Conference Papers  2000 Conference Papers WATER
STABILITY: WHAT DOES IT MEAN AND HOW DO YOU MEASURE
IT?
Peter Gebbie Senior
Process Engineer,
Water Industry Group, Fisher Stewart Pty. Ltd.
DOWNLOAD
(pdf
204K)
ABSTRACT
Currently
there is no requirement under the Australian Drinking
Water Guidelines to produce a stable water that is not
potentially corrosive to water treatment plant equipment
and reticulation systems. This paper discusses the concept
of water stability, describes various indices available
to gauge the corrosivity of a water and their methods
of calculation. An approach is also outlined whereby
the stability of water can be determined following a
particular treatment regime and how it can then be conditioned
to make it less aggressive. Using a worked example,
these concepts are illustrated by examining treatment
and conditioning of a typical water; the Waranga Channel
supply at Rochester, Victoria.
KEY
WORDS
Corrosivity,
Water Stability, Calcium Carbonate, Langelier Index,
CCPP
1.0
INTRODUCTION
Many
surface waters require a coagulant such as alum for
effective treatment. After treatment and disinfection
with chlorine, the water can become aggressive. At present
there is no requirement to produce a water that is stable
(neither scale forming or corrosive) other than having
a pH in the range 6.5 to 8.5 (Australian Drinking Water
Guidelines, 1996).
This
paper describes water stability and shows how several
indices may be calculated to assess the likely corrosivity
of a water. An approach is also outlined whereby the
stability of a water can be determined following a particular
treatment regime and how the water can then be conditioned
to make it less aggressive, using a worked example.
2.0
WATER CHEMISTRY
The
following are the most important water quality parameters
affecting corrosivity, alkalinity, pH and calcium.
The
concentration of various constituents in a water can
be expressed in one of two ways: as the ion or "as is"
or as calcium carbonate. To convert from one form to
the other, the conversion factors listed in Table 1
are used. The chemical formula and formula weights of
these constituents are also listed. It is usual practice
to report the concentration of alkalinity as mg/L CaCO3
rather than "as is".
Positively
charged ions are called "cations" (e.g. calcium ion
or Ca++) and negatively charged ions, "anions" (e.g.
bicarbonate or HCO3). When we sum the concentration
of cations expressed as mg/L CaCO3, the total should
be the same as the sum of the anions, thus giving a
balanced water analysis. Table 2 is a list of chemicals
commonly used in water treatment processes, giving their
chemical formula, formula weight and factor to convert
to equivalent weight as CaCO3.
Table
1: Chemical Formula, Formula Weight and Factors to Convert
Concentration from "AS IS" to "CaCO3" for Common Cations
and Anions
Table
2: Chemical Formula, Formula Weights and Factors to
Convert Concentration from "AS IS" to "CaCO3" for Common
Chemicals Used in Water Treatment
Table
3: Water Analysis Waranga Channel at Rochester
As an example of how to apply these ideas, let's look
at a typical water supply: the Waranga Channel at Rochester,
Victoria (Table 3).
From
this analysis we can conclude:

The sum of the cations and the sum of the anions are
not the same: 87.0 v 82.7 mg/L as CaCO3. However,
in practice we generally label a water analysis balanced
if the (sum of cations / sum of anions) is within
+ or  5%. In this case it is + 5%, so it is acceptable.
 Alkalinity
of the water is 15.0 mg/L, which is relatively low,
suggesting this water will require supplemental alkali
addition for effective treatment when using alum.
 Total
hardness of the water ([Ca]+[Mg]) is equal to 36.2
mg/L as CaCO3; a soft water.
 The
total dissolved solids (TDS) level is approximately
125 mg/L ''as is" derived from the conductivity (EC)
of the water.
3.0
SOME BASIC CONCEPTS
To
determine what happens when we add various chemicals
to a water and to determine if it is corrosive or stable,
we first need to understand several important "rules":

Alkalinity is consumed when an acid is added to
a water. In this instance, an "acid" can be one
in the usual sense, such as sulphuric acid, or more
often than not, a metal cation. Hence, when we add
a cation to a water, say aluminium from alum, we consume
alkalinity. The same is also true when we add chlorine
gas.
 Alkalinity
increases when an alkali is added to a water.
 Carbon
dioxide is produced in a water when alkalinity is
consumed.
 Carbon
dioxide is destroyed when an alkali is added to a
water.
 The
pH of a water will decrease when carbon dioxide is
formed and will increase when CO2 is destroyed.
Knowing
the changes that various chemicals make to alkalinity
and carbon dioxide levels allows prediction of the pH
of the water as a result of different treatment regimes.
Table 4 gives the changes in alkalinity and carbon dioxide
that occur when various chemicals are added to a water.
Table
4: Alkalinity Consumed and Carbon Dioxide Produced per
mg Chemical Dose
We
can then calculate the pH of the water using Equation
(1) which applies for waters with a pH of between 4.5
to 8.5, at 25C.
4.0
WATER STABILITY INDICES
Although
a number of indices have been developed, none has demonstrated
the ability to accurately quantify and predict the corrosivity
or aggressiveness of a water. They can only give a probable
indication of the potential corrosivity of a water.
Experience has shown that if conditions encourage the
formation of a protective calcium carbonate film, then
corrosion will generally be minimized. Several models
and indices are available that use calcium carbonate
chemistry to evaluate water stability (Rossum and Merrill,
1983). Three calcium carbonatebased indices are described
and values for each are calculated with reference to
the Rochester water.
A
commonly used index is the Langelier Saturation Index
(LSI). This index provides a measure of the stability
of a water with respect to its degree of CaCO3 saturation.
If a water has a negative LSI value, it is undersaturated
with respect to calcium carbonate and is potentially
corrosive. Conversely, for waters with a positive LSI,
a protective layer of calcium carbonate can form as
the water is oversaturated with CaCO3 and the water
is scaling. Saturated water has a LSI of zero.
The
pH at which a water is saturated with CaCO3 is known
as the pH of saturation or pHs.
At
25oC and TDS less than 500 mg/L (the case for most Victorian
surface waters), the LSI can be calculated from Equations
(2) and (3):
In
practice, a water is considered to be potentially aggressive
if it has a LSI of less than 1.5.
The
Rochester water has a pH of 7.6, Ca of 7.9 mg/L "as
is" and an alkalinity of 15 mg/L as CaCO3. By substitution,
we can calculate the LSI to be 1.8. We can therefore
say that this water is potentially mildly corrosive.
Another
related parameter is the Ryznar Stability Index,
which is given by:
The
RSI value of a water should be less than 10 for it to
be considered to be stable and noncorrosive. For the
Rochester water, the RSI value is 11.2; again suggesting
this water is mildly corrosive.
The
Calcium Carbonate Precipitation Potential (CCPP)
is a more reliable water stability index to use since
this index provides a quantitative measure of the calcium
carbonate deficit or excess of the water, giving a more
accurate guide as to the likely extent of CaCO3 precipitation.
Previously, CCPP has been less frequently applied because
the longhand calculation procedure is timeconsuming
and quite tedious. The AWWA (1996) released a PCbased
spreadsheet program based on the Rothberg, Tamburini
and Winsor Model, which allows speedy calculation of
a number of corrosivity indices, including CCPP. The
program also allows calculation of the effects of various
chemical additions to a water.
A measure of the corrosivity of a water for different
values of CCPP is presented in Table 5.
Table
5: Corrosivity State of Water for Different CCPP Values
Another
method of determining the CCPP value is a graphical
procedure involving the use of water conditioning diagrams
originally developed by Caldwell and Lawrence (1953).
These diagrams can also be used to solve a wide range
of water treatment and conditioning problems (including
lime and limesoda softening).
Figure
1 is part of a CL Diagram drawn for water at 25oC with
a TDS of 40 mg/L. Although a unique CL Diagram should
be used for the temperature and TDS of the water in
question, in practice Figure 1 can be used over a range
of conditions without serious error.
To
determine the CCPP of a water, two parameters are calculated:
C2=(ALKCa), and ACIDITY. The approximate acidity
of a water may be found using Equation (5).
Figure
1: CaldwellLawrence Diagram : 25oC , 40 mg/L TDS.
Equation
(5) is valid for a water at 25oC, with a TDS of up to
200 mg/L. C2 and ACIDITY are expressed as calcium carbonate.
(ALKCa)
and ACIDITY are plotted on the CL Diagram and at the
point of intersection, we can then read off the Ca concentration
at saturation. The CCPP value is then found from Equation
6:
For
the Waranga Channel water: (ALKCa) = 15.0 19.7= 
4.7 mg/L CaCO3, and ACIDITY = 16.6 mg/L CaCO3 from Equation
(6).
From
Figure 1 (Point A), we can read off the calcium value
at saturation as 25.0 mg/L and hence:
The
precise value of CCPP is  4.7 mg/L. The CCPP value
suggests the water is "passive" and acceptable. Note
that agreement between the two techniques is approximate
only.
Table
6 summarizes the various water corrosivity indices considered
for the Rochester water, compared with accepted values
for a stable, noncorrosive water.
Table
6: Stability Indices for Rochester Water Compared with
Typical Values for Stable Water
Comparing
the calculated indices for the Rochester against acceptable
values, we can conclude this water is probably noncorrosive
to iron and steel.
5.0
WATER CONDITIONING
The water at Rochester has a true colour and turbidity
of 60 Pt/Co units and 40 NTU respectively. We would
anticipate that an alum dose of 5060 mg/L is required
for effective treatment in a conventional WTP. If we
add 50 mg/L of alum to the water, the pH will be too
low for effective coagulation. We must raise the pH
to typically 6.5 by adding an alkali in the form of
hydrated lime (calcium hydroxide), soda ash (sodium
carbonate) or caustic soda (sodium hydroxide).
Starting
with the raw water analysis, we can compute what will
happen as a consequence of dosing 50 mg/L alum with
supplemental alkali addition. Aluminium hydroxide is
formed when alum is added so there is no addition of
cations to the water. The alum consumes alkalinity and
to maintain a balanced water analysis, sulphate will
increase by an equivalent amount.
Assume
we add 12.2 mg/L lime. As a result of this addition,
we will increase both the calcium and alkalinity of
the water. The new water analysis will be:
 cations:
Ca= 19.7+2.5X12.2X40.1/74.1 = 36.2 mg/L CaCO3. (The
factor (40.1/74.1) gives the mg of calcium added per
mg Ca(OH)2 added as CaCO3, from Tables 1 and 2).
 anions:
alkalinity=15.00.45X50+12.2X1.35=9.0 mg/L, sulphate=4.2+0.45X50=26.7
mg/L CaCO3
Next,
we calculate the change to the CO2 level.

Initial CO2: from Equation (1). 7.6=log(2.2X106X15/CO2),
CO2=0.8 mg/L as is
 CO2
added from destruction of alkalinity by alum addition=50X0.4=20.0
mg/L
 CO2
destroyed by lime addition=12.2X1.19= 14.5 mg/L
 Final
CO2 concentration=6.3 mg/L
Now,
we calculate the pH of the water following chemical
addition:
pH=log(2.2X106X9.0/6.3)=6.5.
This
pH is satisfactory for alum coagulation and so the assumed
lime dose is adequate. If the calculated pH was higher
or lower than 6.5, then adjustments to the lime dose
assumed would be required until this value was obtained.
We
can also check the LSI of this water:
pHs = 11.5 log[14.5] log[9.0]=9.4, LSI = 6.5 9.4
= 2.9
This
water is now aggressive and if chlorinated will be corrosive
to the reticulation system. The solution to this potential
corrosivity problem is conditioning using posttreatment
alkali addition. Assume the water is disinfected by
adding 1.5 mg/L chlorine and that we will add 5.8 mg/L
of lime to condition the water. The same procedure as
above is again followed. The new water analysis will
be:

cations: Ca=36.2+2.5X5.8X40.1/74.1=44.0 mg/L CaCO3
 anions:
alkalinity=9.0+5.8X1.351.5X1.41=14.7mg/L
 chloride:
there will be an increase in the chloride level equivalent
to the decrease in alkalinity due to chlorine addition,
i.e. 2.1 mg/L.
Next,
we calculate the change to the CO2 level.
 Initial
CO2: 6.3 mg/L as is
 CO2
added from destruction of alkalinity by chlorine addition=1.5X1.24=1.9
mg/L
 CO2
destroyed by lime addition=5.8X1.19= 6.9 mg/L
 Final
CO2 concentration=1.3 mg/L
The
final pH of the conditioned water will be: pH=log(2.2X106X14.7/1.3)=7.4.
The
LSI of this water will be: pHs = 11.5log[17.6]log[14.7]=9.1,
LSI=7.49.1=1.7.
The
conditioned water now has a pH of 7.4 and a LSI of 1.7,
which is probably satisfactory from a corrosivity standpoint.
We can also check the CCPP value using the graphical
method outlined earlier. In this case:

(ALKCa) = 14.744.0=  29.3 mg/L CaCO3.
 ACIDITY
= 17.2 mg/L CaCO3 from Equation (6).
From
Figure 1 (Point B), we can read off the calcium saturation
value as 48.0 mg/L, and hence CCPP = (44.048.0) = 
4.0 mg/L CaCO3. The precise value of CCPP is  3.6 mg/L,
suggesting the water is "passive" and has been conditioned
to a satisfactory level. The assumed lime dose used
in our calculations is therefore sufficient. We can
also calculate the final sum of the cations and anions
and from this determine the TDS of our conditioned water.
The
new water analysis is shown in Table 7 and the TDS of
the conditioned water will be 164 mg/L as is. Note that
Mg, Na and K levels in the raw water all remain unchanged
as a consequence of water conditioning with lime. The
difference between the sum of the cations and the sum
of the anions is + 4.3 mg/L, which is the same as our
original water analysis and hence our calculations are
correct.
We
could also repeat this procedure using caustic soda
and soda ash for posttreatment pH adjustment.
Table
7: Water Analysis Waranga Channel at Rochester Following
Chemical Conditioning
6.0
CONCLUSIONS
Three
calcium carbonatedbased water stability indices have
been reviewed with a focus on the Langelier Saturation
Index and the Calcium Carbonate Precipitation Potential.
Methods of calculating these parameters have been outlined
as well as an approach to determining the impact different
treatment regimes can have on treated water quality.
The techniques outlined are straightforward and can
be readily adapted to a PCspreadsheet, providing the
WTP Operator with a powerful tool.
Operators
should anticipate that water stability indices will
be used more frequently in the future as the general
trend to improve treated water quality and reduce plant
operating costs continues. The methods presented in
this paper will hopefully contribute towards this goal.
7.0
REFERENCES
Australian
Drinking Water Guidelines (1996), National Water Quality
Management Strategy, NHMRC and ARMCANZ, Canberra
American
Water Works Association (1996), The Rothberg, Tamburini
and Winsor Model for Corrosion Control and Process Chemistry,
Denver, Colorado, USA.
Caldwell,
D.H. and Lawrence, W.B. (1953) "Water Softening and
Conditioning Problems: Solution by Chemical Equilibrium
Methods", Industrial and Engineering Chemistry, 45,
3, 535548.
Rossum,
J.R. and Merrill, D.T. (1983) "An Evaluation of the
Calcium Carbonate Saturation Indexes", Journal AWWA,
February, 95100.
8.0
NOMENCLATURE
Disclaimer
& Copyright
DOWNLOAD
(pdf
204K)
