i:¸ i

j_ . . . . .

?" t ' / 1
DESALINATION
II
ELSEVIER Desalination 152 (2002) 167-174
w ww.elsevier.com/locate/desal

Calco-carbonic equilibrium calculation

B. HamrotmP*, M. Dhahbi b
°D~partement de Chimie, Facultd des Sciences de Tunis, Campus Universitaire, 1060 Tunis, Tunisia
Tel. +216 (71) 872600; Fax +216 (71) 885008; email: [email protected]
bLaboratoire de physico-chimie des interfaces, INRST, B.P 95, 2050 Hammam-Lif, Tunisia
Tel. +216 (71) 430044; Fax +216 (71) 430934; email: [email protected]

Received 1 April 2002; accepted 15 April 2002

Abstract

The Langelier and the Stiff-Davis saturation indices have found wide acceptance as indicators of the degree of
water saturation with respect to calcium carbonate. But the risk of aggressiveness or scaling may be accurately
evaluated from the analytical data by application either of the graphic methods or the complex methods of calculation
made possible by the use of data processing. The present work was devoted to the study of this problem using a
computerized model of Legrand and Poirier, based on the plot of the CO2 concentration vs. Ca2÷ concentration.
Taking into account the formation, in natural waters, of ion pairs as well as the various equilibrium constant variation
with temperature made improvements to this model. Consequently the ionic activity coefficients were calculated
using suitable models. Calculation was developed by means of the graphical LabViewTM programming language.
The calco-carbonic curves obtained constitute a useful tool to visualize if the studied water was in equilibrium or not
and provide quantitative indications of how much reagent, such as lime or soda, must be added to bring water to
equilibrium. Calculations carded out with the program were illustrated by some results and compared with those of
Legrand and Poirier and others studies.

Keywords." Calco-carbonic equilibrium; Saturation indices; Calcium carbonate; Activity coefficient; LabView

1. Introduction air and water and precipitation phenomena in

The most important weak acid-base system is waters. Exact knowledge of different salts solubility
the carbonate one as it is intimately involved in in natural waters, brackish waters, seawater and
the control of the pH, exchange of CO 2 between geothermal waters, is vital since the solubility
determines whether the water is saturated or not
*Corresponding author. [1-5].

Presented at the EuroMed 2002 conference on Desalination Strategies in South Mediterranean Countries:
Cooperation between Mediterranean Countries of Europe and the Southern Rim of the Mediterranean.
Sponsored by the European Desalination Society and Alexandria University Desalination Studies and Technology
Center, Sharm El Sheikh, Egypt, May 4-6, 2002.

0011-9164/02/$- See front matter © 2002 Elsevier Science B.V. All rights reserved
PII: S 0 0 1 1 - 9 1 6 4 ( 0 2 ) 0 1 0 5 9 - 7
168 B. Hamrouni, M. Dhahbi / Desafination 152 (2002) 167-174

The scales commonly encountered in the strength of the treated water and on equilibrium
desalination of natural waters are usually com- constants calculation, were taken into account.
posed of calcium, magnesium and barium salts The results of the program were compared to those
of sulfate, hydroxide, and carbonate. Calcium of Legrand-Poirier and applied to different waters
carbonate may precipitate from water in one or at different temperatures. For representative
more of three forms: calcite, aragonite and vaterite. calculations geothermal water from the south of
Which of these allotropes precipitates depends Tunisia, mineral water and permeate water from
on the pressure, the temperature, the presence of a reverse osmosis desalting plant, were selected.
foreign ions and the rate at which precipitation
takes place [6-9]. The calcite polymorph of calcium
carbonate is the most common scale-forming
mineral. Results from a number of studies involving 2. Thermodynamic aspects of caclco-carbonic
spontaneous precipitation reveal that, at moderate equilibrium
temperatures, vaterite is the solid phase forming 2.1. Carbonic system
first during the precipitation process. The pre-
Basically the carbonic system is derived from
dominance of one crystalline phase over the other
the dissolution of carbon dioxide and carbonate
depends on conditions such as pH, temperature,
minerals into the water. Thus, the carbonate system
and the nature of the antiscalants present [10-13].
is a weak acid-base system which exists in aqueous
At higher temperature, though calcite is the most
solutions as dissolved carbon dioxide CO.zaq,
thermodynamically stable phase, aragonite has
carbonic acid H2CO3, bicarbonate HCO3-, carbonate
been reported to be the first phase to precipitate
ions CO32- and complexes of these ions such as
from solution [14].
Water is said to be saturated with calcium car- CaliCO3 +, CaCO 3.
Dissolved carbon dioxide are hydrated in few
bonate when it is neither dissolve, nor precipitate
minutes [15,16] according to the reaction:
calcium carbonate scale. This equilibrium condition
is based upon undisturbed water, at constant temp- CaCO3aq + H20 ¢:~ CO2, I-I20 (1)
erature, which is allowed to remain undisturbed
The term H2CO3 refers to the composite form,
for an infinite period of time. Water is said to be
which is the sum of the activities of molecularly
undersaturated if it can still dissolve calcium
dissolved carbon dioxide CO2aq, and the hydrated
carbonate. Supersaturated water will precipitate
form CO 2, H20:
calcium carbonate from water if allowed to rest.
The aim of this study is the calculation of the (H2CO3) = (CO2aq) q- (CO 2, H20) (2)
calco-carbonic equilibrium to predict the behavior
This composite form is convenient because of
of the water towards calcium carbonate, by devel- the analytical difficulties of separating out CO2a
oping a microcomputer program. This practical
from CO 2, H20 which constitute about 3% o~
tool may be useful for different users to establish
H2CO3 [17].
the calco-carbonic equilibrium curves which provide
The behavior of carbonic system is intimately
a more accurate way to determine ira water is in involved in the control of the pH and precipitation
calco-carbonic equilibrium or not. If it is aggres- phenomena by the following equilibrium relation-
sive or scaling it is easy to deduce quantitatively ships:
the appropriate chemical treatment. The micro-
computer program written in LabView graphical H2CO 3 ~ H ÷ + H C O 3- (3)
language uses the extended Debye-Htickel model H C O 3- ¢:~ H÷+ CO32- (4)
to calculate ion activity coefficients. The effects
of ion pairing and temperature on the true ionic C a 2+ + CO32- ~:::::~C a C O 3 (5)
 13. Hamrouni, M. Dhahbi / Desalination 152 (2002) 167-174 169

and Stiff-Davis. For example, Langelier saturation

[H+][HC07] =K o ~/u:co, _K, (6)
index (LSI) is the result of a mathematical formula
[H2CO3] TW'YHCO; using arbitrary numbers that are assigned to five
factors - - pH, total alkalinity, calcium hardness,
total dissolved solids and temperature. LSI is applied
-] - Ko
r.co - K2 (7) to low salinity waters and the Stiff-Davis saturation
[HCO3] 3'H+~'CO~- index is mostly used for high salinity waters.
On the other hand, diagrams giving calco-
carbonic curves are proposed by many authors,
[Ca>][CO]-]_ x° _& (8) among them Langelier [18], HaUopeau [19,205,
~/ca-,+~/co? Roubais [21 ] and Legrand and Poirier [22,23.]. In
the present work the more recent representation
(H+)[ OH- ] = K°w(H20) = Kw (9) of Legrand-Poirier was chosen, among others, for
'YOH- its convenience and for allowing the direct pre-
diction of chemical quantities to be added to bring
the given water to equilibrium.
The behaviour of specific water towards
CaCO 3 (marble) defines its scaling or aggressive
2.3. Legrand-Poirier presentation
nature. If we consider, for a given temperature,
the following equilibrium: For establishing their diagram ofcalco-carbonic
equilibrium, Legrand and Poirier [22,23] dis-
CaCO 3 + H2CO 3 ¢:~ 2HCOf + Ca 2+ (10)
tinguished between what they call:
When the amount ofH2CO 3increases for some • Fundamental elements found in all natural
reason the equilibrium will move to dissolve the waters and result of atmospheric CO 2 inter-
calcium carbonate (the water is said aggressive). action with calcium: H30 +, HCO3- , OH-, H2CO3,
When the quantity ofH2CO3 is less than the quantity Ca 2+, CO32-.
needed for the equilibrium then CaCO 3 precipi- • Characteristic elements whose presence
tates (the water is said scaling). characterizes any natural water: Mg > , K +,Na +,
NOr, SO42-, Cl-.
2.2. Calco-carbonic equilibrium representation
The following relation gives the electric neut-
The risk of aggressiveness or scaling of calcium rality of the medium:
carbonate may be evaluated by different ways:
• From the analytical data by application of the 2[Ca2+1+ [H30+] + 2[Mg >] + [K+]
graphic methods and indices calculations, + [Na +] = [NO,-] + 2[so4=-] + [Cl-] (11)
• From the analytical data by the complex methods + [HCO3-] + 2[CO32-] + [OH-]
of calculation made possible by the use of micro-
where concentrations are expressed in mole or
computers,
mmol/l of solution. The authors used the parameters
• From laboratory tests.
N, P and ~, as follows:
Many saturation indices are used as indicators
of the degree of water saturation with respect to p = 2[Mg >1 + [K+] + [Na+1
calcium carbonate [5]: Langelier saturation index, N = [NO,-] + 2[8042-] + [El-] (12)
Ryznar stability index, Puckorius scaling index,
Larson-Skold index and Oddo-Tomson index. N-P
The most widely used indices are those of Langelier 2
170 B. Hamrouni, M. Dhahbi/ Desalination 152 (2002) 167-174

Neutrality equation is: 3. R e s u l t s a n d d i s c u s s i o n

2~CaZ+ ] - Z,) 3.1. Improvement to Legrand-Poirier repre-

(13) sentation
= [ H C O 3 ] + 2[CO ~-] + [OH- ] - [H 30 + ]
The calco-carbonic equilibrium study was baseA
with on the graphical representation of Legrand and
Poirier [22,23], of the total dioxide carbon concen-
[CO2 ]r = [H2CO3 ] + [ H C O 3 ] + [CO23- ] (14) tration [COJr vs. calcium concentration [Ca2*].
To improve this representation calco-carbonic
and calculation was modified by introducing the effect
[CO2 ]r = 2([ Ca2+] - k) + [H3O+ ] of ion pairing on the ionic strength and on the
(15) apparent equilibrium constants, and by using a
- [OH- ]- [CO~- ] + [H 2 C 0 3 ] more accurate model for activity coefficients
depending on ion nature and temperature.
The free carbonic species concentrations are: Our previous calculated distribution of the
[CO2]r[H+] 2 major constituents of seawater and brackish water
[H2CO3]- [H+] 2 +K~[H+]+K, K2 (16) showed the importance of ion pairs formation in
seawater as well as in brackish water [3,4]. In
these two cases CaCO32-, HCO3-, SO42-and Mg~+
[CO2]r[H+]KI are the most affected by ion pairing formation. The
[HCO3] = [H+]2 +K,[H+]+K, K2 (17) most important associated ionic species encountered
are: CaSO2, MgSO4°, NaSO4-, KSO4-, Na2SO4°,
K~SO4°, MgHCO3+, CaliCO3+, MgCO3°, CaCO3°,
[CO2]TKIK2 NaCO3-, MgOH% CaOH +, NaOH °.
[CO]-] = [H+] 2 +K~[H+]+K~K2 (18)
In this way the true ionic strength takes into
account the free and associated (ion pairs) charged
From these equations and the solubility product species:
of calcium carbonate [Eq. (8)], the authors built
their caico-carbonic equilibrium diagram by plotting 1 2
the total dioxide carbon concentration [CO2]r vs. I = - ~ Z i [ionic free species]L
2 i
calcium concentration [Ca2+]. Each point of the
curve obtained represents particular water in + 1 ,~ Z2[charged ion pairs ] (19)
equilibrium. Thus the space is subdivided into two
regions: aggressive waters on the left and scaling At finite dilution deviations from ideality are
waters on the right. observed and the activity and concentrations are
In their representation Legrand and Poirier no longer equal and they are linked by the activity
made the following hypothesis: coefficients. Among the available models for the
• All the ions are considered to be free without activity coefficients calculation [3,4], the extended
ion pairing Debye-Htickel model was used according to the
• Activity coefficients are supposed to be in- following equation:
dependent of temperature, and the same for
different ions.
logy i = -
A.z, 47 +b~ .I
l+aO.B~ (20)
These hypothesis lead to erroneous ionic strength
and apparent equilibrium constants. When the concentration [COJr is not available,
 B. Hamrouni, M. Dhahbi / Desalination 152 (2002) 167-174 171

the total carbonic species concentration calculation f) for each value of total carbon dioxide con-
may be based on the knowledge of the total alkalinity centrations, steps (c) to (e) were repeated.
(TA), Eq. (14) and the following relations: g) output: calco-carbonic equilibrium curves
COsT =f([Ca2+]) and calco-carbonic treatment to
[HCO3]r = [HCO3]L + ~-'~M,HC03 (21) bring water to equilibrium.
i

[CO2-]r = [CO~-]L + )-'~M, CO3 (22) 3.3. Calco-carbonic curves establishing

i
Different kinds of water were used to perform
where M are cations susceptible to give ion pairs representative calculations. Composition of these
with H C O [ and CO3 z-. waters, namely, synthetic water, desalinated water
The total alkalinity may be linked to species and geothermal water, are reported in Table 1. The
concentration by le following expression: synthetic water given as an example by Legrand
and Poirier [22] was used for comparison. Desali-
T A = [ O H - ] + [ M g O H +] + [CaOH +]
nated and geothermal waters were from South
+ [HCO 3 ] + [CaliCOS ] + [MgHCO~ ] Tunisia.
(23)
+ 2[CaCO ° ] + 2[MgCO~ ] + 2 [ N a C 0 3 ] Table 1
+ [ N a O H ] + [ N a H C O ] ] - [H + ] Composition and analyses of the different waters used for
illustrations

Parameters Type of water

3.2. Calculation procedure
Synthetic Desalinated Geothermal
The program developed in this work checks pH -- 7.5 ?
the introduced water composition according to the Temperature,°C 18 23 62
electroneutrality of the salt aqueous solutions. The Na+, mmol/L 0.12 0.74 14.6
different steps of the iterative procedure necessary K ÷, mmol/L 0.1 0.05 1.07
Ca2÷,mmol/L -- 0.29 6.3
to perform the calculations were:
Mg2+,mmol/L 0.05 0.20 2.91
a) input of total molar concentrations of prin- CI-, mmol/L 0.3 1.13 19.0
cipal species, thermodynamic equilibrium constants SO42-, mmol/L 0.8 0.06 7.29
for acid-base species, associated ions (ion pairs, CO32-, mmol/L -- 0 0
triple ions and complexes) and solubility product HCO3-, mmol/L -- 0.68 2.33
for calcium carbonate, temperature.
b) ionic strength and lambda were calculated The simplified version of the developed program
based on total species concentrations choice of permits reaching more easily and more quickly
starting value for total carbon dioxide concentra- the results of Legrand and Poirier. Calculations
tions. obtained by the modified version were compared
c) after pH initialization, activity coefficients with those given for synthetic water by the same
and apparent equilibrium constants were calculated. authors. Table 2 summarizes the results indicating
d) free and associated species concentrations important deviation that confirms the importance
and the ionic strength were determined. of taking into account the ion pair formation. In
e) after pH increment, steps (c) and (d) were this case at least 10% of total calcium are associ-
repeated until calcium carbonate ionic product ated with bicarbonate, carbonate and sulfate ions.
equalled apparent equilibrium constant for its An example of establishing equilibrium calco-
solubility. carbonic curves is given in Fig. 1 in the case of
172 B. Hamrouni, M. Dhahbi / Desalination 152 (2002) 167-174

Table 2
Comparison between our results and those of Legrand and Poirier [22]

Parameters
[CO2]r 1.0 2.0 3.0 6.0 9.0 12.0 15.0 20.0 30.0
pHs 8.1 7.7 7.4 7.0 6.8 6.6 6.5 6.4 6.2
Ca2+[22] 1.30 1.77 2.21 3.37 4.32 5.12 5.83 6.83 8.45
Ca2+this work 1.21 1.68 2.11 3.19 4.02 4.68 5.25 6.03 7.24
Deviation, % 6.9 5.2 4.7 5.7 7.6 9.4 11.0 13.3 16.8

2.0- • Table 3
1.8- Two series of marble test for the desalinatedwater
16- regionwherewaters
1.4- (CaCO3arc
corroslvedissolution) j
Parameter pH TA, °F TH, °F
1,2- / regionwherewatershave Without CaCO3 5.9 0.46 0.21
/ tendencytoscale
~1.0- rep..... taivepointJ (CaCO3-Vtecipitation) 5.5 0.45 0.18
408-
With CaCO3 8.2 3.95 3.94
-- 02-0"4"0'6- ~rbo!c _ equilibri
l ........ 8.2 3.75 3.40

0.0- , i l , , i , n ~ ,
0.0 01 0.2 0.3 0.4 0.5 0.6 0.7 0.8 09 1.0 20.0-
[Clt2+] I ~ L

Fig. 1. Calco-carbonicequilibriumcurvesestablishing,case ioo_ / / !

of a permeate water at 25°C from South Tunisia. ~ 14.0-
12.0-
desalinated water by reverse osmosis from South 10.0-
Tunisia. This figure shows the position of the water ~ 8.0-
representative point (dark point) in relation to the 6.0-
4.0-
equilibrium curve dividing the space in two regions
2.0-
corresponding to aggressive waters on the left and
0.¢'
scaling waters on the right. s.o 6'.o 71o 8'.0 9'.o 1;.o 11.o
i~2+ I IUT
The experimental results of the marble tests,
obtained after 24 h for the desalinated water, are Fig. 2. Calco-carbonicequilibrium curves for a geothermal
given in Table 3. Increasing values observed for water from South Tunisia, at 62 and 18°C.
pH, total alkalinity (TA) and total hardness (TH)
temperature in the study of the behaviour o f given
mean that the water has tendency to dissolve CaCO 3.
water towards calcium carbonate.
This conclusion confirms the position of the repre-
sentative point on the left of the calco-carbonic
3.4. Example o f calco-carbonic treatment
curve as predicted in Fig. 1.
In Fig. 2, two equilibrium curves were drawn To bring water to calco-carbonic equilibrium
for the geothermal water taken at its extraction many reagents may be added such as lime, caustic
temperature 62°C and when cooled to 18°C. It soda, calcium carbonate and sodium carbonate.
appears that this water is nearly in calco-carbonic For example, calculations were deduced from calco-
equilibrium at 62°C but becomes aggressive at carbonic curve of desalinated water in the case of
18°C. This indicates the important effect of post-treatment by lime, calcium carbonate and
 B. Hamrouni, M. Dhahbi / Desalination 152 (2002) 167-174 173

After the improvements and refinement made, the

1.8-
computerized representation of Legrand and Poirier
1.6-
2.0- m / ' 1 ~ chosen was convenient and quantitative. Improve-
ment was made by taking into account the ion pair
I I.2-
formation, temperature effect and by performing
adequate activity coefficient calculations. The
o.6- / comparison of our results with the original method
0.4- confirms the importance of the modifications
0.2- made. The application of the program to different
0.0 , , - kinds of water at different temperatures shows
o'.~ o b c.3 0.4 o.s o'.6 '
0.7 "0.8 Oig I
[Ca2~h,mml [CaS+l I ~ that it constitutes a more accurate and rapid way
~. [Ca2+Jn,,al
to know the behaviour of a given water towards
Fig. 3. Determinationof lime quantity necessary to bring calcium carbonate. Moreover, it permits, using
the desalinatedwater to caico-carbonicequilibrium. the calco-carbonic curves, choosing the treatment
and calculating the quantity of reagent necessary
2.0- . to bring water to calco-carbonic equilibrium. Taking
1.8- into account kinetic considerations may complete
1.6- the thermodynamic approach followed in this work,
1.4-
which can be improved.
I 12-
o.o-
Symbols
4,B - - D e b y e - H t i c k e l parameters for the
activity coefficients
O0 .I
ol0 o/:!,.o:2
f-=-~ r- •
o,
,
06
, , ,
00
,
,0 a i° - - Distance of closed approach between
L. [Ca2~]initinl =l let'*n ,moUL hydrated ions
L~ [Ca?L]final__~___~ bi - - Adjustable parameter for ion i added
Fig. 4. Determination of calcium carbonate quantity to the Debye--Hackel equation
necessary to bring the desalinatedwater to calco-carbonic 1 Ionic strength of the solution
equilibrium. [i], [i]r - - Free and total concentration ofi species,
respectively
sodium hydroxide. For illustration, Figs. 3 and 4 (0 - - Activity of species i
show the examples of treatment by lime and ro,& Thermodynamic and apparent equi-
calcium carbonate respectively. It is easy to deduce librium constant for calcium carbonate
that the quantities needed to bring the studied respectively
water to calco-carbonic equilibrium are per liter: K 0,Kv - - T h e r m o d y n a m i c and apparent equi-
4.5 mg of lime or 8 mg of sodium hydroxide or librium constant for water ionization
13 mg of calcium carbonate. respectively
K °, K Thermodynamic and apparent equi-
librium constant for the ionization
4. Conclusion
number i of carbonic acid respectively.
The different saturation indices and other pH.~ - - log(H+)s, pH of solution saturated by
graphical methods constitute simple indicators for CaCO 3
the behaviour of water toward calcium carbonate. TA - - Total alkalinity
174 B. Hamrouni, M. Dhahbi / Desalination 152 (2002) 167-174

TH - - Total hardness [10] M.M. Reddy, in: Chemistry of Wastewater Techno-

Zi -- Charge o f ion i logy, A.J.Rubin, Ed,, Ann Arbor, 1978.
Activity coefficient o f i species [ll] P.G Koutsoukos and C.G. Kontoyannis, J. Crystal
v, Growth, 69 (1984) 367.
[12] A.G.Xyla and P.G Koutsoukos, J. Chem. Soc. Faraday
Trans., 83 (1987) 1477.
References
[13] T. Ogino, T. Suzuki and K. Sawada. J. Crystal Growth,
100 (1990) 29.
[1] B. Hamrouni and M. Dhahbi, J. Soc. Chim. Tunisie, [14] F. Lippmann, Sedimentary carbonate minerals, in:
4(6) (1999) 461-474. Mineral Rocks and Inorganic Materials; Springer-
[2] B. Hamrouni and M. Dhahbi, Desalination, 136 (2001) Verlag, New York, 1973.
225. [15] D.M. Kern, J. Chem. Ed., 37 (1960) 14.
[3] B. Hamrouni and M. Dhahbi, Desalination, 137 (2001) [16] P. Jones, M.L. Hagget and J.L. Longridge, J. Chem.
275. Ed., 41 (1964) 610.
[4] B. Hamrouni and M. Dhahbi, Activity coefficients [17] W. Stumm and J.J. Morgan, Aquatic Chemistry,Wiley
calculation: application to salt aqueous solutions, Interscience, N.Y., 1970.
Entropie, 237 (2002) 52. [18] W.F. Langelier and J. Zamer, J. AWWA, 30 (1936)
[5] P.R. Roberge, Handbook of Corrosion Engineering, 1802.
McGraw-Hill, 1999. [19] J. Hallopeau, La Technique Sanitalre, 1 (1949) 1.
[6] R.E. Loewental and G.V.RMamis, Carbonate Chemis- [20] J. Hallopeau, Terre des Eaux,13 (1961) 2.
try of Aqueous Systems, Ann Arbor, Michigan, 1976. [21] M. Roubaix, Bull. Soc. Chim. Beige, 54 (1945) 10.
[7] K.B. Krauskopf, Introduction to Geochemistry, [22] L. Legrand and G. Poirier, Chimie des Eaux Naturelles,
McGraw-Hill, N.Y., 1967. Eyrolles, Pads, 1976.
[8] R.A.Robbie and D.R Waldbaum, Geol. Survey Bull., [23] L. Legrand, (3. Pokier and P. Leroy, Les Equilibres
(1968) 1259. Carboniques et l'Equilibre Calco-Carbonique dans les
[9] H. Roques and A. Girou, Water Res., 8 (1974) 907. Eaux Natureiles, Eyrolles, Paris, 1983.