Arab J Sci Eng (2011) 36:1171–1177

DOI 10.1007/s13369-011-0115-z

R E S E A R C H A RT I C L E - C H E M I C A L E N G I N E E R I N G

Mokhtar Mahdavi · Amir Hossein Mahvi · Simin Nasseri ·

Masoud Yunesian

Application of Freezing to the Desalination of Saline Water

Received: 22 August 2009 / Accepted: 25 May 2010 / Published online: 21 October 2011
© King Fahd University of Petroleum and Minerals 2011

Abstract Freezing is one of the methods used for desalination. Although different methods, like reverse osmo-
sis and distillation, are used commercially, freezing can be used for remote areas, cold regions and for small and
remote societies. We have investigated the method of crystallization by external cooling via non-direct contact
for desalination of seawater. The steps used for the freezing-melting (FM) process included: crystallization
(at −20◦ C), separation, surface washing and melting of the samples. Three samples were taken from Boshehr
beach. After three cycles of FM, the total dissolved solid (TDS) of the first sample reduced from 37,650 to
15,200, 5,059 and 1,403 mg/L, for cycles one, two and three respectively. For the second sample the TDS
reduced from 40,750 to 17,884, 5,584 and 1,487 mg/L and for the third sample with from 33,351 to 15,885,
4,850 and 1,345 mg/L. This showed that after three cycles of FM, potable water can be produced.

Keywords Desalination · Freezing · Non-direct contact freezing · Persian Gulf

M. Mahdavi · A. H. Mahvi (B) · S. Nasseri · M. Yunesian

School of Public Health and Institute for Environmental Research,
Tehran University of Medical Science, Tehran, Iran
E-mail: [email protected]

A. H. Mahvi
National Institute of Health Research, Tehran University of Medical Sciences,
Tehran, Iran

123
1172 Arab J Sci Eng (2011) 36:1171–1177

1 Introduction

Approximately 97.5% of the water on our planet is located in the oceans and is therefore classified as seawater.
Of the 2.5% of fresh water, approximately 70% is in polar ice and snow, and the remaining 30% is ground
water, river and lake water, and air moisture [1]. Today’s desalination technology for the production of fresh
water can be divided to two sections that depend on the mechanism used for the removal of salts. (a) Separation
processes: membrane processes that use electrical and mechanical forces, such as electro dialysis (ED) and
reverse osmosis (RO). (b) Thermal processes: changes in phase, such as distillation and freezing [2,3]. The
Danish physician Thomas Bartholinus (1616–1680) was the first to report that water obtained by melting ice
formed in seawater was fresh water. Almost at the same time, Robert Boyle reported the same observation
and foresaw the phenomenon as a source of fresh water [4]. At the end of the eighteenth century, the Italian
scientist Anton Maria Lorgna described a method to purify seawater and impure water by freezing and then
melting of the ice [5].
The method of water purification by freezing-melting (FM) was only possible in the coldest regions and
seasons, and not of practical interest, until the development of refrigerating machines. Interest in the process
of obtaining fresh water from seawater by freezing was revived in the late 1930s. The FM process was first
used commercially in the 1950s. Research in the 1960s and 70s in desalination, petroleum and food processing
applications provided many technical innovations [6].
Today, freezing processes have many applications in various areas of industry, such as the concentration
of fruit juice, dairy products, wastewater sludge and desalination. In addition, it can be used as a pretreatment
method for desalination of brackish and saline water to be further treated by other methods, like electro dialysis
and reverses osmosis [7]. In theory, the freezing process can separate water from saline and impure water by
freezing and crystallization of water. In ideal conditions the ice produced should be pure and without salts.
Fresh water can be produced by freezing of seawater, separation of crystals, surface washing and thawing of
crystals [8,7,6,5].
The three broad classes of FM processes are: direct contact freezing, non-direct contact freezing and
vacuum freezing. All of these methods involve crystallization, separation, surface washing and thawing of
the crystals. Separation of ice crystals can be performed by gravity drainage, centrifuge, filtering, and wash
columns [8,7,6,5,9].
Experimental results show that the melted ice water from a single freezing, without a wash step, has three
to six times less salt content than the feed water [10]. Nicholas showed that for frozen source water with
3,000 mg/L NaCl at an ambient temperature of −15◦ C, removal of 80% of the salt was possible after melting
9% of the ice [11].
A combined RO/direct contact freezing system provides an efficient system to reduce the problem of dis-
posal of the rejected brine from inland desalination plants. Results show that the combined system can reduce
the energy consumption by about 13% and 17%, compared with separate RO and direct contact freezing plants
respectively [7]. Although the FM process isn’t widely used commercially, this process has advantages over
currently used processes. Perhaps the greatest potential advantage of desalination by freezing is the very low
energy required compared with other desalination processes. This is because the latent heat of fusion of ice is
only one-seventh the latent heat vaporization of water [12].
The freezing process also has the advantage of a low operating temperature, which minimizes scaling and
corrosion problems. Furthermore, inexpensive plastics or other low cost materials can be used at low temper-
atures [15,14,13]. Disadvantages of the freezing process include: higher initial investment and capital cost;
higher operation costs during ice separation; and the persistence of the primary odor and taste of the water
[16,17].
In this study, the effect of the freezing process on the removal of cations and anions (bicarbonate, sulfate,
chloride, calcium, magnesium, sodium and potassium), total dissolved solid (TDS) and electrical conductivity
(EC) of Persian Gulf water is investigated.

2 Materials and Methods

Three water samples from different locations at Boshehr beach in the Persian Gulf were tested. The sample
volumes were 50 L. Physicochemical characteristics of these three sample, such as bicarbonate, sulfate, chlo-
ride, calcium, magnesium, sodium, potassium ion concentration, TDS and EC, were measured according to
standard methods for examination of water and wastewater [18] (Table 1). The FM cycle was performed on
the samples according to Fig. 1. The method used for desalination was crystallization by non-direct contact

123
Arab J Sci Eng (2011) 36:1171–1177 1173

Table 1 Composition of Persian Gulf samples before and after freezing

Sample HCO−1
3 SO−2
4 Cl−1 Ca+2 Mg+2 Na+ K+ Total TDS EC Sample
(mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) hardness (mg/L) (µS cm−1 ) volume
(mg/L) (L)
1 B.F. 164.7 3,410 16,214 488.8 876 10,602 344 4,794.5 37,650 51,500 50
A.F.F. 69 1,906 8,735 326 604 5,047 228 3,294 15,200 29,500 9.58
A.S.F. 17 691 2,580 201 315 1,102 110 1,794 5,059 11,850 1.87
A.Th.F. 7 248 512 68 102 233 45 588 1,403 2,751 0.4
2 B.F. 182 3,850 17,812 550 962 11,450 497 5,304 40,750 57,103 50
A.F.F. 73 2,040 9,260 346 625 5,180 318 3,427 17,884 32,180 9.37
A.S.F. 19 776 2,685 207 315 1,191 129 1,809 5,584 12,992 1.77
A.Th.F. 10 280 483 72 101 240 54 594 1,487 2,920 0.395
3 B.F. 145 3,002 15,787 392 781 9,110 278 4,182 33,351 46,108 50
A.F.F. 58 1,683 8,718 262 546 4,555 184 2,893 15,885 30,010 10
A.S.F. 12 642 2,609 167 289 1,045 90 1,602 4,850 11,250 1.97
A.Th.F. 5 224 470 58 97 230 38 543 1,345 2,550 0.4
B.F. Before Freezing cycle, A.F.F. After First Freezing cycle, A.S.F. After Second Freezing cycle, A.Th.F. After Third Freezing
cycle

Distilled water
(4°C)

Freezing in
freezer Ice crystals
(At -20°C) And Washing Melting
Saline Fresh
Saline water process of process at
water water
Ice crystals ambient
temperature

Drainage of
Saline water drainage wash water

Results
Fig. 1 Flow diagram of the freezing-melting cycle

external cooling of batches. The freezer used in this study was a Pars-PAMCO-FRZNF170. The energy con-
sumption of the freezer was 0.1 kW/h. The FM cycle was repeated three times for each sample. The separation
of crystals was performed by drainage of concentrated water (rejected water) and the washing of the ice crystals
from the deposited salts was accomplished by simple flowing of cold distilled water over the desalted ice. The
shape of the crystals formed was spherical. For the first cycle of FM, the duration of freezing was 275 min and
200 mL of cold distilled water was used for surface washing. In addition, about 15–25% of the melted crystals
were used for washing the surface of crystals. The crystals were then melted at ambient temperature (25◦ C)
and physicochemical analysis undertaken on the water produced.
For the FM cycle, the freezing was performed on the water produced from the first cycle. In this cycle, the
duration of freezing was 220 min. The process was the same as for the first FM cycle, although only 60 mL
of cold distilled water was used for surface washing of the crystals. For the third cycle, the water produced
from the second cycle was frozen again and about 20 mL of cold distilled water was used for surface washing.
The duration of freezing was 190 min. The same steps were undertaken for the three Persian Gulf seawater
samples.

3 Results

Table 1 shows the composition and quantity of the Persian Gulf samples before and after the three FM cycles.
After the first cycle of freezing, all three samples showed a significant reduction in the amount of dissolved
salts. The TDS of the three samples after this cycle had reduced by 60, 56 and 51% and the EC by 43, 44 and
35%, for samples 1, 2 and 3, respectively. However, at this stage (the first freezing cycle) the water produced
was not suitable for drinking as a matter of high TDS with respect to drinking water standards.

123
1174 Arab J Sci Eng (2011) 36:1171–1177

Fig. 2 Removal efficiency of the first Persian Gulf sample for three FM cycles

Fig. 3 Removal efficiency diagram of the second Boshehr beach sample for three freezing cycles

Fig. 4 Removal efficiency diagram of the third Boshehr beach sample for three freezing cycles

After the second FM cycle, the amount of TDS, EC and minerals had reduced by between 3–4 times
compared with the original sample. The TDS of the samples had reduced by 67, 69 and 69% for samples 1, 2
and 3, respectively. With two cycles of FM, potable water was still not obtained. After the third FM cycle, the
TDS of the samples reduced by 72, 73 and 73% and potable water was produced. It should be mentioned that
all of the removal percentages are with respect to its last cycle.
Figure 2 shows the removal efficiency of three cycles of FM. The water production rate in this study was
15–20% per cycle of FM.
After three cycles of FM, the TDS of sample 1 reduced from 37,650 to 15,200, 5,059 and 1,403 mg/L
for cycles 1, 2 and 3, respectively. For the second sample the TDS reduced from 40,750 to 17,884, 5,584 and
1,487 mg/L and for third sample from 33,351 to 15,885, 4,850 and 1,345 mg/L. The overall removal efficiency
for any of three saline water sample was in the range of 96%.
The freezing duration for the first cycle was prolonged because the concentration of salt and minerals was
high. The duration of the freezing process reduced as the TDS decreased. The freezing duration was 220 min

123
Arab J Sci Eng (2011) 36:1171–1177 1175

Fig. 5 Mean removal efficiency of minerals, TDS and EC of three mixed samples

Fig. 6 Trend of TDS and EC after three FM cycles of Boshehr beach samples

Table 2 Energy consumption for produced water after three freezing cycles (kW/L) and duration of crystallization

Sample 1 Sample 2 Sample 3

Duration of crystallization in first freezing cycle (min) 275 275 275
Duration of crystallization in second freezing cycle (min) 220 220 220
Duration of crystallization in third freezing cycle (min) 190 190 190
Total time of freezing for produced water (first, second and third freezing cycle) per hour 11.47 11.47 11.47
Energy consumption for produced water after three cycles of freezinga (kW/L) 0.456 0.450 0.456
a 0.1 kW/h ∗ 5.6 h ∗ 0.4 L = 0.456 kW/L

Table 3 Comparison of energy consumption for different desalination methods, kW h/m3 [19,20]

MSF RO ED FM process in this study

kW h/m3 35 0.4–7 1 450a
Raw water TDS, mg/L 30,000–100,000 1,000–45,000 100–3,000 37,650
a Although energy consumption is high, this system is very simple and some development is necessary for the prevention of
energy loss. The TDS of the seawater used in this study is high, and it has a significant effect on energy consumption

for the second freezing process and 190 min for the third. In the first and second cycles of FM, most of the
removal efficiency was due to Na+ cations and HCO3− anions, while for the third cycle of FM the removal
efficiency was mostly due to Na+ cations and Cl− anions. In this study, three FM cycles were needed to produce
potable water (according to Iranian standards). After two FM cycles, the seawater changed to brackish water
with low TDS. Figures 2, 3 and 4 show the removal efficiency after the three cycles of FM. Figure 5 shows
the mean effect of freezing on the reduction of TDS and EC. Figure 6 shows the mean removal efficiency
of TDS and EC for total and mixed of three samples. Tables 2 and 3 show the freezing duration and energy
consumption.

123
1176 Arab J Sci Eng (2011) 36:1171–1177

4 Discussion

With an increase in salt concentration the freezing point of water decreases, so a lower temperature, or a
longer time at a constant temperature, is needed for crystallization. For the first freezing, the time needed for
crystallization at −20◦ C was 275 min. For the second freezing, because of a decrease in TDS, the time needed
reduced to 220 min and for the third freezing it reduced further to 190 min. In this study, because of fast
freezing at −20◦ C the crystal shapes were spherical. For the first FM cycle, the FM was not very effective
in removing dissolved solids because the samples had a high concentration of salt. However, for the second
cycle, the water had the composition of brackish water and the freezing resulted in more effective removal of
salt. For the first FM cycle the mean rate reduction of TDS was about 60%, in the second freezing cycle it was
about 67%, and in the third cycle it was about 72%. The amount of water produced related to the amount of
crystallization. When the duration of freezing is longer, more crystals will be produced, although the quality
of produced water will be lower because salts begin to freeze and are trapped among crystals layers. In this
study, after two freezing cycles the seawater changed to low TDS brackish water. Therefore, this could be used
as a pretreatment for desalination of seawater with further treatment by another method, such as RO or ED.
The typical energy consumption in seawater reverse osmosis (RO) plants operating at 40–45% product
water recovery and with energy recovery from the high pressure reject stream is currently about 3–4 kW/m3
[21].The average power consumed by the high pressure pump in Parametric analyses of energy consumption
and loss in seawater reverse osmosis (SWRO) plants using energy recovery devices is in the range 5.56–
7.93 kW h/m3 [22]. SWRO with energy recovery is assumed to consume about 5.2 kW/m3 electrical energy
[23]. Energy consumption for the first sample in the three FM cycles was 456 kW/m3 , for second sample it
was 450 kW/m3 and for the third sample it was 456 kW/m3 .

5 Conclusions

After three FM cycles of Persian Gulf seawater, potable water was produced. However, the water produced
had low pH, calcium and alkalinity, so further processing would be necessary for its consumption. This very
simple method can produce fresh water in remote and small areas. With one or two freezing cycles the Persian
Gulf seawater showed a significant reduction in TDS, so it could be used as a pretreatment method for other
methods of desalination. The salt concentration of the rejected water is about 10%. Therefore; discharge of
this saline water into water resources, such as the sea, would not have a severe effect. The concentration of
rejected water for other desalination processes is high (about 40%). With an increase in salt concentration
the crystal growth rate decreases at a constant temperature and longer time is needed for crystallization. The
energy consumption in this study was very high, although with further development, the energy loss could be
reduced and energy reclamation increased to improve the process.

References

1. Jayh, L.J.: Water Encyclopedia. Wiley Interscience Press (2006)

2. AWWA.: Water Desalting Planning Guide for Water Utilities, 1st edn. Wiley, New York (2004)
3. Spiegler, K.S.: Principle of Desalination. Academic Press, New York (1980)
4. Nebbia, G.; Menozzi, G.N.: Early experiments on water desalination by freezing. Desalination 5, 49–54 (1968)
5. Shafiur, M.; Ahmed, M.: Freezing-melting process and desalination: I. review of the state-of-the-art. Sep. Purif. Rev. 35,
59–96 (2005)
6. Rosen, J.: Freeze concentration beats the heat. Mech. Eng. 112, 46–50 (1990)
7. Madani, A.A.; Aly, S.S.: A combined RO/freezing system to reduce in land rejected brine. Desalination 85, 179–195 (1989)
8. House, A.: Desalination for water supply FR/RO. Found. Water Res. 1–22 (2006)
9. Thijssen, H.A.C.: Freeze Concentration of Food Liquids. Food Science Technology, Washington (1975)
10. Phillip, C.: Desalination by natural freezing. Desalination 13, 147–157 (1973)
11. Nicholas, B.; Davids, S.: Laboratory investigation on freeze separation of saline mine waste water. Cold Reg. Technol. 48,
239–247 (2006)
12. Brian, P.L.T.: Potential advantages and development problems in water desalination by freezing. Chem. Eng. 78, 191–197
(1971)
13. Johnson, W.E.: State-of-the-art of freezing processes, their potential and future. Desalination 19, 349–358 (1976)
14. Heist, J.A.: Freeze crystallization. Chem. Eng. 86, 72–82 (1979)
15. Agnew, C.; Anderson, E.: Water Resources in the Arid Realm. Routledge, London (1992)
16. Braddock, R.J.; Marcy, J.E.: Quality of freeze concentrated orange juice. J. Food Sci. 52, 159–162 (1987)
17. Wiegandt, H.F.; Von Berg, R.L.: Myths about freeze desalting. Desalination 33, 287–297 (1980)

123
Arab J Sci Eng (2011) 36:1171–1177 1177

18. APHA.: Standard Methods for the Examination of Water & Wastewater. APHA Publication, Washington (2005)
19. Zhou, Y.; Tol, S.J.R.: Evaluating the costs of desalination and water transport. 18th IDA Worldwide Desalting Plants Inven-
tory, Working paper FNU-41 revised (2004)
20. Laborde, H.M.; Franca, K.B.; Neff, H.; Lima, A.M.N.: Optimization strategy for small-scale reverse osmosis water desali-
nation system based on solar energy. Desalination 133, 1–12 (2001)
21. Singh, R.: Sustainable fuel cell integrated membrane desalination systems. Desalination 227, 14–33 (2008)
22. Farooque, A.M.; Jamaluddin, A.T.M.; Al-Reweli, A.R.; Jalaluddin, P.A.M.; Al-Marwani, S.M.; Al-Mobayed, A.A.; Qasim,
A,H.: Parametric analyses of energy consumption and losses in SWCC SWRO plants utilizing energy recovery devices.
Desalination 219, 137–159 (2008)
23. Darwish, M.A.; Alotaibi, S.; Alfahad, S.: On the reduction of desalting energy and its cost in Kuwait. Desalination 220,
483–495 (2008)

123