Desalination 184 (2005) 165–171

How many ppm? The unresolved problem of acid dosing

in the reverse osmosis plant design

Giorgio Migliorini*, Carlo Cattani, Elena Luzzo

Fisia Italimpianti, Genoa, Italy
Tel. þ39 010 6096 429; Fax þ39 010 6096 410; email: giorgio_migliorini@fisiait.com

Received 20 January 2005; accepted 21 February 2005

Abstract
Different membrane manufacturer software are available for the designers of RO plants. Software releases
are reissued periodically and distributed to the designers to make simulation of membrane behaviour. If you are
familiar with these software you know that, a part the different products characteristics, depending on the
membrane specification, the acid dosage requests to modify the pH of the feed water are very different for each
supplier and often also for each software release. In the paper, using a data input a typical sea water
Mediterranean Sea analysis, the membrane software are used to simulate different design conditions and the
acid dosage figures are compared with those obtained from an original model developed by the authors, based
on the carbonate salts equilibrium.

Keywords: Reverse osmosis; pH adjustment; Acid dosing; Sea water salinity; Carbonate system; Carbonic acid
equilibrium

1. Introduction the desired value of this parameter: in other

One of the main problem with that the
reverse osmosis plant process engineer has
to deal, when he has to design a new plant,
is the chemical pre-treatment sizing.
In particular he has to define the pH of the
feed to avoid the salts precipitation phenomenon
and as consequence the acid flow rate to reach
*Corresponding author.
words he has to define ‘‘how many ppm’’
The scale control involves the carbonic
species equilibrium calculation that is directly
correlated to the pH trough the well known
equilibrium reactions.
The software releases normally available
by the membrane or chemical products man-
ufacturer make the calculation starting from
the feed pH value and the carbonates or
bicarbonates content.

Presented at the Conference on Desalination and the Environment, Santa Margherita, Italy, 22–26 May 2005.
European Desalination Society.
0011-9164/05/$– See front matter Ó 2005 Elsevier B.V. All rights reserved
166 G. Migliorini et al. / Desalination 184 (2005) 165–171

In the following sections an overview of 2.1. Temperature influence

the different releases approaches is reported The reference acid used for pH adjust-
and compared with the classic equilibrium ment is sulphuric acid at normalized concen-
model to identify the correct acid dosing rate. tration of 100% and the calculations have
been made in the sea water temperature
2. Software different approaches range 20–35 C.
In the following diagrams (Figs. 1–3) the
Software of three membrane manufac-
acid dosing rate versus temperature to modify
turers is used for the comparison.
the feed pH from the raw water value of 8.1
To compare the results the following
to feed water value of 7.5 is reported for the
standard analysis has been assumed as
different manufactures.
reference:
The curves for manufacturers number 2
and 3 have the same profile, but the manu-
Reference analysis facture 2 curve slope is greater than the slope
Sodium (ppm) 11,093 of manufacturer 3 curve.
Potassium (ppm) 507 The dosing rate is decreasing with tem-
Calcium (ppm) 426 perature increasing.
Magnesium (ppm) 1531
The manufacturer 1 dosing rate is from 2
Bicarbonates (ppm) 188
Sulphates (ppm) 3698 to 2.2 times greater than the others and is
Chlorides (ppm) 19,957 increasing with temperature.
pH 8.1
Total dissolved 37,400
solids (1) (ppm)
2.2. Final pH influence
The acid dosing rates vs. final pH starting
The following approaches are used by from initial pH value of 8.1 are reported at
different manufacturers: two different temperatures 20 and 35 C in
Figs. 4, 5 and 6.

Manufacturer/chemical Input Parameter calculated

producer
1 Sea water analysis CO2 dissolved
Element age
Bicarbonate content
Carbonate content
pH
2 Sea water analysis Carbonate content or Bicarbonate content
Fouling factor (year) CO2 dissolved
Bicarbonate or carbonate content
pH
3 Sea water analysis The pH is recalculated to match bicarbonate
Fouling factor (%) and carbonate content
Bicarbonate or carbonate content
pH
 G. Migliorini et al. / Desalination 184 (2005) 165–171 167

10 T 20°C
Acid dosage

H2SO4 Dosage 8
9 Feed CO2
(ppm)

7,5 T 35°C
8

pH
7 7
6
6,5
20 25 30 35
0,00 10,00 20,00 30,00 40,00 50,00 60,00 70,00
Temperature (°C)
H2 SO4 (100%) Dosage (ppm)

Fig. 1. H2SO4 (100%) dosage vs. T( C) pH 8.1–7.5— Fig. 6. H2SO4 (100%) dosage vs. final pH—manu-
manufacturer 1. facturer 3.

Manufacturer 1: The dosing rate is increas-

Acid dosage

27
22 H2 SO4 Dosage
(ppm)

17
12
Feed CO2 ing whereas the final pH is decreasing; the
7
2 curve is concave; the dosing rate is higher
20 25 30 35 for lower temperatures.
Temperature (°C)
Manufacturer 2: The dosing rate is increas-
Fig. 2. H2SO4 (100%) dosage vs. T( C) pH 8.1–7.5— ing whereas the final pH is decreasing; the
manufacturer 2. curve is convex; the dosing rate is higher for
higher temperatures.
H2SO4 Dosage Manufacturer 3: The dosing rate is increas-
10
Feed CO2 ing whereas the final pH is decreasing; the
Acid dosage

curve is concave; the dosing rate is higher

(ppm)

9
8 for lower temperatures.
7 The behaviour is similar to the Manufac-
20 25 30 35
Temperature (°C)
turer 1.

Fig. 3. H2SO4 (100%) dosage vs. T( C) pH 8.1–7.5—

manufacturer 3. 3. Carbonic acid equilibrium
The pH of most natural waters included the
8
T 25°C
sea water is generally assumed to be controlled
7,5 T 35°C by the carbonic acid system [1,2,5,6].
pH

7 The applicable equilibrium reactions are:

6,5
0 10 20 30
H2 SO4 (100%) Dosage (ppm)
40 50 60
CO2 þ H2 O $ ðH2 CO3 Þ $ Hþ þ HCO3

Fig. 4. H2SO4 (100%) dosage vs. final pH—manu- HCO3 ! Hþ þ CO3

facturer 1.
The CO2 =HCO3 =CO2 3 equilibrium in sea
8 T 20°C water can be calculated starting from Alkali-
7,5 T 35°C nity and pH values trough the calculation of
pH

7
the following equations:
6,5
0 10 20 30 40 50 60 70 Kw ¼ ½Hþ  ½OH  .

H2 SO4 (100%) Dosage (ppm)

½Hþ  ½HCO  
.
Fig. 5. H2SO4 (100%) dosage vs. final pH—manu- K1 ¼
facturer 2. ½CO2 
168 G. Migliorini et al. / Desalination 184 (2005) 165–171

½Hþ  ½CO¼ 
.
½Alk þ ½Hþ  
Kw
K2 ¼ ½Hþ  gr  ions
½HCO3  ½HCO3¼  ¼ .
2K 1
1 þ þ2
½H 
where:
K2 gr  ions
½CO3  ¼ . ½HCO3  .
Kw ¼ ionisation product of water ½Hþ  1
6:0486T4471:33 0:017053 ðTw þ273Þ
¼ 10 w þ273
.

½Hþ 
½CO2  ¼ . ½HCO   gr  moles
.
K1
K1 ¼ first dissociation constant
14:8435T3404:71 0:032786 ðTw þ273Þ For a defined pH and alkalinity or carbonate
¼ 10 w þ273
.

content is possible to determine the bicar-

K2 ¼ second dissociation constant bonate content that contribute to the CO2
6:498T2909:39 0:02379 ðTw þ273Þ
formation.
¼ 10 w þ273
.

In conclusion is clear that the parameters

carbonates, bicarbonates, CO2 and pH are cor-
The following relations have to be taken into related and, if three parameters are defined, the
account: fourth is consequently determined.
The acid addition can be defined accord-
total alkalinity as CaCO3 gr  equiv ing to the previous defined equations.
½Alk ¼ .
50  1000 1

gr  ions 4. Calculation model

½Hþ  ¼ hydrogen ions concentration .
1 The model developed has been already
described in the previous work of same authors
¼ 10pH [7,3,4]. In particular for each stream the follow-
  ing parameters are calculated:
 ppm HCO
 .gr  ions  ionic salt balance
HCO3¼ 3
.
61  1000 1  carbonic species equilibrium pH
   saturation pH
 ¼
 ppm CO¼ 3 .gr  ions In the case that an energy recovery device is
CO3 ¼ .
60  1000 1 used the raw water analysis input is corrected
taking into account the PX salt losses to
ppm ½CO2 
. gr  ions make the exact feeding conditions at the
½CO2  ¼ .
44  1000 1 membranes inlet:
The chemicals dosing rate for the feed water
The following ionic balance can be also written: are calculated starting from the raw water ana-
lysis and for each step of the treatment the com-
½Alk þ ½Hþ  ¼ ½HCO3   þ 2½CO3 ¼  þ ½OH  plete carbonic species equilibrium is calculated.
The main feature of the new calculation
And using the ionisation and dissociation system is that the water quality in terms of
constants defined above, the following final pH, TDS and carbonic species during the
relations are obtained: various phases of the process can be followed
 G. Migliorini et al. / Desalination 184 (2005) 165–171 169

and the PX effect on the system performances
is taken into account, carrying out a new
approach to the RO plant design software.
5. Calculation model results
In the following Figs. 7 and 8 the original
model calculation results are reported compared
with the manufacturers models; as usual ‘‘in
medio stat virtus’’.
In other words the theoretical model in a
good agreement with the manufactures 1 and
3, whereas the manufacturer no. 2 that is com-
pletely out of the normal figures.
The consequence is that the acid
consumption is not corrected; this fact implies

8,1

7,9

7,7

7,5
Manufacturer 1
pH

7,3 Manufacturer 2
Manufacturer 3
7,1 Original Model

6,9

6,7

6,5
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70

H2 SO4 (100%) Dosage (ppm)

Fig. 7. H2SO4 (100%) dosage vs. final pH—T 20 C.

8,1

7,9

7,7
Manufacturer 1
7,5 Manufacturer 2
Manufacturer 3
pH

7,3 Original Model

7,1

6,9

6,7

6,5
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70

H2 SO4 (100%) Dosage (ppm)

Fig. 8. H2SO4 (100%) dosage vs. final pH—T 35 C.

170 G. Migliorini et al. / Desalination 184 (2005) 165–171

that the chemical dosing section is oversized with authors. The manufacturer 1 curve profile is
a consequent plant extra cost. in good agreement with the theoretical model,
Moreover the water production cost can whereas the manufacturer 2 curve presents some
be estimated more than the real one. differences.
In particular for Manufacturer 2 carbo-
nate species are present for pH around 8.
6. What is the true? The consequence of the use of a not correct
In the last Figs. 9, 10 and 11 the classical model is already evidenced in the previous
representation of carbonic acid equilibrium are figures, because the acid dosage estimation is
shown for different manufacturers and for the greater than the necessary.

1,00
0,95
0,90
0,85
Percentage of carbonic species

0,80
0,75
0,70
0,65
0,60
CO2
0,55
0,50 Bicarbonates
0,45 Carbonates
0,40
0,35
0,30
0,25
0,20
0,15
0,10
0,05
0,00
6 6,5 7 7,5 8 8,5
pH

Fig. 9. Carbonic species equilibrium at 20 C—manufacturer 1.

1,00
0,95
0,90
0,85
Percentage of carbonic species

0,80
0,75
0,70
0,65
0,60 CO2
0,55 Bicarbonates
0,50
Carbonates
0,45
0,40
0,35
0,30
0,25
0,20
0,15
0,10
0,05
0,00
6 6,5 7 7,5 8 8,5
pH

Fig. 10. Carbonic species equilibrium at 20 C—manufacturer 2.

 G. Migliorini et al. / Desalination 184 (2005) 165–171 171

1,00
0,95
0,90

Percentage of carbonic species

0,85
0,80
0,75
0,70
0,65
0,60
CO2
0,55
0,50 B icarbo nates
0,45 Carbo nates
0,40
0,35
0,30
0,25
0,20
0,15
0,10
0,05
0,00
6 6,5 7 7,5 8 8,5
pH

Fig. 11. Carbonic species equilibrium at 20 C: theoretical model.

7. Conclusions [3] American Water Works Association Research

Foundation. Water Treatment—Membrane
The software produced by manufacturers can Processes, Lyonnaise des Eaux, Water Research
be used with particular attention; the theory can Commission of South Africa, McGraw Hill, New
not be missed and the feed back of the mem- York, 1996.
brane suppliers have to disclaim to the engineer- [4] Josè Antonio Medina San Juan, Desalacion de
ing companies to allow the best design of RO agues salobres y de mar—Osmosis Inversa,
plants to avoid over sizing of chemical pre- Ediciones Mundi–Prensa, 2000.
[5] H.E. Homig, Sea Water and Sea Water Distilla-
treatment sections with unnecessary extra cost.
tion, Vulkan Verlag, Essen, 1978.
[6] High Temperature Scale Inhibitors for Sea
References Water Distillation, Watson Desalination Consul-
tants, October 1979.
[1] American Water Works Association. Water Qual-
[7] G. Migliorini and E. Luzzo, Seawater reverse
ity and Treatment, 5th ed., McGraw Hill, New
osmosis plant using the pressure exchanger for
York, 1999.
energy recovery: a calculation. Desalination,
[2] American Water Works Association. Water Treat-
165 (2004) 289–298.
ment Plant, 3rd ed., McGraw Hill, New York, 1998.