Comparison

- of Indices for the Scaling and

Corrosion Tendency of Water
H. M. MULLER-STEINHAGEN and C. A. BRANCH

Department of Chemical and Materials Engineering, The University of Auckland, Auckland, New Zealand
Three commonly used fouling indices to characterize the scaling and corrosion tendencies of water, namely the Satu-
ration Index, the Langelier Saturation Index and the Ryznar Stability Index, have been compared and checked for con-
sistency. While the first two indices give similar results, major deviations between the predictions of saturation and
stability indices were found.

Nous avons compare trois indices d'encrassement communkment utilists pour caracttriser les tendances de I'eau ?I
I'ecaillage et ?I la corrosion, ?I savoir l'index de saturation, I'index de saturation de Langelier et I'index de stabilitt
de Ryznar, pour lesquels nous avons vCrifiC la cohtrence. Tandis que les deux premiers indices donnent des rksultats
similaires, des tcarts importants ont Ctt trouvCs entre les prkdictions des indices de saturation et de stabilitt.

Keywords: calcium carbonate scaling, saturation index, Langelier Saturation Index, Ryznar Stability Index.

T he performance of heat exchangers using cooling water

is often greatly reduced by the formation of CaCO,
scale on the heat transfer surfaces. Calcium carbonate scaling
occurs once the liquid at the heat transfer surface becomes
supersaturated with respect to CaC03. Being an inverse
solubility salt, CaC03 precipitates on hot surfaces. Scaling
can be reduced by proper design of the heat exchanger and/or
chemical treatment of the cooling water. Since both fouling
mitigation strategies involve major capital and operating
costs, reliable information about the scaling behaviour of a
given cooling water composition must be available.
Fouling Indices
Scaling indices use the composition of an aqueous solu-
tion to indicate whether the solution has the potential for
scaling or corrosion. Although sophisticated methods for the
prediction of CaC03 scaling rates are available (Hasson,
1981), water treatment is usually determined using one of
the following indices (Nalco, 1979; Betz, 1976):
The Saturation Index
with
in his scaling experiments because it is similar to solutions
found in typical cooling water cycles. As can be seen in
Figure 1, the Saturation Index stays close to zero for lower
values of pH and rises sharply at higher pH.
Langelier (1939, 1946) has developed a qualitative for-
mula to predict if CaC03 is precipitated or dissolved in
water. The equation for the Langelier Saturation Index
is based on pH, total alkalinity, hardness and temperature,
which all affect the solubility of calcium carbonate in
water:
L.S.I. = pH - pH, ......................... (3)
The pH at saturation, pH,, can be calculated from
pH, = pK2 - pKSp + p C a f + + p(T.A.) . . . . . . . (4)
Scaling occurs if L.S.I. > 0, corrosion if L.S.I. < 0.
Figure 2 shows the Langelier Saturation Index as a func-
tion of heat transfer surface temperature and pH for the same
solution as used for Figure 1. Scaling tendency increases with
increasing values of L. S. I. , i. e. with increasing temperature
and increasing pH.
The Langelier Saturation Index is very similar to the Satu-
ration Index, which can also be written in logarithmic form
as:

(Ca++.CO,--), = K,,, ...................... (2)

uses the degree of saturation to determine the scaling poten-
tial of a CaC03 solution. Scaling may only occur if the solu-
tion is super-saturated with respect to CaCO,; hence
S.I. > 1 denotes a scaling tendency
S.I. < 1 denotes a corrosion tendency.
Figure 1 shows the saturation index as a function of pH
and heat transfer surface temperature for a solution containing
0.006254 mol/litre C a + + for a total alkalinity of 0.003397
mol/litre. A similar solution was used by Watkinson (1975)
log(S.Z.) = - pCa++-p(T.A. + Hf - OH-)
+ pKsp - p(2 + H+/K2) . . . . . . . . . . . . . . . . . . (5 1
Therefore, similar predictions can be expected for both
indices.
Saturation indices only show the direction of the driving
force but do not indicate if the supersaturation is high enough
to initiate crystallization. Ryznar (19U), therefore, suggested
an index, which is said to be a quantitative as well as a
qualitative measure of the solution's scaling potential. The
Ryznar Stability Index as defined as
R.S.Z. = 2pH, - pH ........................ (6)

THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING, VOLUME 66, DECEMBER, 1988 1005
 400

M
hs -0
-
2 300
C
x
c .-
- 6 -
-w
.-
.Q n
-w
e w i

2 2oo
K 4 \\'

3l
100

0 2
-
4 5 6 7 a 9 4 5 6 7
8 9
PH DH
Figure I - Saturation Index as ;I function of pH and temperature. Figure 3 - Ryznar Stability Index as a function of pH and
C a t = 0.006254 mol/litre. T.A. = 0.003397 mol/litre. temperature.
Cat * = 0.006254 molllitre. T.A. 0.003397 molilitre.

While all concentrations in Equations ( I)-(S) are expressed

2 -
in moles per litre, C a + + and T.A. in Equation (7) arc in
parts per million as Ca and CaC03. respectively. Ryznar
-0
M has shown experimentally that the following criteria can be
L used to determine the fouling potential of a solution:
1 -
R.S.I. < 6 denotes a scaling tendency

0 - R.S.I. > 7 denotes a corrosion tendency.

The Ryznar Stability Index is plotted as a function of pH
b and heat transfer surface temperature in Figure 3. Scaling
=*
Q) -1 - tendency increases with decreasing value of the index which,
0
c may lead to confusion if other indices are used as well.
4
Comparison of Indices

The predictions of the three indices were compared for

-2
-3 incipient scaling conditions, i.e. for S.I. = I , L.S.I. = 0
4 5 6 7 8 9 and R.S.I.= 6 . The results may be interpreted as the pH
which has to be adjusted for a given composition and tem-
PH perature to minimize both scaling and corrosion. Figures
Figure 2 - Langelier Saturation Index as a function of pH and
temperature. 4a-4c show the prediction of the three indices as a function
C a t + = 0.006254 molilitre. T.A. = 0.003397 molllitre. of the heat transfer surface temperature for three different
solutions. Scaling occurs above the curves, corrosion is likely
For this index, the saturation pH is defined using an equa- for conditions below the curves. Two curves have been cal-
tion suggested by Larson and Buswell (1942): culated for the Ryznar Stability Index, based on pH,-valurts
according to Equation (4) and Equation (7). respectively.
pH, = pK2 - pK,,,, + pCa'" + p(T.A.) + 9.3 As expected, the Saturation Index and the Langelier Satu-
2.5 1'' ration Index give comparable values and trends. namely
+ 1 + 5 . 3 (I + 5 . S ) O ~
. . . . . . . . . . . . . . . . . . (7) decreasing pH with increasing temperature. pH-values for
incipient scaling predicted by the Ryznar Stability lndcx differ
considerably from those predicted by the two saturation
with I being the ionic strength of the solution calculated from indices. For the solution with the lowest Ca + concentra-
I = 52' (6Ca" + 2(T.A.) + HCO3- + OH- tion and the lowest alkalinity (Figure 4a). the Ryznar Sta-
bility Index predicts higher pH values than the two saturation
+ H' + 4 C 0 3 - - ) ........................ (8) indices, which indicates that a certain supersaturation is

1006 THE CANADIAN JOURNAL. OF CHEMICAL ENGINEERING, VOLUME 66. DECEMBER. I988
 10
7
9.6
R.S.I. with eq (4)
f 6.6 -
L

R.S.I. w i t h eq ( 4 )
92 R.S.I. w i t h eq (7) R.S.I. w i t h eq (7)
6.2
0.8
58
8.4
5.4
p 8
P H 5
7.6
4.6
72 4.2
6.8 3.8
6.4 3.4
6 3
290300 310 320 330 340 350
Temperature K
8 ""'1""l""l""I""
Nomenclature

7.5 - - H
I
= hydrogen ion concentration, mol/litre
= ionic strength of solution
K, = first molar dissociation constant of carbonic acid, mol/litre
K, - - second molar dissociation constant of carbonic acid,
7 mol/litre
K, = molar solubility product for calcium carbonate.
(moI/Iitre)'
L.S.I. = Langelier Saturation Index
PH 6.5 -
px = -log(X)
R.S.I.= Ryznar Stability Index
S.I. = Saturation Index
6 - T.A. = total alkalinity

Subscripts
5.5 s = at saturation

l l l l l l l l l l l l l l l l l l l I I L I I I
5 Betz Laboratories, Inc.. "Handhook of Industrial Water Condi-
290 300 310 320 330 340
Temperature
Figure 4 - Comparison of incipient scaling conditions predicted
by S.I.. L.S.I., and R.S.I.
a) C a + ' = 0.00207 mol/litre. T.A. = 0.000960 mol/litre
b) Ca" = 0.00625 molilitre. T.A. = 0.003397 mol/litre
c) Cat+ = 0.01 122 molllitre. T.A. = 0.004826 niol/litre.
necessary to initiate scaling. However, for higher concen-
trations (Figures 4b and 4c) and higher heat transfer surface
temperatures. scaling is predicted for CaC03 concentrations
well below the saturation concentration. Obviously. this
result is not in accordance with the basic understanding of
crystallization kinetics.
Results obtained using the Ryznar Stability Index should,
therefore. always be checked against the predictions of one
of the two saturation indices. Preference should be given to
the prediction of saturation indices in situations where the
Ryznar Index predicts scaling for sub-saturated conditions.
For predictions of scaling rates, Hasson's ionic diffusion
model (198 1 ) may be used.
References
350 tioning." 7th ed. (1976).
K Langelier. W. F.. "The Analytical Control of Anti-Corrosion Water
Treatment," Journal of the American Water Works Association
28. 1500-1521 (1939).
Langelier, W. F., "Chemical Equilibria in Water Treatment.''
Journal of the American Water Works Association 38, 169-182
( 1946).
Larson, T. E. and A. M. Buswell, "Calcium Saturation Index and
Alkalinity Interpretation," Journal of the American Water Works
Association 34. 1667- 1680 ( 1942).
Hasson. D., "Precipitation Fouling." in Fouling of Heat Transfer
Equipment. E. F. C. Somerscales and J. G. Knudsen. eds..
Hemisphere, Washington. D.C.. 527-568 (1981).
Nalco Chemical Corp.. "Nalco Water Handbook." 1st ed..
McGraw-Hill (1979).
Ryznar. J . W.. "A New Index for Determining the Amount of Cal-
cium Carbonate Scale Formed by Water.'' Journal of the
American Water Works Association 36, 472-486 (1944).
Watkinson, A. P. and 0. Martinez, "Scaling of Heat Exchanger
Tubes by Calcium Carbonate." Trans. ASME J. Heat Transfer
97. 490-492 (1973.
Manuscript received December 7. 1987: revised manuscript
received April 20. 1988; accepted for publication May 27. 1988.

THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING. VOLUME 66. DECEMBER. 1988 I 007