Desalination 235 (2009) 340351

0011-9164/09/$ See front matter 2008 Elsevier B.V. All rights reserved

*Corresponding author.

A simulation model and parametric study of MEDTVC process

R.K. Kamalia*, A. Abbassia, S.A. Sadough VaniniaaMechanical Engineering Department, Amir Kabir University of Technology,

Hafez Ave., P.O. Box 15875-4413, Tehran, IranTel. +98 (21) 64543409; Fax +98 (21) 66419736; email: kouhikamali@aut.ac.ir

Received 4 January 2007; accepted revised 28 January 2008

Abstract

The present work summarizes the MEDTVC (multi effect desalination with thermal vapor compression)technique associated with the state of the art of modern desalination. In addition, a computer simulation model forall types of evaporation processes is presented. This program provides engineers with cost-effective tools fordesigning, developing and optimizing thermal desalination plants. It is the objective of this article to develop amathematical model which would predict the influence of all factors on heat transfer coefficients, temperature andpressure, total capacity and performance ratio of the system under design and operating conditions. The transientnature of temperature during the seasons is modeled by ordinary differential equations based on mass and energybalance. Heat exchangers and thermo-compressor are designed based on the results of mass and energy balance.The validated model is further used to test the effect of variations in certain parameters in the process in order toinvestigate their influence on the total capacity of the plant. By means of parametric study, the computer simulationtool developed will help designers to achieve the best setting for the desalination process to minimize energyconsumption. The comparison between the simulation results and experimental data well proves the program validity.

Keywords: Desalination; Multi-effect; Thermal vapor compression; Parametric study; Optimization

1. Introduction

The need for high quality water has signifi-cantly increased during the second half of the lastcentury. It has been a complex task to develop aneffective process without actual testing whichusually requires costly test procedures. The de-salination industry is very important for several

countries and zones around the world, especiallythe countries around the Persian Gulf, such as Iran.Expansion in desalination industry is associatedwith reduction in power consumption. Today, ther-mal desalination processes account for more than65% of the production capacity of the desalina-tion industry [1].

The authors very strongly believe that ther-mal desalination processes, especially multi-ef-fect desalination (MED), is one of the best meth-

doi:10.1016/j.desal.2008.0 .011 9

R.K. Kamali et al. / Desalination 235 (2009) 340351 341

ods for desalting seawater to achieve very lowconductivity which is very useful in power plants.For this reason, a general computer code for MEDtype of desalination has been developed and iscurrently used by a number of Iranian companies.El-Dessouky and Ettouney [2], Jernqvist et al. [3]and Ettouney [4] developed a simulation code forthe MED system with shell and tube evaporators.The present work deals with both shell and tubeand plate type evaporators and in addition, thermo-compressor and ejectors are designed too.

2. Mathematical modeling

A schematic of the MEDTVC system isshown in Fig. 1. The system consists of severalevaporators, a condenser, and a thermo-compres-sor. In each effect, two phase flow inside theevaporators is modeled by mathematical equationsto account the pressure drop and flow specifica-tions [57].

Fig. 1. Schematic of the MEDTVC system.

In mathematical modeling, at first mass andenergy balance equations were been developedfor the system and then heat exchangers, thermo-compressor and ejectors were designed based onthe results of mass and energy balance.

2.1. Mass and energy balance

2.1.1. First effect mass and energy balance

As shown in Fig. 2, the mass and energy bal-ance for thr first effect is as follows:

1 1 1B F D (1)

0 rD S D (2)

11B FX B X F (3)

0 1 11 1

,s p F F FD L F C T X T TD Lu u u

u (4)

342 R.K. Kamali et al. / Desalination 235 (2009) 340351

Fig. 2. Effect as a control volume.

The heat capacitance was estimated as a func-tion of temperature and salinity and defined as[8].

2.1.2. Mass and energy balance for other ef-fects

The vapor generated in each effect was passedthrough demisters and entered the next effect toheat transfer to the feed seawater. Then condensedwater was produced. Therefore, mass and energybalance for each effect can be considered as fol-lows:

(5)

1Bn Fn n n BnX X F D X u u (6)

1

1 11

1

n

n n i i Bi

n n n F n F n n

D L F D Cp

T T F Cp T T D L

u u u u u u

(7)

The total desired product can be considered asfollows:

1

n

ii

D D

(8)

2.1.3. Mass and energy balance for the con-denser

As shown in Fig. 3, part of vapor generated inthe last effect is passed through condenser and iscondensed by seawater intake to the system. Inthis process, seawater which is used as feed waterof effects is preheated. The extra seawater whichis passed inside the condenser should be rejected.The mass and energy balance for the condensercan be calculated as follows:

f n rD D D (9)

tMc F R (10)

1

n

t ii

F F

(11)

f n F F swD L Mc Cp T Tu u u (12)Latent heat is estimated as a function of tem-

perature.

Fig. 3. Condenser as a control volume.

1

n

n i ii

B F D

R.K. Kamali et al. / Desalination 235 (2009) 340351 343

It should be noted that the following assump-tions are applied to solve these equations:1) To achieve the optimum operating conditions,

the temperature difference between all units isassumed to be equal [9].

1S FT TTN

'

(13)

1 ST T T ' (14)

1 2i iT T T i N ' (15)

2) To avoid the sediment in each effect, the ratioof the total feed per distilled water is set basedon Eq. (16) [10]:

1 1

1 1 1

B

n B F

F XD X X

(16)

2.2. Evaporators and condenser design

Evaporators are the main part of desalinationunits. Therefore, it is very important to designthem as cost-effective and more efficient.

Two types of configurations applied in desali-nation plants are conventional shell and tube andplate type evaporators. Plate type evaporators isa new technology and they have a compact andportable system.

2.2.1. Shell and tube evaporators

To design shell and tube evaporators, someparameters should be calculated, such as tube size(diameter and length), number of passes and num-ber of tubes.

Fig. 4 illustrates the design algorithm whichwas used to determine these parameters and fi-nalize the evaporator design.

2.2.2. Plate type evaporators

For being compact, easy to clean, efficient andmore flexible, the gasket plate type heat exchang-ers are widely employed in desalination processes.

Fig. 4. Shell and tube evaporators design algorithm.

To design a plate type heat exchanger, some pa-rameters have to be calculated number ofplates, plate size, chevron type, size of gap be-tween the plates and so on. The design algorithmfor the plate type evaporator is illustrated in Fig. 5.

2.2.3. Shell and tube condenser

Part of vapor generated in the last effect ispassed through the condenser and condensed bythe seawater intake to the system inside the tubes.

344 R.K. Kamali et al. / Desalination 235 (2009) 340351

Fig. 5. Plate type evaporators design algorithm.

The design algorithm for the condenser is illus-trated in Fig. 6.

The condenser is usually shell and tube typein the MED system which has the same designalgorithm as shell and tube evaporators in general,but the effect of non-condensable gases on theheat transfer coefficient should be considered asresistance during the design algorithm, and, inaddition, the allowable pressure drop for seawateras single phase liquid inside the tubes is consideredaccording the maximum allowable velocity insidethe tubes that is mentioned in TEMA standard.

2.3. Thermo-compressor and ejectors

Thermal desalination systems operate at pres-sures lower than the atmospheric pressure. There-fore, using vacuum devices in these systems isunavoidable. Ejectors and thermo-compressors arecommon thermal devices which can providevacuum required for these systems.

It should be noted that the thermo-compressoris one kind of ejectors. The ejector is a pumpingdevice which uses jet action of a high pressureand temperature primary motive steam to entrainand accelerate a slower secondary steam (load).Due to the simplicity of design and the absenceof motive parts, ejectors are very reliable, requirepractically no maintenance and have a relativelylow installation cost. The ejectors are powered byheat, which is low-grade energy and it is obvi-ously less expensive to run than electrical or me-chanical-related power. The steam required for thejet ejector is commonly drawn from boilers. Thesedevices are used in vapor compression desalina-tion systems as a heat pump. A thermo vapor com-pression desalination unit mainly comprises asteam jet ejector, a single or multi effect evapora-tor, and a condenser. The thermo compressor isused to compress the vapor from pressure Ps(which is the vapor pressure leaving the last ef-fect or condenser depending on the system de-sign) to P1 (which is the vapor pressure enteringthe first effect) by using an external source ofsteam at a pressure Pe greater than the vapor pres-sure.

Two types of ejectors are usually used in thesystems. They are hogging ejector and NCG ejec-tor. The first one provides the initial vacuum ofthe system and the second one discharges non-condensable gases (NCG) from the system.

The ejector design can be classified into twocategories which are known as constant-area mix-ing ejector and constant-pressure mixing ejector.In this case, the ejectors and thermo-compressorwere designed based on the constant-pressuremixing ejector [11,12]. The thrmo-compressor andejectors design flowchart is shown in Fig. 7.

R.K. Kamali et al. / Desalination 235 (2009) 340351 345

3. Simulation algorithm

Finally, the algorithm of the thermo-hydraulicdesign of the MEDTVC system can be consid-ered as shown in Fig. 8.

The program is modular in structure and includes

Fig. 6. Shell and tube condenser design algorithm.

a number of modules for evaporators, condensers,thermo-compressor, steam jet ejectors, etc.

Each module has its own mathematical model.The program also includes a comprehensive da-tabase for the physical properties of seawater.There is a library containing correlations for heat

346 R.K. Kamali et al. / Desalination 235 (2009) 340351

Fig. 7. Thermo-compressor and ejectors design algorithm.

Inputs: Pm , Pp, PK , Tp, Ps, Ts, sm

Calculate nozzle throat and nozzle outlet diameters. Determine number of required nozzles.

Set Mach number of the nozzle outlet equal to 3.5 and calculate the nozzle outlet diameters, then determine the number of

required nozzle (usually 3)

Set the Mach number of secondary flow at the inlet of mixing section equal to 1 and calculate the pressure and temperature of

both of streams, and determine the diameter of constant area section.

By assuming "constant pressure mixing ", calculate the flow condition before shock (Pressure, Mach number...)

Calculate flow condition after Mach number, and its condition at the outlet of ejector,

By means of diameter of constant area, determine the lengths of mixing section, constant area section and diffuser section

transfer coefficients of different heat transfer sur-faces and flow regime [1315].

4. Results and discussion

4.1. Simulation results

A sample system with a capacity of 1200 m3/dfresh water is designed by the computer simula-

tion code. Table 1 shows the result of thermo hy-draulic design of MED-TVC system with shelland tube heat exchangers, also a verification be-tween design data and actual data is done andshown in this table, and Table 2 shows these re-sults for MED-TVC system with plate type evapo-rators, unfortunately there isnt any available sys-tem in order to verify these data with them, but it

R.K. Kamali et al. / Desalination 235 (2009) 340351 347

Table 1Design and actual data for MED system with shell and tube evaporator

Parameter Design Actual Deviation (%) Total distilled product, ton/d 1200 1200 Seawater temperature, C 35 35 Motive steam, ton/d 178.2 192 7 Motive steam temperature, C 62.5 62.3 Feed water temperature, C 44 43.8 Number of effects 4 4 Motive steam per entrained steam 1.12 1.15 4 Performance ratio 6.73 6.25 7 Number of tubes in each effect 2059 2228 7 Effects tube length, m 4.1 4.1 Effects tube diameter, cm 28.57 28.57 Number of condenser tubes 1481 1463 1 Condenser tube length, m 3 3 Condenser tube outer diameter, cm 19.05 19.05 Boiler pressure, barg 12 12 Boiler temperature, C 188 188 Throat nozzle diameter, cm 2.3 2.3 Nozzle outlet diameter, cm 13 13 Constant area diameter, cm 43 47 8 Shell diameter, m 3.2 3.2 Shell length, m 16 16

Fig. 8. Thermohydraulic design algorithm.

348 R.K. Kamali et al. / Desalination 235 (2009) 340351

Table 2Design data for MED system with plate type evaporator

Parameter Design Total distillated product, ton/d 1200 Seawater temperature, C 35 Motive steam, ton/d 177.7 Motive steam temperature, C 188 Feed temperature, C 44 Number of effects 4 Motive steam per entrained steam 1.12 Performance ratio 6.75 Number of plates in each effect 168 Plate length, m 1.822 Plate width, m 1.242 Gap between plates, cm 1 Number of condenser tube 1481 Condenser tube length, m 3 Condenser tube outer diameter, mm 19.05 Throat nozzle diameter, cm 2.3 Nozzle outlet diameter, cm 13 Constant area diameter, cm 43 Shell diameter, m 3.2 Shell length, m 8

is nice to note here that a new system is going tobe built based on plate type MED-TVC designdata which are shown in Table 2. It should be notedthat condenser is considered as a shell and tubeheat exchanger. According to these two systemsdimensions in these tables, it is obviously clearthat plate type evaporator is more compact thanshell and tube type, it is concluded that to achieve

Fig. 9. Variation of performance ratio with the number of effects.

2

3

4

5

6

7

8

9

3 4 5 6 7 8 9 10 11 12

Number of Effects

PR

a compact and portable system, it is convenientto use plate type evaporators.

4.2. Parametric study

It is appropriate to consider the effect of pa-rameters involved in the system. Performance ra-tio (PR) is the ratio of the amount of distillatedproduct to steam flow rate. It is the objective ofthis program to predict these effects. The effec-tiveness of the number of effects and variation inthe inlet steam temperature in the first effect onthe performance ratio is shown in Figs. 9 and 10.

It is clear that by changing the number of ef-fects the heat transfer area will change as well.

The effect of concentration factor (Xb/XF) onthe performance ratio at different number of ef-fects is observed in Figs. 11 and 12.

It is clear that by changing the number of ef-fects the heat transfer area will change as well.

The effect of the number of effects on the ra-tio of the motive steam per entrained steam isshown in Fig. 13.

Finally, by the parametric study it was con-cluded that: The performance ratio will increase with in-

creasing the number of effects and the heattransfer area.

The performance ratio will increase with de-creasing the steam temperature through the firsteffect.

R.K. Kamali et al. / Desalination 235 (2009) 340351 349

Fig. 10. Variation of the performance ratio with the first effect steam temperature at different numbers of effects.

2

3

4

5

6

7

8

9

40 45 50 55 60 65 70 75

Ts [oC]

PR

No=4No=5No=6No=7No=8No=9

Fig. 11. Variation of the performance ratio with the concentration factor at different numbers of effects.

The performance ratio will increase with in-creasing the concentration factor.

The ratio of the motive steam per entrainedsteam will increase with increasing the num-ber of effects and the heat transfer area.

5. Conclusions

The simulation model provides an effectivetool for engineers to design a MEDTVC sys-tem with any desired capacity.

The simulation model provides an effective

tool for engineers to evaluate the performanceratio of the system.

The method of parametric study is an appro-priate tool to estimate the optimum parameters.

Plate type evaporators make the system com-pact and portable rather than shell and tubetype.

Symbols

B Mass flow rate of brine blow down, m3/hCP Capacity coefficient, kJ/kg C

33.54

4.55

5.56

6.57

7.58

0 0.5 1 1.5 2 2.5 3 3.5

Xb/XF [%]

PR

No=4No=5No=6No=7No=8No=9

350 R.K. Kamali et al. / Desalination 235 (2009) 340351

Fig. 12. Variation of the performance ratio with the num-ber of effects at different concentration factors.

33.54

4.55

5.56

6.57

7.58

8.59

3 4 5 6 7 8 9 10 11

Number of Effects

PR

Xb/XF=1.3

Xb/XF=1.4

Xb/XF=1.5

Xb/XF=1.7

Xb/XF=2

Xb/XF=3

Fig. 13. Motive steam per entrained steam vs. the num-ber of effects.

D Mass flow rate of the distillate product,m3/h

D0 Mass flow rate of steam entering the firsteffect, m3/h

Dr Mass flow rate of the entrained steam,m3/h

F Mass flow rate of seawater feed to theeffects, m3/h

L Latent heat, kJ/kg

M Mass flow rate of sea water, m3/hm Mass flow rateP Steam pressurePR Performance ratioR Mass flow rate of rejected waterS Mass flow rate of the motive steam, m3/hT Temperature of the effect, CX Salt concentration, ppmP Nozzle efficiency

Subscripts and superscripts

B Brinef Condenser productF Feedi Indexn Number of effectsp Primary steams Motive steamsw SeawaterT Total

Acknowledgment

This paper belongs to the Fan-Niro Company(Tehran, Iran), so we appreciate their cooperation,and, additionally, we thank Bonian-Danesh-pajouhan Institute (Tehran, Iran) and our co-work-ers particularly Mrs. Mahmoudi, Mrs. Saadatmandand Mrs. Alishiri.

References

[1] H.M. Ettouney and H. El-Dessouky, A simulator forthermal desalination process, Desalination, 125(1999) 277291.

[2] H.T. El-Dessouky and H.M. Ettouney, Multiple ef-fect evaporation desalination systems: thermalanalysis, Desalination, 125 (1999) 259276.

[3] A. Jernqvist, M. Jernqvist and G. Aly, Simulation ofthermal desalination process, Desalination, 134(2001) 187193.

[4] H. Ettouney, Visual basic computer package for ther-mal and membrane desalination processes, Desali-nation, 165 (2004) 393408.

0

5

10

15

20

25

30

35

40

3 4 5 6 7 8 9 10 11 12

Number of Effects

S/D

r

R.K. Kamali et al. / Desalination 235 (2009) 340351 351

[5] M.B. Ould Didi, N. Kattan and J.R. Thome, Predic-tion of two-phase pressure gradients of refrigerantsin horizontal tubes, Intern. J. Refrigeration, 25(2002) 935947.

[6] W. Yu, D.M. France, M.W. Wambsganss and J.R.Hull, Two-phase pressure drop, boiling heat trans-fer, and critical heat flux to water in a small-diam-eter horizontal tube, Intern. J. Multiphase Flow, 28(2002) 927941.

[7] E.A.D. Saunders, Heat Exchangers, John Wiley &Sons, 1988.

[8] P. Fiorini, E. Sciubba and C. Sommariva, A newformulation for the non-equilibrium allowance inMSF processes, Desalination, 136 (2001) 177188.

[9] M.A. Darwish and A.A. El-Hadik, The multi-effectboiling desalting system and its comparison with themulti-stage flash system, Desalination, 60 (1986)251265.

[10] N.M. Al-Najem, M.A.Darwish and F.A. Youssef,Thermo vapor compression desalters: energy and

availability analysis of single and multi-effect sys-tems, Desalination, 110 (1997) 223238.

[11] B.J. Huang, J.M. Chang, C.P. Wang and V.A.Petrenko, A 1-D analysis of ejector performance,Intern. J. Refrigeration, 22 (1999) 354364.

[12] H. El-Dessouky and H. Ettouney, Evaluation ofsteam jet ejectors, Chem. Eng. Processing, 41 (2002)551561.

[13] W.H. Parken, L.S. Fletcher, V. Sernas and J.C. Han,Heat transfer through falling film evaporation andboiling on horizontal tubes, Heat Transfer J., Trans.ASME, 112 (1990) 744750.

[14] A. Cavallini, G. Censi, D. Del Col, L. Doretti, G.A.Longo, L. Rossetto and C. Zilio, Condensation in-side and outside smooth and enhanced tubes areview of recent research, Intern. J. Refrigeration,26 (2003) 373392.

[15] S. Kakac, A.E. Bergles and F. Mayinger, Heat Ex-changer, Thermal-Hydraulic Fundamentals and De-sign, Hemisphere, Mc Graw-Hill, 1981.