You are on page 1of 5

I.A.

Karimi and Rajagopalan Srinivasan (Editors), Proceedings of the 11th International


Symposium on Process Systems Engineering, 15-19 July 2012, Singapore.
© 2012 Elsevier B.V. All rights reserved.

Seawater Desalination Processes: Optimal Design


of Multi Effect Evaporation Systems.
Paula Druettaa , Sergio Mussatia, Pio Aguirrea
a
lNGAR - Instituto de Desarrollo y Diseño, Avellaneda 3657, Santa Fe
(3000),Argentina

Abstract
This is the first paper of a series of articles that deals with the modeling and
optimization of dual-purpose desalination plants which combine thermal desalination
processes and combined heat and power systems, specifically solid oxide fuel cell
SOFC electricity generators. This paper presents preliminary results obtained for the
multi effect evaporation (MEE) process (stand alone process). The steady state
performance of the MEE system is described by a simplified and no linear programming
(NLP) model. Optimal operating conditions including profiles of temperature, flow-rate
and heat transfer area along the evaporator are analyzed. In addition, the influence of the
effect number on the evaporation efficiency is also investigated.

Keywords: Multi-effect desalination; Optimal design; Non-Linear Programming (NLP)

1. Introduction
In many countries the most pressing issue for water security is in meeting basic
provision of fresh water supply and sanitation. Sea-water desalination plays and will
play in the future an important role in contributing to the provision of fresh-water since
the global requirements for fresh water are increasing rapidly as the global population.
The following are the main desalination processes: a) Thermal: multi stage flash (MSF)
and multi effect (MEE) distillation, vapor compression (VC) and b) reverse osmosis
(RO), electrodialysis (ED), membrane distillation (MD). The thermal desalination
processes are energy intensive (which global requirement is also increasing) and the
efficiency is greatly improved when they are integrated with heat-power plants (co-
generation power desalting plants, CPDP). Thus, the waste heat of the power plant is
used as thermal source to produce fresh-water. The main drawback of such plants,
however, are the high consumption of fossil fuel and high CO2 emissions which lead to
global warming and inevitable climate changes. It is essential therefore to look for cost-
effective and sustainable processes that combine alternative energy sources (wind,
geothermal and solar energy, fuel cells and others) with desalination technologies.
The use of high temperature fuel cells [molten carbonate fuel cells (MCFC) and solid
oxides fuel cells (SOFC)] has emerged as a suitable technology for cogeneration.
Certainly, electricity can be generated in a cleaner and more efficient way than with
conventional technologies. In addition, the temperature of the flue gases is high enough
to produce high-temperature steam to be used as heating medium in energy intensive
processes.
Fig. 1 shows a schematic and feasible flow-sheet which integrates fuel cell and
desalination process. The mathematical model to be presented in this work is the first
basic step of a more ambitious project aimed at determining the optimal synthesis and
design of the integrated process including desalination processes and SOFCs.
Seawater Desalination Processes: Optimal Design of Multi Effect Evaporation Systems. 771

Fig. 1 Integration of SOFC and Multi-Effect Evaporation (MEE) process.


The MEE desalination system illustrated in Fig 1, which is the aim of this paper, is
named “forward feed flow arrangement. As shown, each one of the effects involves a
pre-heater, an evaporator and a distillate flash chamber. Feed seawater enters to the
system and is preheated in the down condenser by the vapor formed in the last effect. A
part of the seawater is then discharged to the sea. In the first effect, part of the vapor
formed by evaporation of the brine is condensed outside the tubes of the pre-heater to
heat the feed and the remaining the vapor is used in the second effect to evaporate part
of the brine coming from the first effect. From the second to the last effect, the vapor
formed by flashing of the brine and distillate is used to pre-heat the feed while the fresh
vapor formed by evaporation is used to evaporate part of the brine in the next effect.
This cycle of evaporation, condensing and heating is repeated at each effect until the
last effect. The first and last stages need external heating and cooling respectively. The
amount of heat removed from the last stage must nearly equal the amount of heat
supplied to the first stage (Q). For sea water desalination purpose, the first stage is
typically operated at a temperature below 70 °C in order to avoid scale formation.

2. Optimization Problem
The proposed optimization problem can be stated as follows. Given the sea-water
conditions (composition, temperature and flow-rate), the goal is to determine the
optimal operating conditions and the heat exchange values for the different heat
exchangers in order to minimize the total heat transfer area in satisfying the fresh-water
demand. The problem is solved for the following two cases: a) equal heat transfer areas
(HTA) in all pre-heaters and evaporation effects and b) variation of HTA along the pre-
heaters and evaporation effects. The output results are compared in detail. Finally, the
influence of the number of effects and the steam temperature supplied in the first effect
(heating utility) on the total heat transfer area is also analyzed. For this analysis, the
optimization problem is systematically solved by varying both parameters.

3. Assumptions and mathematical model


The following are the main assumptions used to develop the mathematical model.
772 P. Druetta et al.

x Heat losses, non-equilibrium allowance (NEA) and Non-condensable gases


effects are neglected.
x A mean value is adopted for the heat capacity coefficients (Cp), the latent heat
of vaporization (O), the overall heat transfer coefficients (U) and for the boiling
point elevation of the brine (BPE), neglecting the effect of the chamber
geometry, temperature, pressure and the fluid properties.
x The vapor form by flashing in the effects and the flashing boxes is used to heat
the feed.
x The heat transfer equations model the HTA in each evaporator as the sum of
the area for brine heating (Ab) and an area for evaporation (Ae).
3.1. Equations and computational aspects
The MEE system is represented as a No Linear Programming (NLP) model. This model
consists in a set of mass and energy balances, and design equations used to compute
heat transfer areas and physical-chemical properties of each one of the streams. There is
mass and energy balances for each one of the pre-heaters, effects and flashing
chambers. The logarithmic mean temperature difference (LMTD) is used to compute the
HTA in pre-heaters. The model also includes non equality constraints which are used to
avoid temperature crossover. In addition, lower and upper bounds have been imposed
for each one of the variables in order to facilitate the model convergence.
The model involves non-linear constraints and bilinear terms which lead to local
optimal solutions and therefore, global optimal solutions can not be guaranteed. In order
to solve the optimizations, a feasible solution taken from Darwish’s (2006) has been
used to solve the first optimization problem. Then in order to solve the next
optimization problem where some model parameters are varied, the previous solution is
used as initialization. Thus, a sequence of several optimization problems is efficiently
and systematically solved.

4. Results
The optimization problem described in section 3 is solved using the parameter values
listed in Table 1. As mentioned earlier, the fresh-water demand (D), steam temperature
(Ts), steam flow-rate (S) and seawater conditions [(Xfeed), (Tfeed)] are given. In addition,
an upper bound for the rejected brine salinity is imposed for environmental restrictions
(XUp). The down condenser eject temperature (Tc) is also assumed as given.
Table 2 and Fig. 2, 3 and 4 compare the output results obtained by considering either
uniform or non uniform distributions of heat transfer areas in pre-heaters and
evaporation effects along the MEE system. In Table 2 are also compare same widely
used variables to measure the system efficiency: the performance ratio (PR), the
conversion ratio (CR), the specific total HTA (sHTA) and the specific cooling water
ratio (sCWR).
Table 2 clearly shows that despite the cooling water flow-rate and the HTA in down
condenser for non uniform area are greater than those required for uniform area, the
total HTA required in pre-heaters and effects is considerably smaller than that required
for uniform area. Fig. 2 compares the distribution of the HTA in pre-heater and
evaporation effect along the MEE system (8 stages). It is possible to observe that the
non uniform distribution of the HTA leads to the minimum total HTA of the process.
Fig. 3 compares the corresponding distribution of the driving force and the fresh water
production in each one of the effects for both cases. It clearly shows that from the first
to fifth effect the driving forces are similar in both cases and then in the last three
effects the driving force for non uniform area is greater than the corresponding to
Seawater Desalination Processes: Optimal Design of Multi Effect Evaporation Systems. 773

uniform distribution. In both cases, the total fresh-water production which is fixed is
similarly distributed along the desaltor.
Fig. 4 compares the distribution of the temperatures in pre-heater and evaporation effect
along the MEE system (8 stages) for both situations.
Table 1. Parameter values Table 2. Optimal values
Parameters Non uniform Uniform
Cp [KJ/Kg ºC] 4 HTA HTA
D [Kg/s] 2500 sHTA [m2/Kg/s] 252.76 285.36
BPE [°C] 1 PR (D/S) 4.16 4.16
O [KJ/Kg] 2333 sCWR (Wc/D) 5.91 4.93
Tfeed [°C] 26 CR (F/D) 2.77 2.77
Tc [°C] 40 Cooling Water 14785.22 12325.65
Ts [°C] 70 Flow [Kg/s]
Xfeed [ppm] 45978.9 Down Condenser 36009.02 21540.53
XUp [ppm] 72000.0 HTA [m2]
U [kW\ m2 ºC] 3 Total HTA [m2] 631901.21 713408.13
N 8
S [Kg/s] 600

Fig. 2. Heat transfer areas vs. effect Fig. 3. Fresh water production and driving
force vs. effect

Fig. 4. Brine and preheaters temperature vs. Fig. 5. Total HTA and heating utility vs.
effect effect

The impact of the number of effects and the steam temperature used for heating utility
on the total HTA and consumption of the heating utility is shown in Fig. 5. The
illustrated results were obtained by minimizing the total HTA considering uniforms
HTA in pre-heaters and effects and a given fresh water demand. In addition, in this case,
the heating utility consumption was considered as a variable. Fig. 5 exhibits a similar
behavior to those obtained via simulation by El-Dessouky (2002), instead of
simultaneous optimization as is here used. As is shown, both the total HTA and heating
774 P. Druetta et al.

utility consumption are strongly influenced by the number of effects and steam
temperature. The total HTA for N = 4 is not strongly influenced by the steam
temperature. However, the total area is significantly increased as the number of effects
increases and the steam temperature decreases. Certainly, for Ts = 65 ºC and N = 4, the
total specific HTA is 192.17 m2/Kg/s, while for same steam temperature but N=12 the
total specific HTA increases to 717.72 m2/Kg/s, a 525 % higher.
By the other side, the heating utility consumption decreases as the number of effects
increases and the steam temperature decreases. Certainly, for Ts = 65 ºC and N = 4, the
total specific heating utility is 682.78kJ/s/Kg/s, while for same steam temperature but
N=12 the total heating utility decreases to 273.57 kJ/s/Kg/s, a 59.93 % lower.

5. Conclusions and future work


A deterministic NLP mathematical model by effects for the MEE system was presented.
Optimal temperature, salinity, areas dimensions and flow-rate evolutions through the
equipment are obtained in order to satisfy a given fresh water demand. The proposed
resulting model is a valuable tool not only to optimize the process but also to simulate
the desalting process if the degree of freedom of the equation system is zero. After
validation, several optimization problems using different criteria have been examined
concluding that the total HTA require can be decreased a 11.4% by allowing HTA
taking a non uniform distribution. It was verified that the minimization of the total HTA
by fixing the fresh water demand and external heat, is equivalent to the minimization of
the external heat by fixing the total HTA and fresh water demand. The effect of the
number of effects and team temperature on the optimal solutions has been also
investigated, obtaining similar results to those obtained by other authors. HTA and
heating utility consumption are strongly influenced by the number of effects and steam
temperature: the total HTA increased as the number of effects increases and the steam
temperature decreases, while the heating utility consumption decreases as the number of
effects increases and the steam temperature decreases.
At present, the proposed model is being expanded in order to consider the number of
effects as an additional model variable (discrete decision), as well as the inclusion of the
geometry, pressure drop, BPE, NEA and non-condensable gases. The coupling between
the MEE system with SOFT cells and the minimization of the investment and operating
costs of the entire process will be further investigated in detail. Finally, the application
of a global optimum algorithm to solve the resulting model will be also addressed.

References
S. A1-Hallaj, Fuad Alasfour, Sandeep Parekh, Shabab Amiruddin, J. Robert Selman, Hossein
Ghezel-Ayagh, 2004, Conceptual design of a novel hybrid fuel cell/desalination system,
Desalination, 164, 19-31
M.A. Darwish, F. Al-Juwayhel, H. K. Abdulraheim, 2006, Multi-effect boiling systems from an
energy viewpoint, Desalination, 194, 22–39.
H.T. El-Dessouky, I. Alatiqi, S. Bingulac, H.M. Ettouney, 1998, Steady-state analysis of the
multiple effect evaporation desalination process, Chem. Eng. Technol, 21, 15±29.
H.T. El-Dessouky, H.M. Ettouney, 2002, Multiple Effect Evaporation, Fundamentals of Salt
Water Desalination, Capitulo 4.2.3, ELSEVIER SCIENCE B.V., 202-205.
H.T. El-Dessouky, H.M. Ettouney, 1999, Multiple-effect evaporation desalination systems:
thermal analysis, Desalination, 125, 259-276.
H.M. Ettouney, 2006, Design of single-effect mechanical vapor compression, Desalination, 190,
1–15.
P. Lisbona, J. Uche, L. Serra, 2005, High-temperature fuel cells for fresh water production,
Desalination, 182, 471–482.

You might also like