from an average
of 2.3m to 10.7m, with the average water
depth during the whole period being 12.4m.
The buoyancy frequency drops from an
average of 0.039s-1 in the baseline scenario to
0.013s-1 in the strong aeration scenario,
indicating that the mixing has taken place
and mixing efficiency is higher than in the
shallow lake scenario. However, the surface
temperature in the deep lake also remains
unchanged and the total heat stored in the
dam also increases. In this scenario, the
bottom water temperature increases from an
average of 18oC (baseline scenario) to 25oC
after destratification.
Temperature (oC) under strong aeration conditions – deep lake
Depth (m)
29-Sep-2009 02-Dec-2009 04-Feb-2010 11-Apr-2010
Fig. 8. Simulated water temperature profile for deep
lake (maximum depth = 13.5m) from 29/09/2009 to
11/04/2010 after strong aeration
The reduction in surface temperature after
strong aeration (Fig. 9) is higher at the
beginning of the simulation period, extending
to the beginning of December. Higher
reductions in surface temperature after
destratification are positively correlated with
the difference between bottom and surface
temperatures during this period. From the
middle of December until mid-February,
despite the high difference between surface
and bottom temperatures, the reduction in
surface temperature after destratification is
not significant. This can be explained by the
difference between the air and water surface
temperatures after destratification during this
time. The cold water is brought from the
Page 7 of 8
bottom to the surface by the destratification
system. Heat exchanges at the surface will be
higher, because of the difference between
surface and air temperatures. The higher this
difference, the higher the input of heat into
the water due to sensible heat flux. This
theory indicates that a destratification system
would be more efficient in cooling the
surface water, and keeping it cool, during
mild weather conditions. This would be when
the surface temperature (resulting from the
mixing between the cold water brought to the
surface and the warm water from the surface)
would be in equilibrium with the air
temperature. A situation like this, however,
would only be observed during short periods
in a year, during spring for instance.
Therefore, the installation of a
destratification system would not be
justifiable, although this theory would require
further investigation.
Fig. 9. Temperature difference between bottom and
surface waters for the deep water lake
Discussion and Conclusions
The efficiency of continuous destratification
by air-bubble plumes in reducing surface
temperature, and consequently evaporation
from open waters, is a function of the
occurrence of a difference in temperature
between bottom and surface waters. The
results from this study also indicate that this
efficiency will be more noticeable during
mild meteorological conditions. Under the
existence of a significant difference between
bottom and surface water temperatures in
association with cool weather, the technique
successfully reduces surface temperature.
The surface temperature is then kept cool
because of the weather, and evaporation rates
reduce significantly. As the weather becomes
hot, the surface water that contains the cold
water brought up by the destratification
system is rapidly heated due to sensible heat
flux, and the destratification system simply
contributes to increasing the heat stored
along the depth of the reservoir even more.
This behaviour was observed for both
scenarios considered in this study, shallow
and deep lakes. The main difference is that
for the shallow lake, the heat added at the
surface is slightly higher than in the deep
lake case. This happens because the bottom
water in a deep lake is always, or almost
always, colder than the surface. This fact
may reduce evaporation from deep, stratified
reservoirs, but the effectiveness, as explained
before, will be minor if the air temperature is
constantly hot (that is, during summer).
Concerning the model used, DYRESM has
been shown to be a potential tool for
predicting the change in water temperature
and estimating evaporation rates from open
waters under destratification conditions. A
study to evaluate the seasonal effectiveness
of destratification systems in reducing lake
surface temperature and evaporation is
recommended for a better understanding of
artificial destratification in reservoirs.
Cooley, K.R. (1983). Evaporation reduction:
summary of long-term tank studies. J. Irrig.
Drain. ASCE, 109, 89-98.
Craig, I., Green, A., Scobie, M. & Schmidt,
E. (2005). Controlling evaporation loss
from water storages. Publication 1000580/1
for the National Centre for Engineering in
Agriculture, Toowoomba.
Department of Natural Resources and Mines
(2005). Australian synthetic daily Class A
pan evaporation. Report for the Managing
Climate Variability Program, Brisbane.
Page 8 of 8
Fisher, H. B., List, E. J., Koh, R. Y. C.,
Imberger, J. & Brooks, N. H. (1979).
Mixing in inland and coastal waters.
Academic Press, New York.
Hamilton, D. & Schladow, S. (1997)
Prediction of water quality in lakes and
reservoirs. Part 1 - model description. Ecol.
Model., 96, 91-110.
Helfer, F., Zhang, H. & Lemckert, C. (2009).
Evaporation reduction by windbreaks:
overview, modelling and efficiency.
Technical Report no. 16 for the Urban
Water Security Research Alliance,
Helfer, F., Lemckert, C. & Zhang, H. (2011).
Assessing the effectiveness of air-bubble
plume aeration in reducing evaporation
from farm dams in Australia using
modelling. In: 6th Water Resources
Management Conference, 23- 25 May 2011,
Riverside, USA.
Hipsey, M. (2006). Numerical investigation
into the significance of night time
evaporation from irrigation farm dams
across Australia (final report). Report
prepared for the Land & Water Australia,
Centre for Water Research, The University
of Western Australia.
Imberger, J., Patterson, J.C., Hebbert, B. &
Loh, I. (1978) Dynamics of reservoir of
medium size. J. Hydraul. Div. ASCE, 104,
Imberger, J. & Patterson, J.C. (1981). A
dynamic reservoir simulation model -
DYRESM. In Fischer, H.B. (Ed.) Transport
Models for Inland and Coastal Waters. New
York, Academic Press.
Ivey, G.N. & Patterson, J.C. (1984). A model
of the vertical mixing in Lake Erie in
summer. Limnol. Oceanogr., 29, 553-563.
Koberg, G.E. & Ford, M.E. (1965).
Elimination of thermal stratification in
reservoirs and resulting benefits. Geol.
Surv. Water Supply Pap. 1809-M,
Washington DC.
Lemckert, C. J. & Imberger, J. (1993).
Energetic bubble plumes in arbitrary
stratification. J. Hydraul. Eng., 119, 680-
Lewis, D.M. (2004). Surface mixers for
destratification and management of
Anabaena circinalis. University of
Adelaide. (PhD Thesis).
Mcjannet, D., Cook, F. & Burn, S. (2008a).
Evaporation reduction by manipulation of
surface area to volume ratios: overview,
analysis and effectiveness. Technical Report
8 for the Urban Water Security Research
Alliance, Brisbane.
Mcjannet, D., Cook, F., Hartcher, M.,
Knight, J. & Burn, S. (2008b). Evaporation
reduction by monolayers: overview,
modelling and effectiveness. Technical
Report 6 for the Urban Water Security
Research Alliance, Brisbane.
Moshfeghi, H., Etemad-Shahidi, A. &
Imberger, J. (2005). Modelling of bubble
plume destratification using DYRESM. J.
Water Supply Res. T., 54, 37-46.
Patterson, J.C., Hamblin, P.F. & Imberger, J.
(1984). Classification and dynamic
simulation of the vertical density structure
of lakes. Limnol. Oceanogr., 29, 845-861.
Patterson, J.C. & Imberger, J. (1989).
Simulation of bubble plume destratification
systems in reservoirs. Aquat. Sci., 51, 3-18.
Yao, X., Zhang, H., Lemckert, C., Brook, A.
& Schouten, P. (2010). Evaporation
reduction by suspended and floating covers:
overview, modelling and efficiency,
Technical Report No. 28 for the Urban
Water Security Research Alliance,
We thank the Urban Water Security Research
Alliance and CSIRO for providing us with the
meteorological data. Funding for this project was
provided by Griffith University Postgraduate
Research School, the Australian Government –
Department of Innovation, Industry, Science and
Research, the Griffith School of Engineering and
the Urban Water Research Security Alliance.
... Although the first attempts of artificial destratification aimed at improving the water quality, the potential of this technique in controlling evaporative losses due to thermal destratification is now well known. In fact, induced mixing of water layers can lift up the colder hypolimnion layers and mix it with warmer epilimnion where strongly controls evaporative losses (Helfer et al., 2012). In 1999, Cox (1999 investigated the effect of thermal mixing of water layers on evaporation rate of open reservoirs in Cyprus. ...
... Sherman et al. (2010) performed a numerical analysis on the effect of thermal mixing of layers on evaporation reduction of open reservoirs and showed that the artificial destratification has an acceptable performance in reducing the evaporation rate of deep storages. The effectiveness of destratification by air-bubble plumes in reducing evaporation from shallow and deep reservoirs was assessed by Helfer et al. (2012). The results showed that destratification can decrease surface temperature and thus evaporation from both deep and shallow water bodies with a rather higher impact on deep reservoirs. ...
The need for protection of freshwater resources with population growth and frequent droughts has become an important issue, especially in arid and semiarid regions. Investigations show that evaporation is the reason of significant losses in water reservoirs and thus suppressing it could effectively enhance water storage strategies. Covering reservoirs with modular floating elements, such as floating balls, offers a simple and reliable measure for suppression of evaporative losses. Despite extensive studies on the application of floating covers, its performance in the presence of surface flows while coverage fraction might change due to the cover density has not yet been studied systematically. The present study thus aims to experimentally address the effect of cover fraction of floating balls (induced by changing balls’ density and thus immersion depth) on the evaporation suppression efficiency and potential impacts on the change in the surface energy balance due to surface flows. For this purpose, a small water basin with 1.13m2 surface area and 0.5m depth is fully covered with white and black polyethylene balls with diameter of 70mm and two different masses, i.e., 90g and 127g (providing 0.83 and 0.77 surface cover fraction, respectively). Overall, the white balls showed the highest evaporation suppression efficiency. Also, it is found that with increasing balls’ density so that more than half of the volume sinks into the water, surface coverage and thus suppression efficiency decreases. Interestingly, the results showed that with increasing water surface flow rate, the evaporative loss is first reduced until a certain flow rate (the so-called optimal water flow-rate) and then it increases gradually with associated effects on reservoirs energy balance where cover-water energy exchanges may alter with changing balls density.
... The period chosen for the simulations (based on available and consistent data) was from 29 September 2009 to 20 July 2010 (295 days). The model had been already calibrated for previous studies (Helfer et al., 2011;Helfer et al., 2010a), with a mean absolute error of 0.79 o C; a mean relative error of 2.4%; and a coefficient of determination, R 2 , of 0.93 in comparison with field measurements of water temperature. ...
Full-text available
One of the main concerns regarding water storage in Australia, and other semi-arid countries, is the high rate of evaporation that inevitably leads to significant water loss. In this paper, the use of air-bubble plume systems to reduce evaporation from large reservoirs is assessed. A destratification system was designed for a large dam based on its depth and stratification strength with the intention of destratifying the reservoir in a short time period. The model DYRESM was then used to simulate the water dynamics under destratification conditions. Different strategies for the operation of the aeration system were assessed, from 10-days operation periods at times of high evaporation rates to continuous operation over longer time spans. The modelled water column temperatures and evaporation rates were analysed and it was found that artificial destratification was only effective in reducing evaporation in spring. In summer, heat is added to the water at a rapid rate, and artificial destratification only helps reduce evaporation in the initial days of operation. The effect of artificial destratification in reducing evaporation in autumn depends on the operation of the system during summer. If operated in summer, the rates of evaporation in autumn will increase due to the additional heat added to the water during the summer. In winter, overturn takes place and artificial destratification has no influence on water temperatures and evaporation. It was concluded that aeration by air-bubble plumes would only be effective in reducing evaporation if the hypolimnetic water does not become warm when mixing takes place. This is an ideal situation, but is unlikely to happen in practice.
Full-text available
This paper aims to assess the effectiveness of artificial destratification by airbubble plumes in reducing evaporation from farm dams in Australia. A onedimensional model was applied to simulate the change in water temperature and evaporation rates of a real farm dam in Queensland under aeration conditions. Results show that destratification systems can reduce surface temperature, but the highest reduction in evaporation for the studied reservoir would be only 2.5%. The main conclusion is that it is unlikely that the technique will be feasible for small farm dams, given the high costs involved with the operation of an aeration system and the small quantity of water saved through evaporation reduction. The results also indicate that the technique may be effective for reservoirs that experience long periods of accentuated thermal stratification, such as large, deep dams, in which the mixing process would lead to higher reductions in surface temperature and, consequently, in evaporation rates. Keywords: DYRESM, thermal stratification, water aeration, evaporation, water temperature, destratification. 1 Introduction Australian farm dams’ capacity is approximately 7000 GL (7 x 109 m3), representing 9% of the total water stored across the country. It is estimated however that 40% of the volume stored is lost every year due to the high rates of evaporation in agricultural areas in New South Wales and Queensland (Craig et al. [1]). Extremely high radiation and air vapour deficits make these areas experience evaporation rates that can surpass 2000 mm per year. As a
A study was conducted to find out the consequence of night time evaporation from irrigation farm dams across Australia and to provide quantitative estimates of its significance across a range of climatic regions. The model involved in this study included the hydrodynamic model DYRESM (DYnamic REServoir Simulation Model) to simulate the water balance and thermal structure of the dams and to assess the evaporative fluxes. The three important points, which were taken into consideration included to resolve sub-daily evaporation variability, the effect of evaporation on water temperature to predict evaporative losses, and the influence of atmospheric motions on the evaporative flux. The result projected the importance of meteorological data for the estimation of evaporation flux, and the necessity of nighttime fraction in determining in water balance investigations and during the design of evaporation amelioration strategies.
DYRESM is a one dimensional numerical model for the prediction of temperature and salinity in small to medium sized reservoirs and lakes. The model makes use of a layer concept, in which the reservoir is modelled as a system of horizontal layers of uniform property. These layers move up and down, adjusting their thickness in accordance with the volume-depth relationship, as inflow and withdrawal increase and decrease the reservoir volume. The result is a model which is computationally very simple and yet conceptually more realistic than a fixed grid approach. Because of the computational simplicity, the algorithm is economical, and long term simulations which incorporate the finest scales become a realistic proposition. (from paper)
A Lagrangian one-dimensional model of the dynamics of a reservoir is presented. The model utilizes new formulas to describe the dynamics of both the inflow and outflow. In addition, recent advances pioneered in the oceanographic literature have been modified and applied to the dynamics of the mixed layer. Finally, a new model is proposed for deep hypolimnionic mixing. The model so constructed contains only four universal constants. Data from one complete annual cycle of the salinity and temperature distribution from a Western Australian reservoir are used to calibrate the model.
Field data from two lakes of widely differing geometry and size are analyzed in terms of four nondimensional numbers which allow the principal mixing processes in each lake to be identified. The numbers are based on basin geometry, density stratification, wind stress, and rates of inflow and outflow. The procedure highlights the differences in the dynamics of the two lakes and allows assessment of the validity of the assumption of one-dimensionality. The result is that both lakes were dominated by one-dimensional, but different processes. The dynamics of the epilimnion of the smaller lake were dominated by stirring from surface wind and cooling, whereas shear at the pycnocline was also significant in the larger lake. In neither case did the effects of the earth’s rotation, inflow, or outflow generate significant horizontal gradients. A one-dimensional numerical model (DYRESM) was used to simulate the vertical temperature and salinity structures of both lakes over lengthy periods, with good results. The model is based on the parameterization of the important physical processes in a framework of horizontal layers of variable thickness and was applied in both lakes without alteration. The interpretative power of the model is demonstrated by examination of the formation and erosion of a thermal inversion in the larger lake.
A vertical one-dimensional model is proposed to describe the vertical mixing in the central portion of the central basin of Lake Erie in summer. The model explicitly describes the various mechanisms which operate in the well mixed upper and bottom layers. The model predictions are in good agreement with the observed thermal structure over the entire water column during a month-long simulation period. The central basin of Lake Erie has been the subject of intensive investigation in recent years due, in large part, to concern over the depletion of dissolved oxygen in the hypolimnion during the stratified summer period (Burns and Ross 1972; Burns 1976). The central basin encompasses an area about 80 km wide and 200 km long but with a mean depth of only 18 m and a maximum depth of 26 m. During summer, the lake is strongly stratified and is characterized by an upper mixed layer (UML), a strong temperature gradient region, and a near-isothermal bottom mixed layer (BML) (see