Mechanism of Calcium Carbonate Scale Deposition under Subcooled Flow Boiling Conditions

Mechanism of Calcium Carbonate Scale Deposition under Subcooled Flow Boiling Conditions screenshot

Text-only Preview

Chinese J. Chem. Eng., 13 (4) 464 - 470 (2005)
Mechanism of Calcium Carbonate Scale Deposition under
Subcooled Flow Boiling Conditions
XING Xiaokai MA Chongfang and CHEN Yongchang
College of Environmental and Energy Engineering, Beijing University of Technology, Beijing 100022, China
Abstract Fouling of heat transfer surfaces during subcooled flow boiling is a frequent engineering problem in
process industries. It has been generally observed that the deposits in such industrial systems consist mainly
of calcium carbonate (CaCO3), which has inverse solubility characteristics. This investigation focused on the
mechanism to control deposition and the morphology of crystalline deposits. A series of experiments were carried
out at different surface and bulk temperatures, fluid velocities and salt ion concentrations. It is shown that the
deposition rate is controlled by different mechanism in the range of experimental parameters, depending on salt ion
concentration. At higher ion concentration, the fouling rate increases linearly with surface temperature and the
effect of flow velocity on deposition rate is quite strong, suggesting that mass diffusion controls the fouling process.
On the contrary, at lower ion concentration, the fouling rate increases exponentially with surface temperature and
is independent of the velocity, illustrating that surface reaction controls the fouling process. By analysis of the
morphology of scale, two types of crystal (calcite and aragonite) are formed. The lower the temperature and ion
concentration, the longer the induction period and the higher the percentage of calcite precipitated.
Keywords heat transfer, fouling, calcium carbonate, subcooled flow boiling
below its saturation temperature, while the tempera-
Fouling is generally defined as an unwanted depo-
ture of heat transfer surface is above the saturation
sition of suspended, dissolved, or chemically generated
temperature. Bubbles detaching from the heat trans-
material in process fluids onto the heat transfer sur-
fer surface will then collapse and condense in the sub-
faces. Once the scale builds up on a heat transfer
cooled bulk liquid. Subcooled flow boiling can occur
surface, at least two problems associated occurs[1,2].
over a considerable length of heat transfer equipment
The first problem is the degradation in the perfor-
and may represent up to 50% of the total heat duty[4].
mance of heat transfer equipment. The second less
In evaporators, fouling is usually more severe in the
critical problem is that a small change in tube diam-
subcooled boiling region, due to the low flow velocity
eter substantially decreases the flow rate or increases
and turbulence level.
the pressure drop in the heat transfer equipment.
Several investigators have studied fouling mech-
Fouling is a time-dependent phenomenon. Never-
anisms to understand, quantify, and develop reme-
theless, the current method for designing industrial
dial or preventive treatment. Most of these studies
boilers and heat exchangers exposed to a fouling envi-
have been devoted to fouling during forced convective
ronment is to use empirical constants for fouling resis-
heat transfer and hardly any information is available
tance By using a constant value for the fouling resis-
on fouling during sub-cooled flow boiling. Oufer et
tance at the design stage, one can estimate what may
al.[5] investigated fouling during subcooled flow boil-
happen to the equipment performance but not when
ing of organic fluids. However, their results did not
it will happen. Thus, it is probable that the equip-
lead to any generalized conclusion and cannot be ap-
ment will have to be taken out of service for cleaning
plied for scale formation from aqueous solution. Under
at an inconvenient and economically undesirable time.
most conditions, fouling is more severe during boil-
In order to provide a satisfactory surface area for an
ing heat transfer because of the mechanisms of bub-
acceptable period of operation, it is necessary to be
ble formation and detachment during boiling[3]. A
able to predict the dependence of fouling resistance
recent literature review by Jamialahmadi and Miiller-
on both time and operation parameters.
Steinhagen[6] on the mechanisms of boiler fouling re-
The nucleate boiling increases the formation of de-
veals that experimental fouling data under boiling
posit on heat transfer surfaces
conditions in general and under subcooled flow boiling
[3]. Boiling can be di-
vided into pool boiling and flow boiling, according to
in particular are scarce and incomplete.
the motion of the fluid. Subcooled flow boiling occurs
Hasson et al.[7,8] analyzed the mechanism of cal-
when the liquid enters the equipment at a temperature
cium carbonate scale deposition from a laminar falling
Received 2004-08-12, accepted 2005-03-23.
* Supported by the Special Funds for Major State Basic Research Projects of China (G2000026304).
** To whom correspondence should be addressed. E-mail: [email protected]

Mechanism of Calcium Carbonate Scale Deposition under Subcooled Flow Boiling Conditions
film under evaporative non-boiling conditions. Theo-
rates were controlled by three control valves in front
retical models were presented by taking into account
of the test section and measured by three flow me-
the effects of water composition and operating con-
ters, which were calibrated using precision weighing
ditions. Based on the analysis of the experimental
balance. The supply tank was maintained at a prede-
results it was concluded that diffusion effects could
termined temperature with an internal temperature-
be of importance even in thin-film flow. Hasson et
controlled band heater. The bulk temperature was
al.[9] investigated calcium carbonate scale deposition
measured with T-type thermocouples located in mix-
on the surface of an annular constant heat flux ex-
ing chambers before and after the test section.
changer. They examined the effect of parameters such
as flow velocity, scale surface temperature, and water
composition on scale growth by measuring the scale
deposition rates. It was found that calcium carbon-
ate deposition is mainly controlled by the diffusion of
Ca2+ and HCO3- ions. Chan and Ghassemi[10] used
conservation equations and surface reaction kinetics
to model scaling of heat transfer surfaces by calcium
carbonate. Multi-species reaction rates were used in
their model to predict the calcium carbonate fouling
Recently, Najibi et al.[11,12] performed many ex-
periments on scale deposition of calcium sulphate and
calcium carbonate during subcooled flow boiling in a
vertical annulus. In the investigated range of flow ve-
locity, an almost linear relation between fouling resis-
tance and time was observed, except during the initial
period of experiments. The deposition rate was found
to be controlled by different mechanisms, depending
Figure 1 Schematic diagram of experimental setup
on flow velocity and surface temperature. Different
1—storage tank; 2—heat exchanger;
trends were observed for conditions where convective
3—control valve; 4—flow meter;
5—mixing chamber; 6—test seetion; 7—centrifugal pump;
heat transfer or nucleate boiling dominates.
The main objective of the present investigation is
to study systematically the mechanism of calcium car-
bonate scale formation during subcooled flow boiling
by measuring the drop in overall heat transfer coef-
ficient in a wide range of salt concentration, flow ve-
locity, bulk and surface temperatures. The deposit
formed at varied operating parameters is analyzed by
scanning electron microscope (SEM).
Figure 2 Schematic diagram of test section
2.1 Experimental setup
The measurements were performed in a flow cir-
2.2 Test solution
cuit, including three parallel annular test sections as
Calcium carbonate crystals have an inverse solu-
shown schematically in Fig. 1. Each test section con-
bility with temperature. In the experiments of this
sists of an electrically heated cylindrical stainless steel
work, because the heated surface temperature is al-
rod. This test heater, shown in Fig. 2, is located con-
ways higher than the bulk temperature, the satura-
centrically within the surrounding vertical pipe and
tion concentration will decrease near the heated sur-
the test liquid flows through the annulus in upward
face. Thus, calcium carbonate is easy to deposit on
direction. A miniature stainless steel sheathed re-
the heat transfer surface.
sistance wire is fitted into the centre of the heater
Since calcium carbonate crystals do not dissolve
rod. Three T-type thermocouples are located closely
easily in water, the simulated hard water was used as
below the surface of the stainless steel layer to ob-
test solution, which was produced by dissolving a cer-
tain information about the surface temperature. The
tain quantity of calcium chloride (CaCl2) and sodium
fluid was pumped from a temperature-controlled sup-
bicarbonate (NaHCO3) in water and the molar ratio
ply tank through the annular test sections. The flow
of the former and the latter is 0.5. This method is
Chinese J. Ch. E. 13 (4) 464 (2005)

Chinese J. Ch. E. (Vol. 13, No. 4)
Table 1 Range of operating parameters
preferable because it provides high enough ion concen-
The local bulk temperature Tb, was the arithmetic
tration in solution. In the experiments, each run was
average value of the temperature in the mixing cham-
started with 50 L of solution and salt concentrations
ber at the inlet and the outlet:
were of measured by EDTA titration. All chemicals
used were of analytical grade. As the test solution
was circulated, the hardness decreased because solid
The heat transfer coefficient a was calculated from
3 precipitated from the test solution, adhering
to the heated surface or settling at the bottom of the
reservoir tank. Therefore, the test solution was re-
placed by new solution from another tank after 4 h of
The overall fouling resistance was calculated from
the heat transfer coefficient at the beginning of each
2.3 Experimental procedure and data acquisi-
experiment and the actual heat transfer coefficient af-
ter a certain operational period, according to the fol-
A certain quantity of NaHCO3 was added to the
lowing equation:
supply tank with 50 L of deionized water for start-
ing the test. The pump was turned on, and the flow
rate was regulated to given values. The temperature
of solution, the specified heat flux and surface tem-
perature were adjusted to desired values. The range
The uncertainty in experimental results was esti-
of the experimental parameters is given in Table 1.
mated, based on the uncertainty in various primary
Altogether, 32 fouling runs were performed, most of
experimental measurements such as flow rate and tem-
them in the subcooled flow boiling region. The steady
perature. The basic uncertainties of the heat flux,
state conditions were obtained when the above param-
temperature and flow rate were found to be ±1.5%,
eters remained unchanged. Then a certain quantity of
±0.5% and ±0.5%, respectively. The uncertainties for
the over all heat transfer coefficient and fouling resis-
2 was added to the supply tank. After the test
solution was stirred for half a minute, the tempera-
tance were estimated to be 5% and 10%, respectively.
ture values of thermocouples were recorded. During
the experiment, inlet solution temperature, flow veloc-
ity and heat flux were kept constant by adjusting the
3.1 Fouling curves
corresponding instruments. The surface temperature
A fouling curve shows the relationship between the
of the electrically heated test piece changed during the
thermal resistance of the fouling deposit and time, as
experiment due to the formation of deposits. Surface
shown in Fig. 3. The shape of fouling curves is in-
temperatures indicated in the various diagrams always
dicative of the phenomena occurring during the foul-
refer to the initial temperature.
ing process. Typical curves at lower ion concentra-
A data acquisition system was used to mea-
tions show a linear relation between fouling resistance
sure temperatures, velocities and heat fluxes at pre-
and time, which indicates that the deposition rate is
selected time intervals. All data were recorded with a
constant and there is no removal. On the contrary,
DELL personal computer in connection with an Agli-
parabolic curves are observed at higher ion concentra-
gent Benchlink Data Logger with 20 channel multi-
tion. The parabolic relationship is characteristic for
poor coherence and tenacity of the deposit owing to
higher growing rate and the outer fouling layer may
The surface temperature TS was calculated from
be removed.
the measured average values TW of three thermocou-
3.2 Effect of surface temperature on fouling
The variation in fouling resistance with surface
where s(=1mm) is the distance between the location
temperature at constant bulk temperature and liq-
of thermocouple and the heat transfer surface and
uid velocity is shown in Fig. 3. The results show
that fouling rates depend strongly on the heat trans-
-1·K-1) is the thermal conductivity of the
stainless steel. The power supplied to the test heater
fer surface temperature especially at lower ion con-
was calculated from the measured current and voltage.
centrations. Since mass transfer coefficient increases
August, 2005

Mechanism of Calcium Carbonate Scale Deposition under Subcooled Flow Boiling Conditions
linearly with temperature[9], fouling rates should in-
crease linearly with surface temperature for the mass
transfer controlled operating conditions. This trend
was indeed observed for Ca2+ ion concentration above
5mmol·L-1. For Ca2+ ion concentration below
5mmol·L-1, the fouling rate increases exponentially
with surface temperature, which illustrates that foul-
ing is occurring under reaction-controlled conditions.
The results shown in Fig. 4 confirm the above hypoth-
esis that there are different mechanisms to control the
process of scale formation at different ion concentra-
Figure 4 Effect of surface temperature on the fouling
rate for Ca2+ concentration of (a) Smmol·L-1 (b)
3.3 Effect of fluid velocity on fouling resistance
This effect of fluid velocity is quite clear for differ-
ent ion concentrations, as indicated in Fig. 5. At low
ion concentration, an induction period is observed in
the scale growth process, which tends to increase with
decreasing ion concentration. This is a direct conse-
quence of the fact that only heterogeneous surface nu-
cleation takes place, at relatively small rate. Further-
more, "surface reaction" appears to control the initial
growth of nuclei/crystallites, which are isolated and
distributed on the surface. Under these conditions,
individual crystal growth, and in particular surface
coverage are relatively slow. The controlling process
here implies that there is no effect of flow velocity on
the deposition rate, as shown in Fig. 5(a). At higher
ion concentration, the rate of nucleation is fairly large
and the surface is rapidly covered by deposited mass.
Thus, there is no induction period. The effect of flow
Figure 3 Effect of surface temperature on the fouling
velocity on deposition rate is quite strong, suggesting
resistance for Ca2+ concentraton of (a) 5mmol·L-1
(b) 10mmol·L
that convective diffusion controls the scale formation.
Chinese J. Ch. E. 13 (4) 464 (2005)

Chinese J. Ch. E. (Vol. 13, No. 4)
Figure 5 Effect of fluid velocity on the fouling
resistance for Ca2+ concentration of (a) Smmol·L-1
Figure 6 Effect of bulk temperature on the fouling
(b) 10mmol·L-1
resistance for Ca2+ concentration of (a) 5mmol·L-1
3.4 Effect of bulk temperature on fouling re-
3.5 Analysis of scale structure
3.5.1 Macrostructure of the crystalline fouling layer
The effect of bulk temperature is shown in Fig. 6
There can be seen differences in the macrostruc-
for constant surface temperature and fluid velocity.
ture of the crystalline fouling layer at different sur-
The rate of fouling is independent of bulk tempera-
face temperature. At higher temperature, the layer
ture only at lower ion concentration where the fouling
is dense and shows a strong adhesion on the stainless
process is controlled by surface reaction. At higher ion
steel surface, Fig. 7 as indicates.
concentration, mass transfer controls the fouling pro-
3.5.2 Morphology of calcium carbonate scale at dif-
cess and the mass transfer coefficient increases with
ferent temperature and ion concentration
the bulk temperature. On the other hand, bulk pre-
It is well known that below 35°C, calcite is essen-
cipitation phenomena appear to take place and col-
tially the only polymorph, mainly in prismatic form,
loidal particles form in the bulk solution at high bulk
exhibiting typical rhombohedral faces on the top.
temperature and supersaturation, then these particles
Above 35°C, aragonite appears, exhibiting character-
transport towards and deposit on the surface. There-
istic dendritic formations emerging out of (and adher-
fore, the respective effects may occur concurrently,
ing onto) the metallic substrate. However, during the
creating the increase of fouling rate with bulk tem-
induction period, small calcite crystals tend to cover
the substrate forming a coherent "base layer" (even
August, 2005

Mechanism of Calcium Carbonate Scale Deposition under Subcooled Flow Boiling Conditions
at these higher temperatures)[13], as also reported by
Yang et al.[14,15]. Calcite is the most thermodynam-
ically stable form of CaCO3, and aragonite is the
metastable form. Because the growing rate of CaCO3
is lower during the induction period, it is favorable for
forming thermodynamically stable form. During the
post-induction period, the growing rate of CaCO3 is
higher, and it is favorable for forming the metastable
form. As a result, the lower the temperature and
ion concentration, the longer the induction period and
the higher the percentage of calcite precipitated. The
SEM photomicrographs of scale from the present ex-
periments show evidence of the viewpoint, as shown
in Fig. 8.
Figure 7 Photos of the crystalline fouling layer
([Ca2+]=10mmol·L-1, Tb = 81°C, velocity=0.9m·s-1)
The calcium carbonate fouling experiments were
performed to determine the mechanisms of controlling
deposition and analyze the fouling morphology. The
results show that there are different mechanisms to
control the process of scale formation owing to dif-
ferent ion concentrations. At lower ion concentra-
tion, chemical reaction controls the fouling rate and
the control process implies that there are no effects of
flow velocity and bulk temperature on the deposition
rate, as observed in the present experiments. On the
contrary, at higher ion concentration, the fouling rate
increases linearly with surface temperature and the ef-
fect of flow velocity on deposition rate is quite strong,
suggesting that mass diffusion controls the fouling pro-
cess. By analysis of morphology of scale, as a result,
two types of crystal (calcite and aragonite) are formed
owing to different fouling rates between the induction
and post-induction time.
q heat flux, W·m-2
Chinese J. Ch. E. 13 (4) 464 (2005)

Chinese J. Ch. E. (Vol. 13, No. 4)
fouling resistance, m2·K·W-1
U.S.A. (1990).
radial distance from the hole to outer furface of the
6 Jamialahmadi, M., H., "Scale formation
rod, mm
during nucleate boiling-A review", Corrosion Rev., 11, 25—
average temperature of solution, K
54 (1993).
outer surface temperature, K
7 Hasson, D., Perl, I., "Scale deposition in a laminar falling-
temperature in the wall, K
film system", Desalination, 37, 279—292 (1981).
inlet temperature, K
8 Gazit, E., Hasson, D., "Scale deposition from an evaporat-
outlet temperature, K
ing falling-film", Desalination, 17, 339—351 (1975).
fouling rate, m
9 Hasson, D., Avriel, M., Resnick, W., Rozenman, T.,
Winderich, S., "Mechanism of calcium carbonate scale de-
total heat transfer coefficient, W·m-2·K-1
position on heat transfer surfaces", Ind. Eng. Chem.
heat conductive factor of the stainless steel,
Fund., 7, 59—65 (1968).
10 Chan, S.H., Ghassemi, K.F., "Analytical modelling of cal-
cium carbonate deposition for laminar falling films and tur-
bulent flow in annuli, Part I. Formulation and single species
model", J. Heat Trans., 113, 735—740 (1991).
1 Branch, C.A.,
H., "Influence of scal-
11 Najibi, S.H., H., Jamialahmadi, M.,
ing on the performance of shell-and-tube heat exchangers",
"Calcium sulphate scale formation during sub-cooled flow
Heat Transfer Eng., 12, 37—45 (1991).
boiling", Chem. Eng. Sci., 52, 1265—1284 (1997).
2 Munson, B.R., Young, D.F., Okiishi, T.H., Fundamentals
12 Najibi, S.H., H., Jamialahmadi, M.,
of Fluid Mechanics, 2nd Edition, Wiley, New York (1994).
"Calcium carbonate scale formation during sub-cooled flow
3 Jamialahmadi, M., R., H.,
boiling", Trans. ASME J. Heat Trans., 119, 767—775
"Bubble dynamics and scale formation during boiling of
aqueous calcium sulphate solutions", Chem. Eng. Process,
13 Andritsos, N., Karabelas, A.J., Koutsoukos, P.G., "Mor-
26, 15—26 (1989).
phology and structure of CaCO3 scale layers formed under
4 Helalizadeh, A., H., Jamialahmadi, M.,
isothermal conditions", Langmuir, 13, 2873—2879 (1997).
"Mixed salt crystallization fouling", Chem. Eng. Process,
14 Yang, Q,F., Ding, J., Shen, Z.Q., "Investigation on fouling
39, 29—43 (2000).
behaviors of low-energy surface and fouling fractal charac-
5 Oufer, L., Breber, G., Knudsen, J., "A study of chemi-
teristics", Chem. Eng. Sci, 55, 797—805 (2000).
cal reaction fouling under subcooled boiling conditions",
15 Yang, Q.F., Liu, Y.Q., Gua, A.Z., et al., "Investigation
AIChE Symposium on Heat Exchanger Fouling in Chemical
of induction period and morphology of CaCO3 fouling on
and Food Processing, AIChE National Meeting, San Diego,
heated surface", Chem. Eng. Sci., 57, 921—931 (2002).