- Email: [email protected]
LiBrH2O absorption heat pump for single-effect evaporation desalination process
DESALINATION ELSEVIER
Desalination 128 (2000) 161-176 www.elsevier.com/locate/desal
LiBr-H20 absorption heat pump for single-effect evaporation desalination process •
•a*
Falsal Mandam , Hisham Ettouney b, Hisham E1-Dessouky b aCollege of Technological Studies, PO Box 42325, Shuwaikh 70654, Kuwait Fax +965 481-3691 bChemical Engineering Department, College of Engineering and Petroleum, Kuwait University, PO Box 5969, Safat 13060, Kuwait Received 28 October 1999; accepted 1 December 1999
Abstract
A new configuration is presented for the single-effect evaporation process combined with lithium bromide water (LiBr-H20) solution absorption heat pump. Performance evaluation of the process is presented in terms of variations in the thermal performance ratio, the specific heat transfer area, the specific flow rate of cooling water, and the conversion ratio. These variations are presented as a function of the heating steam temperature, the lithium bromide mass fraction in the concentrated solution, and the temperature difference of the heating steam and the boiling brine. Results show thermal performance ratios of 2.4-2.8, which are higher by 50-70% than those for the single-effect thermal vapor compression system. Operation at higiier heating steam temperatures and large temperature difference between the heating steam and the boiling brine is highly attractive due to drastic reduction in the specific heat transfer area and the specific flow rate of cooling water. Also, at these conditions slight reduction occurs in the system thermal performance ratio. In search for low-energy and low-cost production units, the superior performance of high temperature operation for the absorption heat pump single-effect system makes it extremely suitable for adoption by remote and small communities. Keywords." Water desalination; Single-effect evaporation; Vapor compression; Absorption heat pump; Modeling; Simulation
1. I n t r o d u c t i o n At the turn o f the millennium, securing sustainable resources for fresh water and energy *Corresponding author,
is a major challenge for mankind. Desalination of sea and brackish water has developed considerably during the 20th century to provide and support humane and industrial activities in several zones around the world, especially in the
0011-9164/00/$- See front matter © 2000 Elsevier Science B.V. All rights reserved PII: S0011 - 9 1 6 4 ( 0 0 ) 0 0 0 3 I - X
162
F. Mandani et al. / Desalination 128 (2000) 161 176
Gulf States, other middle eastern countries, southern California and Florida, Italy, and Spain. Developments have resulted in configuration of several forms for thermal desalination, which could be based on evaporation or flashing, operation in single or multiple effect, or designed as a stand-alone system or combined with vapor compression heat pumps [1]. Regardless, the desalination industry remains to be expensive for the majority of the world population because of its high capital and production cost. Recent quotes for funding requirements of a 20migd desalination plant based on reverse osmosis (RO), multi-stage flash desalination (MSF), single-effect mechanical vapor compression (MVC), or multiple-effect evaporation (MEE) vary between $69 to $95 million [2]. In addition, Bednarski and Minamide [3] reported a unit production cost of $0.8/m 3 for 6migd MSF, $0.72-0.93/m 3for RO (the cost range depends on the pretreatment cost), and $0.45/m 3 for lowtemperature MEE. This relatively high unit product cost is a result of the high energy cost, which may vary between 40-60% of the unit product cost [4]. In an attempt to reduce the energy cost in desalination and other industrial process, use of heat pumps has recently been considered. Heat pumps receive energy from a low-temperature source and upgrade to a higher temperature before performing the desired function. Several forms of heat pumps exist in the literature and are suitable for operation withthermal-based desalination processes. Examples include thermal vapor compression (TVC), MVC, adsorption vapor compression (ADVC), and absorption vapor compression (ABVC). Examples for performance of these heat pumps in thermal desalination processes can be found in the studies by AI-Juwayhel et al. [5] on single-effect evaporation, EI-Dessouky and Ettouney [6] on MEE systems, and E1-Dessouky et ai. [7] on MSF combined with thermal vapor compression.
Single-effect evaporation desalination units combined with vapor compression have several attractive features that make them suitable for providing sustainable water supplies for industrial sites, remote small communities, and scattered remote populations. Currently, the best available technology in this category is the single-effect MVC process with unit capacities up to 5000m3/d [8]. The process has highly competitive specific power consumption against the RO process with values ranging between 48kWh/m 3. Similar to other thermal desalination processes, the MVC process has reliable operation, the ability to withstand harsh operating conditions of high feed temperature and high salinity, and has a high plant factor [9]. The major drawback of the MVC process is that its main energy source is electric current, which is very expensive in comparison with saturated steam used in other thermal desalination processes. In addition, a major component in the MVC process is the mechanical compressor, which has moving parts and therefore requires specialized and more frequent maintenance frequency. Other single-effect evaporation desalination processes include the TVC system, the ABVC, and the ADVC. The single-effect TVC process is not found on industrial scale because of its relatively lowthermalperformanceratio, defined as the ratio of product water per unit mass heating steam, with values averaging 1.5 [10]. However, the multiple-effect thermal vapor compression process is the standard of the MEE industry. The multiple-effect configurations have flexible design features that allow operation in a stand-alone mode or combined with the thermal vapor compression heat pump. The thermal performance ratio of the process is high; for example, a 12-effect system has a performance ratio of 8 in a stand-alone mode and of 16.7 in the vapor compression mode [ 11]. Weinberg and Ophir [12] reported similar trend for a stand-
F. Mandani et al. / Desalination 128 (2000) 161-176
alone six-effect system with a performance ratio of 5.7. Also, Michles [13] reported a thermal performance ratio of 8 for a four-effect system combined with TVC. Kronenberg [ 14] showed that unit product cost for various combinations of the multiple-effect system with power generation units, i.e., diesel, steam turbines, or gas turbine cycles, varies between $0.27/m 3 to $0.56/m 3. On the other hand, the unit product costs for the RO and MSF system are higher with values of $0.68/m 3 and $0.89/m 3. It should be stressed that multiple-effect TVC is the industrial standard because of its higher thermal performance ratio in comparison with single-effect TVC. On the other hand, the singleeffect MVC is the industrial standard. This is because the specific power consumption has the same value for the single- and multiple-effect MVC configurations [5-15]. The main difference between the single- and multiple-effect MVC systems is the increase in the system capacity for the multiple-effect systems due to the increase in the compression range. Although, the combination of the ABVC and ADVC systems with thermal desalination provide high performance in comparison with conventional configurations, their use is not found on full commercial or industrial scale. The literature includes a large number of studies on development, innovation, and performance evaluation of ABVC and ADVC for refrigeration and air conditioning processes [16]. On the other hand, evaluation of the combined systems of these heat pumps and various thermal desalination processes is limited to a small number of studies in the literature. Weinberg et al. [ 17] evaluated the performance of a coupled system of MEE vacuum freezing with a lithium bromide absorption heat pump. The system is thought to enhance the performance ratio of the MEE to high values of 18-20 and operating temperatures between 0-60°C. The low-temperature operation minimizes corrosion and scaling problems. Alefeld
163
and Ziegler [18] proposed a fully integrated desalination system combined with LiBr-H20 absorption heat pump. The system includes three stages which process seawater and generate fresh water. Aly [19] and Fathalah and Aly [20] analyzed a solar-powered LiBr-H20 heat pump, which generates high-grade steam to operate a MEE desalination system. In their analysis, more emphasis was given to the performance of the solar power unit and air conditioning in the evaporator unit. Yanniotis and Pilavachi [21] modeled the performance of sodium hydroxide heat pumps in the MEE systems. Model results are validatedagainstexperimentalmeasurements and were found to have reasonable agreement. A1-Juwayhel et al. [5] studied the performance of single-effect evaporation desalination systems combined with various types of heat pumps. As discussed before, results for the ABVC and ADVC gave thermal performance ratios close to three times higher than the single-effect TVC system. On the other hand, E1-Dessouky and Ettouney [6] showed a 50% increase in the thermal performance ratio for the MEE-ABVC and MEE-ADVC systems over the MEE-TVC with values close to 20. This paper evaluates the performance of a novel process constituting a single-effect evaporation desalination combined with absorption heat pumps. This process integrates the exothermic heat of absorption within the desalination system, instead of providing hot utility water. The study includes development of the proposed process and mathematical model. Results and system performance are presented as a function of the design and operating parameters that have a strong effect on the product water cost.
2. Description of the ABVC process Elements of the ABVC system are shown in Fig. 1. In this system, the evaporator constitutes
164
F. Mandani et al. / Desalination 128 (2000) 161-176
~l Cooling ~---~ Il EntrainedI- -gap°r' T [~l"Mev' ' Tv I
~
Mcw' Seawater Xcw' Tf
I FeedSoawator M, X.w ", I Absorber
Heat
Intake Seawater Mf + Mcw, Xcw, Tcw
~
I Dilute LiBr-H20 Solution Mo, Xo
Feed brine, | [ Ma, Xa, Ta [---]
! ~ ~iiWl --
<
SolutionC°nc" LiBr-H20 Mg, Xg
..~ ~
Vapor, D
~
~ 7 ......
Down Condenser
.... ~
~ 7
[ Heating
,,i ~: : ~ ,imi .............. '
~ Motive Steam, Tin, Pm
Fa Film Evaporator
I ~1 ~ Blowdown Brine Mb' Xb' Tb
I Generator ~'~ [Cond . . . . te l
Fig. 1. Schematic of single-effect absorption vapor compression desalination process (ABVC).
horizontal falling film tubes, brine spray nozzles, demister, and the brine pool. The down condenser has a similar shell and tube configuration where condensation takes place on the shell side. The absorber is also a shell and tube falling film configuration. Absorption of water vapor by the LiBr-H20 solution occurs on the shell side of the absorber, while heating of the feed seawater and vapor formation take place on the tube side of the absorber. The generator has a similar layout to the evaporator where dilute LiBr-H20 solution forms a falling film on the outside surface of the tubes and the motive steam is condensed inside the tubes. The heat exchange unit between the concentrated and diluted LiBr-H20 solution is used to recover part of the energy from the higher temperature andconcentratedLiBr-HzOsolution, This improves the overall process efficiency, The process is described in the following points: • The intake seawater stream flows through the down condenser, where it condenses part of the vapor formed in the evaporator. The
•
•
•
remaining part of the vapor is fed to the shell side of the absorber in the heat pump. The intake seawater temperature increases from (Tcw) to (Ts) as it absorbs the latent heat of condensation of the condensing vapor. Part of the feed seawater is rejected back to the sea (Mc,~) which is known as the cooling seawater. The remaining portion of the intake seawater is the feed seawater stream (Ms), which is chemically treated and deaerated before being fed to the tube side of the absorber. The concentrated LiBr-H20 solution absorbs the vapor stream entering the absorber. The absorption process is exothermic and releases a sufficient amount of heat that sustains an increase of the feed seawater temperature to the saturation temperature. Also, vapor is formed from the feed seawater within the absorber. This vaporformspartoftheheating steam in the evaporator. The temperature of the absorbed vapor and the concentration of the outlet dilute LiBr-H20 solution define the equilibrium
F. Mandani et al. / Desalination 128 (2000) 161-176
•
•
•
•
conditions in the absorber. It should be noted that the boiling point elevation for the LiBr-H20, or the temperature difference between the dilute LiBr-H20 solution and the absorbed vapor varies over a range of 1050°C as the mass fraction of LiBr-H20 in the dilute solution is increased from 0.25 to 0.45. The dilute LiBr-H20 solution enters the generator where it is sprayed on the outside surface of the tubes. The solution absorbs the latent heat of motive steam that condenses on the tube side of the generator. The heating process increases the temperature of the LiBr-H20 solution to saturation and results in evaporating the same amount of water absorbed by the solution in the absorber. The concentration of the concentrated LiBr-H20 solution and the temperature of the formed vapor define the equilibrium conditions in the generator. The combined vapor formed in the generator and absorber derives the evaporation process in the evaporator. The brine stream leaving the absorber is sprayed on the outside surface of the evaporator tubes where it absorbs the latent heat of condensation from the condensing steam on the tube side of the evaporator. The concentrated brine leavingthe evaporator is rejected back to the sea and the formed vapor is routed to the down condenser. The sum of the condensate of the heating steam and the condensate in the down condenser forms the distillate product stream. Demisters in the generator, absorber, and evaporator prevent droplet entrainment of the LiBr-H20 solution or brine in the vapor stream.
3. Mathematical model of ABVC The steady-state model includes a set of material and energy balances, heat transfer equations, and thermodynamic relations.
165
Assumptions used in the model include: • The vaporformedintheevaporator, absorber and generator is salt free; this assumes that the entrainment of brine droplets by the vapor stream is negligible and has no effect on the salinity of the distillate product. • Energy losses from the evaporator to the surroundings are negligible; this is because of operation at relatively low temperatures, between 100-40 ° C, and the evaporator is well insulated. • The physicalproperties of various streams are calculated at the average temperature and concentration of influent and effluent streams. The overall material and salt balances are given by Mf = M d + M b Mb --
Mf~(f/Xb)
(1) (2)
where M is the mass flow rate, X is the salt concentration, and the subscripts b, d, and f denote the brine, distillate, and feed seawater. In Eq. (2) the brine blowdown salinity (Xb) is set at 90% of the saturation salinity of the CaSO4 solution X b = 0.9 (457,628.5 - 11,304.11 Tb + 107.5781 Tb2 -0.360747 Tb3)
(3)
This equation is obtained by curve fitting of the salinity/temperaturerelation for the solubility of CaSO4 [ 15]. In the evaporator the saturated falling brine film absorbs the latent heat of the condensing steam. This evaporates a controlled mass of vapor, D at T~, where M Z : D~. v
(4)
F. Mandani et aI. / Desalination 128 (2000) 161-176
166
where )~ is the latent heat. The subscripts s and v denote the heating steam and the vapor formed, respectively. In the evaporator, absorber, and generator the boiling temperatures are higher than the corresponding vapor saturation temperatures by the boiling point elevation (BPE), and the temperature rise is caused by the hydrostatic pressure head, ATy. This is Tb = Tv +
BPE + A T
(5)
The term ATy is negligible in horizontal falling films because of the very small thickness of the boiling film. The condensation temperature in the down condenser, Tc, is lower than the boiling temperature, Tb, by the saturation temperature depressions associatedwithpressurelossesinthe demister (APp), transmission lines (AP 0, and vapor condensation inside the tubes (APe). The resulting condensation temperature is
This energy balance is given by
(D-Mev))~c=(Mw+Mf)Cp(Tf-Tw)
where the subscripts c, cw, and ev denote the condensing vapors, the intake seawater, and the entrained vapor in the absorber. The following relation gives the flow rate of the heating steam: M = M v +MaD
(9)
where Mev is the amount of entrained vapor in the absorber, subsequently released in the generator as a part of the heating steam. Mab is remaining part of the heating steam generated in the absorber. Inspection of Fig. 1 shows thatthetotal distillate flow rate is given by Ma = D + M b
T : T b -(BPE + A T + ATt + ATe)
(8)
(lO)
(6) The energy balance around the absorber is given by
The pressure drop during condensation, APc, is defined as the algebraic sum of the decrease caused by friction, APr, and the increase caused by gravity, Apg, and vapor deceleration AP a. This relation is given by
AP c = (AP r - AP i - APo)
(7)
Correlations for the pressure drop components App, APt, APr, Apg, and APa are given in the study by EI-Dessouky et al. [22]. In the down condenser, the temperature of the intake seawater, MCw+Mf, is increased from T~wto Tf. Condensing part of the vapors formed in the evaporator provides the required heating energy,
Mg/-/g + MevH"ev + MfHf
= (rag + m v ) H + mab H" s + M aH a
(11)
where Mg is the flow rate of the concentrated LiBr-H20 solution entering the absorber, and M. is the brine mass flow leaving the absorber. The water vapor saturation enthalpies H"ev and H" s are obtained at the condensation temperature in the down condenser (To) and the heating steam temperature in the evaporator (Ts), respectively. In Eq. (1 I) Hg and/4o are the enthalpies of the concentrated and diluted LiBr-H20 solution evaluated at (Cg, Tg) and (Co,To). It should be noted that Tg and To are obtained from the
F. Mandani et al. / Desalination 128 (2000) 161-176
equilibrium relation [Eq. (32) in the appendix] at water vapor saturation temperatures of T~ and To, respectively. The enthalpies of the feed seawater and the feed brine, He and Ha, are calculated at (Tf, Xow) and (Ta, Xa), respectively. The heating steam temperature is related to the feed brine temperature by the boiling point elevation, or
T= T~-BPE~(a,T )
(12)
condenser. For the evaporator, the heat transfer area, Ae, is M )~ A e
(18) Ue (T_Tb)
where U is the overall heat transfer coefficient, and the subscript e refers to the evaporator. The heat transfer area of the down condenser is given by
The energy equation for the generator balances the amount of input energy in the motive steam and the dilute LiBr solution against the amount of output energy in the concentrated LiBr solution and the heating steam. This relation
A -
is given by
(LMTD) c _
(J~lfg+mev)Ho + mmO"m = mgHg, + McvH" s
(13)
167
(D-Mv))~ c
c U (LMTD)c
(19)
Tf- T~w In Tc T~w
(201
re- L.
where M m and H " m are the mass flow rate and enthalpy of motive steam. The material and salt balances around the absorber for the feed seawater and the feed brine are given by
EI-Dessouky et al. [22] developed the correlations for the overall heat transfer coefficient in the evaporator and condenser: Ue = 1.9394 + 1.40562 x 10 -3 Tb -2.0752
Mf = Ma + (M~-M~v)
(14)
(21)
xlO 4(rb)2 +2.3186x 10 6(rb)3 Xf -~iff = MaX a
(15) Uc = 1.6175 +0.1537x 10 -3 T +0.1825 (22)
Similarly, the following relations give the total mass and salt balances for the concentrated and diluted LiBr-H20 solutions: M o = Mg +Me v
(16)
MoC ° =MgCg
(17)
The design equations for the heat transfer area are developed for the evaporators and the down
x 10 3(T)2-8.026 × 10-8(T~)3 where Ue and Uc are the overall heat transfer coefficient in the evaporator and down condenser in kW/m 2 °C, Tb is the brine boiling temperature, and T~ is the vapor condensation temperature in the condenser. All temperatures in the above correlations are in °C. The standard deviations for the above correlations are 2.03 % and 1.76%. These correlations are tested and proved to be
168
b~ Mandani et al. /Desalination 128 (2000) 161-176
reliable through comparison against other correlations in the literature and available experimental and design data.
4. Performance parameters The system performance is defined in terms of the following parameters: • performance ratio, which is defined as the amount of the distillate product per unit mass of the motive steam P R = Md/Mm
•
specific flow rate of cooling water, which is defined as the amount of the cooling water per unit mass of distillate product
sMcw = M~w/Md
•
(24)
specific heat transfer area, which is defined as the ratio of the total heat transfer area of the evaporator and condenser to the total flow rate of distillate product
sA = ( A e + A , ) / M d
•
(23)
(25)
conversion ratio, which is defined as the amount of distillate product per unit mass of feed seawater
CR = MJMf
(26)
5. Solution method The solution procedure is shown in Fig. 2. Solution of the model equations requires definition of the following system variables: • distillate flow rate, Md, is 1 kg/s • intake seawater temperature, Tow,is 25 °C • heating steam temperature, Ts, is higher than brine boiling temperature, Tb, by 2-10°C
•
feed seawater temperature, Tf, is lower than the vapor condensation temperature Tc by 5 °C • feed brine temperature, Ta, is lower than the temperature of the dilute LiBr-HzO solution To by 5°C • motive steam temperature, Tin, iS higher than the temperature of the concentrated LiBr-H20 solution Tg by 5 °C • range for the heating steam temperature, Ts, is 50_100°C • feed seawater salinity, Xf, is 36,000 ppm. As shown in Fig. 2, the solution sequence proceeds as follows: • The system capacity, stream temperatures, and stream salinity are defined as specified above. • Eqs. (1)-(3) are solved to determine the feed and brine flow rates and the salinity of the brine blowdown. • An initial guess is assumed for the following: (1) evaporator area; (2) flow rates of the heating steam, motive steam, entrained vapor, vapor formed in the absorber, dilute LiBr-H20, and the concentrated LiBr-H20; (3) concentrations of the dilute LiBr-H20 solution and brine leaving the absorber. • The above variables are calculated by solution of Eqs. ( 4 ) a n d (9)-(18). Solution proceeds iteratively using Newton's method. Iterations continue until the tolerance criterion is achieved with a value of 1 × 10 4 for e. • The flow rate of the cooling seawater and the heat transfer area of the condenser are calculated from Eqs. (8) and (19), respectively. • The performance parameters are calculated from Eqs. (23)-(26).
6. Results System performance is evaluated as a function of the heating steam temperature, the temperature difference of the heating steam and the boiling
169
F. Mandani et al. /Desalination 128 (2000) 16l--176
Define System Temperatures and Stream Salinity:
Md, Tb, Tcw, Tin, Tf, Ts, Cg, Xf, Xd
Calculate Brine salinity and Flow Rates of Brine and Feed Streams:
Mfand Mb from Eqs. (1 and 2) Xb from Eq. (3)
Calculate Initial Guess (X°):
Ae, Ms, Mev, Mab, Mm, Ma, Mo, Mg, Xa, Co, D
Solve the E q u a t i o n s a n d O b t a i n N e w P r o f i l e s (X1): 1
Ae, Ms, Mev, Mab, Mm, Ma, Mo, Mg, Xa, Co, D
C h e c k I t e r a t i o n s Error: m
( y
No
(x ° _
<
i=1 Yes~ Design the Down Condenser: Calculate Ac and Mcw from Eqns. 19 and 8
Calculate Performance Parameters:
Eqs. (23-26) PR, sMcw, sA, and CR
Fig. 2. Solutionalgorithmof the absorptionheatpumpand the singleeffectevaporationdesalinationsystem.
170
F. Mandani et al. / Desalination 128 (2000) 161-176
brine, and the mass fraction of LiBr in the concentrated solution. Effects of the heating steam temperature and the mass fraction of the LiBr in the concentrated solution are shown in Figs. 3-5 for the variations in the performance ratio, the specific heat transfer area, and the specific flow rate o f cooling water. The three figures show insensitive and independent behavior o f the system parameters on the mass fraction of LiBr in the concentrated solution. The following cause this behavior: • Various system temperatures including the heating steam, the boiling temperature, the feed seawater, and the feed brine, are independent o f the LiBr mass fraction in the concentrated solution. These temperatures affect the system variables used to calculate the system performance parameters. • Increase in the LiBr mass fraction in the concentrated solution affects only the flow rate of the concentrated solution. At higher
4 Yb =Ts-3 xf= 36000 plan
3.5 ~ 3 ~ ~ 2.5 :2
Tcw = 25 °C
__ ~ *=====~" "
2
1.5 |
o.4
i
i
0.5
0.6
0.7
~450 ~c ~ E 400
~ ~, " ~
~ 35o ~
cc • 0.45 , 0.55 •A 0.5 0.6
absorber and generator.
~' 250
Tb=Ys - 3 Xf = 36000 ppm
'~
Tcw = 25
N
~ 200 40
. 50
~
.
. 60
. 70
. . 80 90
Heating Steam Temperature,
100
110
°C
Fig. 4. Variation in the specific heat transfer as a function of the mass fraction of LiBr in concentrated solution and the heating steam temperature. .
• Selection between the above two points is dependent upon several factors including: • availability ofhighpressure steam, i.e., 17bar vs. 3 bar
0.8
Mass Fraction of L~3r h Concentrated So krtion
Fig. 3. Variation in the performance ratio as a function of the mass fraction of LiBr in concentrated solution and the heating steam temperature.
The main effect o f increasing the LiBr mass fraction in the concentrated solution is the need for higher pressure motive steam. This is illus-
•
i
~ 300
Cg = 0.75, Tg = 201.87°C, Tm = 206.87°C, Pm=1789 kPa (17.89 bar), P R = 2.4, sA = 220.9, sMcw = O, CR = 0.092. Cg=O.45, T g = 1 2 2 . 9 3 ° C , T m = 1 2 7 . 9 3 ° C , P ~ 253 kPa (2.53 bar), P R = 2.3, sA = 212.9, sMcw = O, CR = 0.092.
~
Ts
.~
•
~
• 100 -- ~- - 8 050 --,- 65
concentrations, the solution enthalpy increases and results in reduction of the concentrated solution flow rate. This is necessary to balance the system energy in the
trated in the following data obtained for a heating steam temperature of 100°C:
*
•
increase in the system second law efficiency upon operation at low steam pressures [4-23] use of smaller tube diameter for higher pressure steam [7] elimination o f the control loops required for reduction of the steam pressure to lower values of 3 bar [24].
F. Mandani et al. / Desalination 128 (2000) 161-176
1.75
16
E+05
14 ~: 12
Cc --*-- 0.45 ~ 0.55 • 0.5 -----0.6
\ \
"N~10
6 ~:
~
171
~
Tb =
4
'xL_
2
Ts - 3
xf = 36000 pwn
CaSO4.2H20
~
g
+
~
--60
T ~ , , ,T~w . = 25 oC
r~ g] 0
40
"i 50
~i 60
"i 70
i
80
~
-,,~
--
90
a
1.40E+05 HeatingSteam E Temperature ~ ---~-100c
SO4.0.SH~
70 C C
3.50E+04
-
100
110
00oz+00 0
HeatingSteam Terr~rature, °C
20
40
60
80
100
120
T e m p e r a t u r e , ,<2
Fig. 5. Variation in the specific flow rate of cooling as a function of the mass fraction of LiBr in concentrated solution and the heating steam temperature. As shown in Figs. 3-5, the effects of the heating steam temperature on the system performance are more pronounced than the mass fraction of the LiBr in the concentrated solution. This gives a dramatic reduction in the specific heat transfer area and the specific flow rate of cooling water. As shown in Fig. 4, the specific heat transfer area has values above 400 m2/(kg/s) at heating steam temperatures close to 50°C. On the other hand, the specific heat transfer area decreases to values between 200-250mZ/(kg/s), which are considered the industrial practice, as the temperature is increased to values between 80-100 ° C. This behavior is primarily caused by an increase in the overall heat transfer coefficient at higher temperatures. This enhances the heat transfer rate and results in reduction of the required heat transfer area. Another factor with a lesser effect is the reduction in the latent heat of evaporation or condensation at higher temperatures, which results in reduction in the thermal load of the evaporator and the down condenser, It should be stressed that the temperature difference or the driving force between heating steam and boiling brine has no effect here because it is kept constant in the calculations.
Fig. 6. Calcium sulfate solubility and top brine temperature for the ABVC system.
3.5 3 .~ 2.5 ~ ~o
2
AT=Ts "Tb •* 4 2
1.5
.L 6
1 4o
. 8 x 10 . . . . 50 60 70
Cc =0.55 xf = 34000 ppm Tcw=25°C . . 80 90 100 110
Heathg Steam Temperature, °C
Fig. 7. Variation in the performance ratio as a function of the temperature difference of heating steam and top brine temperature and the heating steam temperature.
At higher heating steam temperatures, the specific flow rate of cooling water is zero (Fig. 5). This is because of the limitations imposed on the salinity of the brine blowdown stream (Fig. 6 ) w h e r e at h i g h e r t e m p e r a t u r e s t h e difference in the salinity o f the feed seawater and the brine blowdown is less than 1000-2000 ppm. As a result of the constant production capacity,
1~ Mandani et al. // Desalination 128 (2000) 161 176
172 ~, 800
50
700
C~=055 Xf= 36000 ppm T~,, o =25 c
'k, 'I
600
<
500
~
.
.
1
.
A T = Ts - Tb • 2 • 4 A 6
~ -~ ~: = ~
40
AT = T s - T b • 2 • 4 a 6
--
8
.8
0
w
400
30
x
2
10
\~,
3oo
Cc:O.
"~ zZ 200
~
100 0 40
,
,
,
,
,
,
50
60
70
80
90
100
110
Heating Steam Temperature, °C
~ -E ~
Xf = 36000 ppm 10
~
0
Tcw=25°C
40
,
,
,
,
50
60
70
80
-
-~ 90
100
110
Heating Steam Temperature, °C
Fig. 8. Variation in the s p e c i f i c heat transfer area as a
Fig. 9. Variation in the specific f l o w rate o f c o o l i n g w a t e r
function of the temperature difference of heating steam and top brine temperature and the heating steam temperature,
as a function of the temperature difference of heating steam and top brine temperature and the heating steam temperature.
the feed flow rate of seawater increases to higher values and reduces the flow rate of the cooling seawater. The opposite behavior occurs at lower temperature, where higher conversion ratios are achieved. This reduces the feed flow rate and results in the increase in the cooling seawater flow rate. Variations in the system performance as a function of the heating steam temperature and the temperature difference between the heating steam and the boiling brine are shown in Figs. 7-9. As is shown the two parameters have strong effect on the specific heat transfer area and the specific flow rate of cooling water. As shown in Fig. 8, the increase in the temperature reduces the specific heat transfer area. At larger temperature differences, the driving force for heat transfer increases and results in reduction in the heat transfer area. Simultaneously, this effect increases the amount of distillate product, which increases conversion ratio. As discussed before, increase in the conversion ratio is associated with simultaneous decrease in the feed seawater flow rate and increase in the flow rate of cooling seawater (Fig. 9).
7. Conclusions A novel system for absorption heat pumps combined with single-effect evaporation is analyzed as a function design and operating parameters. The following conclusions are made in light of the results and analysis: • The thermal performance ratio is insensitive to various design and operating parameters. • The thermal performance ratio varies over a range of 2.4-2.8 and is close to 50-70% higher than that of the single-effect thermal vapor compression [ 10]. • Effects of the LiBr mass fraction in the concentrated solution are minimal on the system performance. However, choice of this parameter is dependent on steam availability. • The specific heat transfer area decreases with an increase in the heating steam temperature and the temperature difference of the heating steam and boiling brine. • The specific flow rate of cooling water decreases at higher heating steam temperatures and lower temperature difference between the heating steam and boiling brine.
F. Mandani et al. / Desalination 128 (2000) 161-176 In s u m m a r y , selection o f the o p t i m u m design and o p e r a t i n g c o n d i t i o n s s h o u l d take into consideration attractive features for s y s t e m operation
173
Subscripts a ab
---
Absorber H e a t i n g steam
ac
--
absorber Deceleration component
area. B o t h factors result in c o n s i d e r a b l e savings in the capital and p r o d u c t i o n cost.
b c cw d
-----
Brine C o n d e n s a t e or c o n d e n s e r C o o l i n g water Distillate v a p o r
8. S y m b o l s
e ev f g
-----
Evaporator Entrained v a p o r Feed seawater G e n e r a t o r or c o n c e n t r a t e d solution
H e a t capacity, kJ/kg °C C o n v e r s i o n ratio, defined as M J M j Distillate f l o w rate generated in the evaporator, kg/s E n t h a l p y o f liquid phase, kJ/kg E n t h a l p y o f v a p o r phase, kJ/kg T h i c k n e s s o f d e m i s t e r pad, m m L o g a r i t h m i c m e a n t e m p e r a t u r e dif-
gr m o p r s t v
---------
Gravitational c o m p o n e n t M o t i v e steam in heat p u m p Dilute L i B r solution Demister Frictional c o m p o n e n t H e a t i n g steam T r a n s m i s s i o n line Vapor
ference, °C M a s s f l o w rate, kg/s Pressure, k P a P e r f o r m a n c e ratio, defined as M J Mm
y
--
Static head
at h i g h e r t e m p e r a t u r e s . A t these conditions, a drastic r e d u c t i o n o c c u r s in the specific f l o w rate o f c o o l i n g w a t e r and the specific heat transfer
A BPE C
---
Cp CR D
----
H H" L LMTD
-----
M P PR
----
- -
AP s sA
----
SMcw AT T U
-----
v
H e a t transfer area, m 2 B o i l i n g p o i n t e l e v a t i o n , °C M a s s fraction o f L i B r in solution
generated
in the
LiBr
R e f e r e n c e s
P r e s s u r e drop, kPa Salt c o n c e n t r a t i o n , mg/l Specific heat transfer surface area, m2/(kg/s) Specific f l o w rate o f c o o l i n g water T e m p e r a t u r e drop, °C T e m p e r a t u r e , °C Overall heat transfer coefficient, W/m2K
[1] H.M. Ettouney, H.T. El-Dessouky and I. Alatiqi, Chem. Eng. Prog., 95 (1999) 43. [2] G.F. Leitner, Int. Desalination Water Reuse Q., 7 (1998) 10. [3] J. Bednarski, M. Minamide and O.J. Morin, Proc., IDA World Congress on Desalination and Water Sciences, Madrid, 1 (1997) 227. [4] M.A.DarwishandH. E1-Dessouky, Applied Thermal Engineering, 18 (1996) 523.
--
V a p o r specific v o l u m e , m3/kg
V X
---
V a p o r velocity, m/s Salinity, p p m
Greek P )~
--
Density, kg/m 3 L a t e n t heat, k J / k g
[5] F. A1-Juwayhel, H.T. EI-Dessouky and H.M. Ettouney, Desalination, 114 (1997) 253. [6] H.T. E1-Dessouky and H.M. Ettouney, Simulation of combined multiple effect evaporation-vapor compression desalination processes, 1st IDA Int. Desalination Conference, Cairo, September, 1997. [7] H.T. EI-Dessouky, H.M. Ettouney, H. A1-Fulaij and F. Mandani, Chem. Eng. Proc., in press.
- -
174
F. Mandani et al. / Desalination 128 (2000) 161 176
[8] H.M. Ettouney, H.T. E1-Dessoukyand Y. AI-Roumi, Int. J. Energy Res., 23 (1999) 431. [9] J.M. Veza, Desalination, 101 (1995) I. [10] H.T. EI-Dessouky and H.M. Ettouney, Heat Transfer Eng., 20 (1999)52. [11] C. Temstet, G. Canton, J. Laborie and A. Durante, Desalination, 105 (1996) 109. [12] J. Weinberg and A. Ophir, Ashdod experience and other dual-purpose desalination plants based on multi-effect desalination with aluminum tubes, Symp., Desalination of Seawater with Nuclear Energy, Taejon, Republic of Korea, 1997. [13] Y. Michels, Desalination, 93 (1993) 111. [14] G. Kronenberg, Proc., IDA World Congress on Desalination and Water Science, Abu Dhabi, 3 (1995) 459. [15] H.T. E1-Dessouky, H.M. Ettouney and F. AIJuwayhel, Trans. I. Chem. E., in press, [16] J.L. Yhrelkeld, Thermal Environmental Engineering, 2nd ed., Prentice-Hall, Englewood, N J, USA, 1972. [17] J. Weinberg, A. Ophir and U. Fisher, Proc., 7th Int. Symposium on Fresh Water from the Sea, 1 (1980) 283. [18] G. Alefeld, and F. Ziegler, ASHRAE Tech. Data Bull., June 1985, pp. 11-24. [19] S.E. Aly, Desalination, 68 (1988) 57. [20] K. Fatbalah and S.E. Aly, Energy Convers., 31 (1991) 529. [21] S. Yanniotis and P.A. Pilavachi, Chem. Eng. Technol., 19 (1996) 448. [22] H.T. El-Dessouky, I. Alatiqi, S. BingulacandH.M. Ettouney, Chem. Eng. Yech., 21 (1998) 15. [23] O.A. Hamed, A.M. Zamamiri, S. Aly and N. Lior, Energy Convers. Mgmt, 37 (1996) 379. [24] I. Alatiqi, H.M. Ettouney and H.T. EI-Dessouky, Desalination, 126 (1999) 33. [25] H.T. E1-Dessouky, I.M. Alatiqi, H.M. Ettouney and N.S. AI-Deffeeri, Chem. Eng. Process., 39 (2000) 129. [26] H.M. Hellmann and G. Grossman, ASHRAE Trans Symp., 1996, pp. 980-997.
Appendix A: Model correlations The correlation for the demister pressure drop was developed by E1-Dessouky et ai. [25]. The correlation was conducted over the following parameter range, 1 _
(27)
where 9p is the demister pad density, Lp is the pad thickness, V is the vapor velocity in the demister pad. The BPE is obtained as a function of the brine salinity and temperature. The value of BPE is obtained from the following empirical correlation, which is valid for 20,000 _
(28)
with B =-(6.71 +6.34 x 10 2 T + 9.74 × 10-5 T2) 10-3
C =(22.238+9.59x10 3T+9.42× 10-5T2) 10 8
where the BPE is in °C. The seawater specific heat, Cp, is given by the following correlation: C = (A + B T + C T 2 +DT3)x l0 -3 P
(29)
The variables A, B, C and D are evaluated as a function of the water salinity as follows: A = 4206.8 - 6.6197S+1.2288 × 10-2S e
175
F. Mandani et al. / Desalination 128 (2000) 161-176
B = - 1 . 1 2 6 2 + 5.4178x10 -2S-2.2719x10 4S 2 C=1.2026x10 -2- 5.3566×10 4S+ 1.8906x10-6S2 D -- 6.8777 x 10 -7 + 1.517 x 10 -6 S- 4.4268 × 10 9 ~/2 where C o is in kJ/kg°C, Tin °C, and S is the water salinity in gm/kg. The above correlation is valid over salinity and temperature ranges of 20,000 < X _<160,000ppm and 20_< T_< 180°C, respectively, The correlation for the water vapor saturation pressure is given by
where v is in m3/kg and T is in °C. The above correlation is valid over a temperature range of 25-110°C. The percentage errors for the calculated vs. the steam table values are less than 10%. The latent heat correlation for the water vapor is )~ = 2589.583 + 0.9156 T-4.8343×10 2 T 2
(33)
where T is in °C and ~. is in kJ/kg. The above correlation is valid over a temperature range of 10-140 °C with errors less than 0.4% for the calculated and the steam table values.
P = 10.17246-0.6167302(7) + 1.832249 The density of seawater is given by x 10 -2 (Z) 2 - 1 . 7 7 3 7 6 x 10 -4 (T) 3
(30)
+ 1.47068× 10 6(7)4
9=103(A,Fl+A2F2+A3F3+A4F4)
(34)
The parameters in the above equation are given by where P is in kPa and T is in °C. The above correlation is developed over a temperature range of 10-110°C with percentage errors less than 2% for the calculated and the steam table values. The saturation temperature correlation is given by 3892.7 T=
42.6776- [ln(P/l~.48654])
) |
B = [(2)(s)/1000- 150]/150 G1 = 0.5 G~ = B G3 = 2B2 - 1
(31)
A~ = 4 . 0 3 2 2 1 9 G j + 0 . 1 1 5 3 1 3 G 2 + 3 . 2 6 x 1 0 4 G 3
-273.15 A2 = -0.108199 G1 + 1.571 x 10 -3 G2- 4.23 × 10 -4 G3 where P is in kPa and T is in °C. The above correlation is valid for the calculated saturation
A3 = - 0 . 0 1 2 2 4 7 G ~ + l . 7 4 x 1 0 -3 G2-9x10-6G3
temperature over a pressure range o f 10-1750 kPa.
A 4 = 6.92x10 -4 G~ -8.7×10 5 G2 _5.3×10-5 G3
The percentage errors for the calculated vs. the steam table values are less than 0.1%. The correlation for the saturation water volume
A -- [(2)(7) - 200]/160
is given by
F~ = 0.5, F 2 =A, F3 = 2 A 2 - 1 , F 4 = 4A3 - 3 A
v = 163.3453019 - 8.041421773 T +0.171021164 T 2 -.001.878124x10 3T3 (32) + 1"03842x10-5 T4 -2"28215x10-8 T5
where 9 is in kg/m 3, s is in gm/kg, and Tis in °C. This correlation is valid over 0_
1~ Mandani et al. / Desalination 128 (2000) 161-176
176
The v a p o r enthalpy o f pure water is given by
c~ = 40.2847, c2~ = 39.9142, c3~ = 33.3572
% = 13.1032, co2 = - 0 . 1 8 6 0 5 1 , c12 = - 0 . 1 9 1 1 9 8
H " = 2 5 0 0 . 1 5 2 + 1.947036 T - 1.945387 (35)
× 1 0 -3 T 2
c22 = 0.199213, c32 = - 0.178258, c42 = - 0.0775101 with a range o f 0.01 - 145°C and R z = 0.9999. The liquid enthalpy o f pure water is given
co3 = - 7.51277 E - 6 , c~3 = 0, c23 = 0, c33 = 0, c43
=
0
by H = 0.5802129 + 4.151904 T + 3 . 5 3 6 6 5 9 (36)
x 1 0 -4 T 2
with a range o f 0 . 0 1 - 1 4 5 °C and R 2 = 0.9999. H e l l m a n n and G r o s s m a n [26] developed the e n t h a l p y correlation for saturated LiBr solution, w h i c h is given by 4
~o lo T (P, C) -- ~ a~X ~+ T w ~ b~C ~ i~o i-o
(38)
2
H ( C , T) = X~_, a, T ~+(1 -C) ~ i-0
H e l l m a n n and G r o s s m a n [26] developed the boiling temperature o f the LiBr solution, which is given by
b~ T ~
i=o
In Eq. (38) Tw is the saturation temperature o f pure (3 7)
4 3 + X ( 1 - X ) ~ ~ cij(2C-1)~T j 0 ~-o
water at pressure P, C is the mass fraction o f L i B r in the solution, P is the pressure, and T is the boiling temperature o f the LiBr solution. The constants in the above relation are as follows:
w h e r e C is the m a s s fraction o f LiBr and T i s the solution temperature. The constants in the above relation are as follows:
a0 = 0, al = 16.634856, a 2 = - 5 5 3 . 3 8 1 6 9
ao = 508.668, a~ = 18.6241, a2 = 0.0985946,
a 6 = - 2 1 1 1 2 5 6 . 7 , a7 = 4,385,190.1
a3 = - 2 . 5 0 0 9 7 9 E - 5 , a4 = 4.15801 E - 8
a8 = -5,409,811.5, a 9 = 3,626,674.2
bl = 1.617155702, b 2 = 4 . 1 0 1 8 7 4 8 5 ,
alo = -1,015,305.9, bo = 1, bl = - 0 . 0 6 8 2 4 2 8 2 1
b3 = 0.000717667
b2 = 5.873619, b3 = - 102.78186, b4 = 930.32374
Coo -- - 1021.61, C~o = - 533.08, C2o = 483.628
b5 = - 4822.394,
C3o = 1155.13, C4o = 640.622, col = 36.8773
b8 = 34100.528,
a3 = 11228.338, a4 = - 110,283.9, a5 = 621,094.64,
b 6=
b9=
15,189.038,
b 7=
-
29,412.863
- 2 1 , 6 7 1 . 4 8 , blo = 5799.56
Desalination 128 (2000) 161-176 www.elsevier.com/locate/desal
LiBr-H20 absorption heat pump for single-effect evaporation desalination process •
•a*
Falsal Mandam , Hisham Ettouney b, Hisham E1-Dessouky b aCollege of Technological Studies, PO Box 42325, Shuwaikh 70654, Kuwait Fax +965 481-3691 bChemical Engineering Department, College of Engineering and Petroleum, Kuwait University, PO Box 5969, Safat 13060, Kuwait Received 28 October 1999; accepted 1 December 1999
Abstract
A new configuration is presented for the single-effect evaporation process combined with lithium bromide water (LiBr-H20) solution absorption heat pump. Performance evaluation of the process is presented in terms of variations in the thermal performance ratio, the specific heat transfer area, the specific flow rate of cooling water, and the conversion ratio. These variations are presented as a function of the heating steam temperature, the lithium bromide mass fraction in the concentrated solution, and the temperature difference of the heating steam and the boiling brine. Results show thermal performance ratios of 2.4-2.8, which are higher by 50-70% than those for the single-effect thermal vapor compression system. Operation at higiier heating steam temperatures and large temperature difference between the heating steam and the boiling brine is highly attractive due to drastic reduction in the specific heat transfer area and the specific flow rate of cooling water. Also, at these conditions slight reduction occurs in the system thermal performance ratio. In search for low-energy and low-cost production units, the superior performance of high temperature operation for the absorption heat pump single-effect system makes it extremely suitable for adoption by remote and small communities. Keywords." Water desalination; Single-effect evaporation; Vapor compression; Absorption heat pump; Modeling; Simulation
1. I n t r o d u c t i o n At the turn o f the millennium, securing sustainable resources for fresh water and energy *Corresponding author,
is a major challenge for mankind. Desalination of sea and brackish water has developed considerably during the 20th century to provide and support humane and industrial activities in several zones around the world, especially in the
0011-9164/00/$- See front matter © 2000 Elsevier Science B.V. All rights reserved PII: S0011 - 9 1 6 4 ( 0 0 ) 0 0 0 3 I - X
162
F. Mandani et al. / Desalination 128 (2000) 161 176
Gulf States, other middle eastern countries, southern California and Florida, Italy, and Spain. Developments have resulted in configuration of several forms for thermal desalination, which could be based on evaporation or flashing, operation in single or multiple effect, or designed as a stand-alone system or combined with vapor compression heat pumps [1]. Regardless, the desalination industry remains to be expensive for the majority of the world population because of its high capital and production cost. Recent quotes for funding requirements of a 20migd desalination plant based on reverse osmosis (RO), multi-stage flash desalination (MSF), single-effect mechanical vapor compression (MVC), or multiple-effect evaporation (MEE) vary between $69 to $95 million [2]. In addition, Bednarski and Minamide [3] reported a unit production cost of $0.8/m 3 for 6migd MSF, $0.72-0.93/m 3for RO (the cost range depends on the pretreatment cost), and $0.45/m 3 for lowtemperature MEE. This relatively high unit product cost is a result of the high energy cost, which may vary between 40-60% of the unit product cost [4]. In an attempt to reduce the energy cost in desalination and other industrial process, use of heat pumps has recently been considered. Heat pumps receive energy from a low-temperature source and upgrade to a higher temperature before performing the desired function. Several forms of heat pumps exist in the literature and are suitable for operation withthermal-based desalination processes. Examples include thermal vapor compression (TVC), MVC, adsorption vapor compression (ADVC), and absorption vapor compression (ABVC). Examples for performance of these heat pumps in thermal desalination processes can be found in the studies by AI-Juwayhel et al. [5] on single-effect evaporation, EI-Dessouky and Ettouney [6] on MEE systems, and E1-Dessouky et ai. [7] on MSF combined with thermal vapor compression.
Single-effect evaporation desalination units combined with vapor compression have several attractive features that make them suitable for providing sustainable water supplies for industrial sites, remote small communities, and scattered remote populations. Currently, the best available technology in this category is the single-effect MVC process with unit capacities up to 5000m3/d [8]. The process has highly competitive specific power consumption against the RO process with values ranging between 48kWh/m 3. Similar to other thermal desalination processes, the MVC process has reliable operation, the ability to withstand harsh operating conditions of high feed temperature and high salinity, and has a high plant factor [9]. The major drawback of the MVC process is that its main energy source is electric current, which is very expensive in comparison with saturated steam used in other thermal desalination processes. In addition, a major component in the MVC process is the mechanical compressor, which has moving parts and therefore requires specialized and more frequent maintenance frequency. Other single-effect evaporation desalination processes include the TVC system, the ABVC, and the ADVC. The single-effect TVC process is not found on industrial scale because of its relatively lowthermalperformanceratio, defined as the ratio of product water per unit mass heating steam, with values averaging 1.5 [10]. However, the multiple-effect thermal vapor compression process is the standard of the MEE industry. The multiple-effect configurations have flexible design features that allow operation in a stand-alone mode or combined with the thermal vapor compression heat pump. The thermal performance ratio of the process is high; for example, a 12-effect system has a performance ratio of 8 in a stand-alone mode and of 16.7 in the vapor compression mode [ 11]. Weinberg and Ophir [12] reported similar trend for a stand-
F. Mandani et al. / Desalination 128 (2000) 161-176
alone six-effect system with a performance ratio of 5.7. Also, Michles [13] reported a thermal performance ratio of 8 for a four-effect system combined with TVC. Kronenberg [ 14] showed that unit product cost for various combinations of the multiple-effect system with power generation units, i.e., diesel, steam turbines, or gas turbine cycles, varies between $0.27/m 3 to $0.56/m 3. On the other hand, the unit product costs for the RO and MSF system are higher with values of $0.68/m 3 and $0.89/m 3. It should be stressed that multiple-effect TVC is the industrial standard because of its higher thermal performance ratio in comparison with single-effect TVC. On the other hand, the singleeffect MVC is the industrial standard. This is because the specific power consumption has the same value for the single- and multiple-effect MVC configurations [5-15]. The main difference between the single- and multiple-effect MVC systems is the increase in the system capacity for the multiple-effect systems due to the increase in the compression range. Although, the combination of the ABVC and ADVC systems with thermal desalination provide high performance in comparison with conventional configurations, their use is not found on full commercial or industrial scale. The literature includes a large number of studies on development, innovation, and performance evaluation of ABVC and ADVC for refrigeration and air conditioning processes [16]. On the other hand, evaluation of the combined systems of these heat pumps and various thermal desalination processes is limited to a small number of studies in the literature. Weinberg et al. [ 17] evaluated the performance of a coupled system of MEE vacuum freezing with a lithium bromide absorption heat pump. The system is thought to enhance the performance ratio of the MEE to high values of 18-20 and operating temperatures between 0-60°C. The low-temperature operation minimizes corrosion and scaling problems. Alefeld
163
and Ziegler [18] proposed a fully integrated desalination system combined with LiBr-H20 absorption heat pump. The system includes three stages which process seawater and generate fresh water. Aly [19] and Fathalah and Aly [20] analyzed a solar-powered LiBr-H20 heat pump, which generates high-grade steam to operate a MEE desalination system. In their analysis, more emphasis was given to the performance of the solar power unit and air conditioning in the evaporator unit. Yanniotis and Pilavachi [21] modeled the performance of sodium hydroxide heat pumps in the MEE systems. Model results are validatedagainstexperimentalmeasurements and were found to have reasonable agreement. A1-Juwayhel et al. [5] studied the performance of single-effect evaporation desalination systems combined with various types of heat pumps. As discussed before, results for the ABVC and ADVC gave thermal performance ratios close to three times higher than the single-effect TVC system. On the other hand, E1-Dessouky and Ettouney [6] showed a 50% increase in the thermal performance ratio for the MEE-ABVC and MEE-ADVC systems over the MEE-TVC with values close to 20. This paper evaluates the performance of a novel process constituting a single-effect evaporation desalination combined with absorption heat pumps. This process integrates the exothermic heat of absorption within the desalination system, instead of providing hot utility water. The study includes development of the proposed process and mathematical model. Results and system performance are presented as a function of the design and operating parameters that have a strong effect on the product water cost.
2. Description of the ABVC process Elements of the ABVC system are shown in Fig. 1. In this system, the evaporator constitutes
164
F. Mandani et al. / Desalination 128 (2000) 161-176
~l Cooling ~---~ Il EntrainedI- -gap°r' T [~l"Mev' ' Tv I
~
Mcw' Seawater Xcw' Tf
I FeedSoawator M, X.w ", I Absorber
Heat
Intake Seawater Mf + Mcw, Xcw, Tcw
~
I Dilute LiBr-H20 Solution Mo, Xo
Feed brine, | [ Ma, Xa, Ta [---]
! ~ ~iiWl --
<
SolutionC°nc" LiBr-H20 Mg, Xg
..~ ~
Vapor, D
~
~ 7 ......
Down Condenser
.... ~
~ 7
[ Heating
,,i ~: : ~ ,imi .............. '
~ Motive Steam, Tin, Pm
Fa Film Evaporator
I ~1 ~ Blowdown Brine Mb' Xb' Tb
I Generator ~'~ [Cond . . . . te l
Fig. 1. Schematic of single-effect absorption vapor compression desalination process (ABVC).
horizontal falling film tubes, brine spray nozzles, demister, and the brine pool. The down condenser has a similar shell and tube configuration where condensation takes place on the shell side. The absorber is also a shell and tube falling film configuration. Absorption of water vapor by the LiBr-H20 solution occurs on the shell side of the absorber, while heating of the feed seawater and vapor formation take place on the tube side of the absorber. The generator has a similar layout to the evaporator where dilute LiBr-H20 solution forms a falling film on the outside surface of the tubes and the motive steam is condensed inside the tubes. The heat exchange unit between the concentrated and diluted LiBr-H20 solution is used to recover part of the energy from the higher temperature andconcentratedLiBr-HzOsolution, This improves the overall process efficiency, The process is described in the following points: • The intake seawater stream flows through the down condenser, where it condenses part of the vapor formed in the evaporator. The
•
•
•
remaining part of the vapor is fed to the shell side of the absorber in the heat pump. The intake seawater temperature increases from (Tcw) to (Ts) as it absorbs the latent heat of condensation of the condensing vapor. Part of the feed seawater is rejected back to the sea (Mc,~) which is known as the cooling seawater. The remaining portion of the intake seawater is the feed seawater stream (Ms), which is chemically treated and deaerated before being fed to the tube side of the absorber. The concentrated LiBr-H20 solution absorbs the vapor stream entering the absorber. The absorption process is exothermic and releases a sufficient amount of heat that sustains an increase of the feed seawater temperature to the saturation temperature. Also, vapor is formed from the feed seawater within the absorber. This vaporformspartoftheheating steam in the evaporator. The temperature of the absorbed vapor and the concentration of the outlet dilute LiBr-H20 solution define the equilibrium
F. Mandani et al. / Desalination 128 (2000) 161-176
•
•
•
•
conditions in the absorber. It should be noted that the boiling point elevation for the LiBr-H20, or the temperature difference between the dilute LiBr-H20 solution and the absorbed vapor varies over a range of 1050°C as the mass fraction of LiBr-H20 in the dilute solution is increased from 0.25 to 0.45. The dilute LiBr-H20 solution enters the generator where it is sprayed on the outside surface of the tubes. The solution absorbs the latent heat of motive steam that condenses on the tube side of the generator. The heating process increases the temperature of the LiBr-H20 solution to saturation and results in evaporating the same amount of water absorbed by the solution in the absorber. The concentration of the concentrated LiBr-H20 solution and the temperature of the formed vapor define the equilibrium conditions in the generator. The combined vapor formed in the generator and absorber derives the evaporation process in the evaporator. The brine stream leaving the absorber is sprayed on the outside surface of the evaporator tubes where it absorbs the latent heat of condensation from the condensing steam on the tube side of the evaporator. The concentrated brine leavingthe evaporator is rejected back to the sea and the formed vapor is routed to the down condenser. The sum of the condensate of the heating steam and the condensate in the down condenser forms the distillate product stream. Demisters in the generator, absorber, and evaporator prevent droplet entrainment of the LiBr-H20 solution or brine in the vapor stream.
3. Mathematical model of ABVC The steady-state model includes a set of material and energy balances, heat transfer equations, and thermodynamic relations.
165
Assumptions used in the model include: • The vaporformedintheevaporator, absorber and generator is salt free; this assumes that the entrainment of brine droplets by the vapor stream is negligible and has no effect on the salinity of the distillate product. • Energy losses from the evaporator to the surroundings are negligible; this is because of operation at relatively low temperatures, between 100-40 ° C, and the evaporator is well insulated. • The physicalproperties of various streams are calculated at the average temperature and concentration of influent and effluent streams. The overall material and salt balances are given by Mf = M d + M b Mb --
Mf~(f/Xb)
(1) (2)
where M is the mass flow rate, X is the salt concentration, and the subscripts b, d, and f denote the brine, distillate, and feed seawater. In Eq. (2) the brine blowdown salinity (Xb) is set at 90% of the saturation salinity of the CaSO4 solution X b = 0.9 (457,628.5 - 11,304.11 Tb + 107.5781 Tb2 -0.360747 Tb3)
(3)
This equation is obtained by curve fitting of the salinity/temperaturerelation for the solubility of CaSO4 [ 15]. In the evaporator the saturated falling brine film absorbs the latent heat of the condensing steam. This evaporates a controlled mass of vapor, D at T~, where M Z : D~. v
(4)
F. Mandani et aI. / Desalination 128 (2000) 161-176
166
where )~ is the latent heat. The subscripts s and v denote the heating steam and the vapor formed, respectively. In the evaporator, absorber, and generator the boiling temperatures are higher than the corresponding vapor saturation temperatures by the boiling point elevation (BPE), and the temperature rise is caused by the hydrostatic pressure head, ATy. This is Tb = Tv +
BPE + A T
(5)
The term ATy is negligible in horizontal falling films because of the very small thickness of the boiling film. The condensation temperature in the down condenser, Tc, is lower than the boiling temperature, Tb, by the saturation temperature depressions associatedwithpressurelossesinthe demister (APp), transmission lines (AP 0, and vapor condensation inside the tubes (APe). The resulting condensation temperature is
This energy balance is given by
(D-Mev))~c=(Mw+Mf)Cp(Tf-Tw)
where the subscripts c, cw, and ev denote the condensing vapors, the intake seawater, and the entrained vapor in the absorber. The following relation gives the flow rate of the heating steam: M = M v +MaD
(9)
where Mev is the amount of entrained vapor in the absorber, subsequently released in the generator as a part of the heating steam. Mab is remaining part of the heating steam generated in the absorber. Inspection of Fig. 1 shows thatthetotal distillate flow rate is given by Ma = D + M b
T : T b -(BPE + A T + ATt + ATe)
(8)
(lO)
(6) The energy balance around the absorber is given by
The pressure drop during condensation, APc, is defined as the algebraic sum of the decrease caused by friction, APr, and the increase caused by gravity, Apg, and vapor deceleration AP a. This relation is given by
AP c = (AP r - AP i - APo)
(7)
Correlations for the pressure drop components App, APt, APr, Apg, and APa are given in the study by EI-Dessouky et al. [22]. In the down condenser, the temperature of the intake seawater, MCw+Mf, is increased from T~wto Tf. Condensing part of the vapors formed in the evaporator provides the required heating energy,
Mg/-/g + MevH"ev + MfHf
= (rag + m v ) H + mab H" s + M aH a
(11)
where Mg is the flow rate of the concentrated LiBr-H20 solution entering the absorber, and M. is the brine mass flow leaving the absorber. The water vapor saturation enthalpies H"ev and H" s are obtained at the condensation temperature in the down condenser (To) and the heating steam temperature in the evaporator (Ts), respectively. In Eq. (1 I) Hg and/4o are the enthalpies of the concentrated and diluted LiBr-H20 solution evaluated at (Cg, Tg) and (Co,To). It should be noted that Tg and To are obtained from the
F. Mandani et al. / Desalination 128 (2000) 161-176
equilibrium relation [Eq. (32) in the appendix] at water vapor saturation temperatures of T~ and To, respectively. The enthalpies of the feed seawater and the feed brine, He and Ha, are calculated at (Tf, Xow) and (Ta, Xa), respectively. The heating steam temperature is related to the feed brine temperature by the boiling point elevation, or
T= T~-BPE~(a,T )
(12)
condenser. For the evaporator, the heat transfer area, Ae, is M )~ A e
(18) Ue (T_Tb)
where U is the overall heat transfer coefficient, and the subscript e refers to the evaporator. The heat transfer area of the down condenser is given by
The energy equation for the generator balances the amount of input energy in the motive steam and the dilute LiBr solution against the amount of output energy in the concentrated LiBr solution and the heating steam. This relation
A -
is given by
(LMTD) c _
(J~lfg+mev)Ho + mmO"m = mgHg, + McvH" s
(13)
167
(D-Mv))~ c
c U (LMTD)c
(19)
Tf- T~w In Tc T~w
(201
re- L.
where M m and H " m are the mass flow rate and enthalpy of motive steam. The material and salt balances around the absorber for the feed seawater and the feed brine are given by
EI-Dessouky et al. [22] developed the correlations for the overall heat transfer coefficient in the evaporator and condenser: Ue = 1.9394 + 1.40562 x 10 -3 Tb -2.0752
Mf = Ma + (M~-M~v)
(14)
(21)
xlO 4(rb)2 +2.3186x 10 6(rb)3 Xf -~iff = MaX a
(15) Uc = 1.6175 +0.1537x 10 -3 T +0.1825 (22)
Similarly, the following relations give the total mass and salt balances for the concentrated and diluted LiBr-H20 solutions: M o = Mg +Me v
(16)
MoC ° =MgCg
(17)
The design equations for the heat transfer area are developed for the evaporators and the down
x 10 3(T)2-8.026 × 10-8(T~)3 where Ue and Uc are the overall heat transfer coefficient in the evaporator and down condenser in kW/m 2 °C, Tb is the brine boiling temperature, and T~ is the vapor condensation temperature in the condenser. All temperatures in the above correlations are in °C. The standard deviations for the above correlations are 2.03 % and 1.76%. These correlations are tested and proved to be
168
b~ Mandani et al. /Desalination 128 (2000) 161-176
reliable through comparison against other correlations in the literature and available experimental and design data.
4. Performance parameters The system performance is defined in terms of the following parameters: • performance ratio, which is defined as the amount of the distillate product per unit mass of the motive steam P R = Md/Mm
•
specific flow rate of cooling water, which is defined as the amount of the cooling water per unit mass of distillate product
sMcw = M~w/Md
•
(24)
specific heat transfer area, which is defined as the ratio of the total heat transfer area of the evaporator and condenser to the total flow rate of distillate product
sA = ( A e + A , ) / M d
•
(23)
(25)
conversion ratio, which is defined as the amount of distillate product per unit mass of feed seawater
CR = MJMf
(26)
5. Solution method The solution procedure is shown in Fig. 2. Solution of the model equations requires definition of the following system variables: • distillate flow rate, Md, is 1 kg/s • intake seawater temperature, Tow,is 25 °C • heating steam temperature, Ts, is higher than brine boiling temperature, Tb, by 2-10°C
•
feed seawater temperature, Tf, is lower than the vapor condensation temperature Tc by 5 °C • feed brine temperature, Ta, is lower than the temperature of the dilute LiBr-HzO solution To by 5°C • motive steam temperature, Tin, iS higher than the temperature of the concentrated LiBr-H20 solution Tg by 5 °C • range for the heating steam temperature, Ts, is 50_100°C • feed seawater salinity, Xf, is 36,000 ppm. As shown in Fig. 2, the solution sequence proceeds as follows: • The system capacity, stream temperatures, and stream salinity are defined as specified above. • Eqs. (1)-(3) are solved to determine the feed and brine flow rates and the salinity of the brine blowdown. • An initial guess is assumed for the following: (1) evaporator area; (2) flow rates of the heating steam, motive steam, entrained vapor, vapor formed in the absorber, dilute LiBr-H20, and the concentrated LiBr-H20; (3) concentrations of the dilute LiBr-H20 solution and brine leaving the absorber. • The above variables are calculated by solution of Eqs. ( 4 ) a n d (9)-(18). Solution proceeds iteratively using Newton's method. Iterations continue until the tolerance criterion is achieved with a value of 1 × 10 4 for e. • The flow rate of the cooling seawater and the heat transfer area of the condenser are calculated from Eqs. (8) and (19), respectively. • The performance parameters are calculated from Eqs. (23)-(26).
6. Results System performance is evaluated as a function of the heating steam temperature, the temperature difference of the heating steam and the boiling
169
F. Mandani et al. /Desalination 128 (2000) 16l--176
Define System Temperatures and Stream Salinity:
Md, Tb, Tcw, Tin, Tf, Ts, Cg, Xf, Xd
Calculate Brine salinity and Flow Rates of Brine and Feed Streams:
Mfand Mb from Eqs. (1 and 2) Xb from Eq. (3)
Calculate Initial Guess (X°):
Ae, Ms, Mev, Mab, Mm, Ma, Mo, Mg, Xa, Co, D
Solve the E q u a t i o n s a n d O b t a i n N e w P r o f i l e s (X1): 1
Ae, Ms, Mev, Mab, Mm, Ma, Mo, Mg, Xa, Co, D
C h e c k I t e r a t i o n s Error: m
( y
No
(x ° _
<
i=1 Yes~ Design the Down Condenser: Calculate Ac and Mcw from Eqns. 19 and 8
Calculate Performance Parameters:
Eqs. (23-26) PR, sMcw, sA, and CR
Fig. 2. Solutionalgorithmof the absorptionheatpumpand the singleeffectevaporationdesalinationsystem.
170
F. Mandani et al. / Desalination 128 (2000) 161-176
brine, and the mass fraction of LiBr in the concentrated solution. Effects of the heating steam temperature and the mass fraction of the LiBr in the concentrated solution are shown in Figs. 3-5 for the variations in the performance ratio, the specific heat transfer area, and the specific flow rate o f cooling water. The three figures show insensitive and independent behavior o f the system parameters on the mass fraction of LiBr in the concentrated solution. The following cause this behavior: • Various system temperatures including the heating steam, the boiling temperature, the feed seawater, and the feed brine, are independent o f the LiBr mass fraction in the concentrated solution. These temperatures affect the system variables used to calculate the system performance parameters. • Increase in the LiBr mass fraction in the concentrated solution affects only the flow rate of the concentrated solution. At higher
4 Yb =Ts-3 xf= 36000 plan
3.5 ~ 3 ~ ~ 2.5 :2
Tcw = 25 °C
__ ~ *=====~" "
2
1.5 |
o.4
i
i
0.5
0.6
0.7
~450 ~c ~ E 400
~ ~, " ~
~ 35o ~
cc • 0.45 , 0.55 •A 0.5 0.6
absorber and generator.
~' 250
Tb=Ys - 3 Xf = 36000 ppm
'~
Tcw = 25
N
~ 200 40
. 50
~
.
. 60
. 70
. . 80 90
Heating Steam Temperature,
100
110
°C
Fig. 4. Variation in the specific heat transfer as a function of the mass fraction of LiBr in concentrated solution and the heating steam temperature. .
• Selection between the above two points is dependent upon several factors including: • availability ofhighpressure steam, i.e., 17bar vs. 3 bar
0.8
Mass Fraction of L~3r h Concentrated So krtion
Fig. 3. Variation in the performance ratio as a function of the mass fraction of LiBr in concentrated solution and the heating steam temperature.
The main effect o f increasing the LiBr mass fraction in the concentrated solution is the need for higher pressure motive steam. This is illus-
•
i
~ 300
Cg = 0.75, Tg = 201.87°C, Tm = 206.87°C, Pm=1789 kPa (17.89 bar), P R = 2.4, sA = 220.9, sMcw = O, CR = 0.092. Cg=O.45, T g = 1 2 2 . 9 3 ° C , T m = 1 2 7 . 9 3 ° C , P ~ 253 kPa (2.53 bar), P R = 2.3, sA = 212.9, sMcw = O, CR = 0.092.
~
Ts
.~
•
~
• 100 -- ~- - 8 050 --,- 65
concentrations, the solution enthalpy increases and results in reduction of the concentrated solution flow rate. This is necessary to balance the system energy in the
trated in the following data obtained for a heating steam temperature of 100°C:
*
•
increase in the system second law efficiency upon operation at low steam pressures [4-23] use of smaller tube diameter for higher pressure steam [7] elimination o f the control loops required for reduction of the steam pressure to lower values of 3 bar [24].
F. Mandani et al. / Desalination 128 (2000) 161-176
1.75
16
E+05
14 ~: 12
Cc --*-- 0.45 ~ 0.55 • 0.5 -----0.6
\ \
"N~10
6 ~:
~
171
~
Tb =
4
'xL_
2
Ts - 3
xf = 36000 pwn
CaSO4.2H20
~
g
+
~
--60
T ~ , , ,T~w . = 25 oC
r~ g] 0
40
"i 50
~i 60
"i 70
i
80
~
-,,~
--
90
a
1.40E+05 HeatingSteam E Temperature ~ ---~-100c
SO4.0.SH~
70 C C
3.50E+04
-
100
110
00oz+00 0
HeatingSteam Terr~rature, °C
20
40
60
80
100
120
T e m p e r a t u r e , ,<2
Fig. 5. Variation in the specific flow rate of cooling as a function of the mass fraction of LiBr in concentrated solution and the heating steam temperature. As shown in Figs. 3-5, the effects of the heating steam temperature on the system performance are more pronounced than the mass fraction of the LiBr in the concentrated solution. This gives a dramatic reduction in the specific heat transfer area and the specific flow rate of cooling water. As shown in Fig. 4, the specific heat transfer area has values above 400 m2/(kg/s) at heating steam temperatures close to 50°C. On the other hand, the specific heat transfer area decreases to values between 200-250mZ/(kg/s), which are considered the industrial practice, as the temperature is increased to values between 80-100 ° C. This behavior is primarily caused by an increase in the overall heat transfer coefficient at higher temperatures. This enhances the heat transfer rate and results in reduction of the required heat transfer area. Another factor with a lesser effect is the reduction in the latent heat of evaporation or condensation at higher temperatures, which results in reduction in the thermal load of the evaporator and the down condenser, It should be stressed that the temperature difference or the driving force between heating steam and boiling brine has no effect here because it is kept constant in the calculations.
Fig. 6. Calcium sulfate solubility and top brine temperature for the ABVC system.
3.5 3 .~ 2.5 ~ ~o
2
AT=Ts "Tb •* 4 2
1.5
.L 6
1 4o
. 8 x 10 . . . . 50 60 70
Cc =0.55 xf = 34000 ppm Tcw=25°C . . 80 90 100 110
Heathg Steam Temperature, °C
Fig. 7. Variation in the performance ratio as a function of the temperature difference of heating steam and top brine temperature and the heating steam temperature.
At higher heating steam temperatures, the specific flow rate of cooling water is zero (Fig. 5). This is because of the limitations imposed on the salinity of the brine blowdown stream (Fig. 6 ) w h e r e at h i g h e r t e m p e r a t u r e s t h e difference in the salinity o f the feed seawater and the brine blowdown is less than 1000-2000 ppm. As a result of the constant production capacity,
1~ Mandani et al. // Desalination 128 (2000) 161 176
172 ~, 800
50
700
C~=055 Xf= 36000 ppm T~,, o =25 c
'k, 'I
600
<
500
~
.
.
1
.
A T = Ts - Tb • 2 • 4 A 6
~ -~ ~: = ~
40
AT = T s - T b • 2 • 4 a 6
--
8
.8
0
w
400
30
x
2
10
\~,
3oo
Cc:O.
"~ zZ 200
~
100 0 40
,
,
,
,
,
,
50
60
70
80
90
100
110
Heating Steam Temperature, °C
~ -E ~
Xf = 36000 ppm 10
~
0
Tcw=25°C
40
,
,
,
,
50
60
70
80
-
-~ 90
100
110
Heating Steam Temperature, °C
Fig. 8. Variation in the s p e c i f i c heat transfer area as a
Fig. 9. Variation in the specific f l o w rate o f c o o l i n g w a t e r
function of the temperature difference of heating steam and top brine temperature and the heating steam temperature,
as a function of the temperature difference of heating steam and top brine temperature and the heating steam temperature.
the feed flow rate of seawater increases to higher values and reduces the flow rate of the cooling seawater. The opposite behavior occurs at lower temperature, where higher conversion ratios are achieved. This reduces the feed flow rate and results in the increase in the cooling seawater flow rate. Variations in the system performance as a function of the heating steam temperature and the temperature difference between the heating steam and the boiling brine are shown in Figs. 7-9. As is shown the two parameters have strong effect on the specific heat transfer area and the specific flow rate of cooling water. As shown in Fig. 8, the increase in the temperature reduces the specific heat transfer area. At larger temperature differences, the driving force for heat transfer increases and results in reduction in the heat transfer area. Simultaneously, this effect increases the amount of distillate product, which increases conversion ratio. As discussed before, increase in the conversion ratio is associated with simultaneous decrease in the feed seawater flow rate and increase in the flow rate of cooling seawater (Fig. 9).
7. Conclusions A novel system for absorption heat pumps combined with single-effect evaporation is analyzed as a function design and operating parameters. The following conclusions are made in light of the results and analysis: • The thermal performance ratio is insensitive to various design and operating parameters. • The thermal performance ratio varies over a range of 2.4-2.8 and is close to 50-70% higher than that of the single-effect thermal vapor compression [ 10]. • Effects of the LiBr mass fraction in the concentrated solution are minimal on the system performance. However, choice of this parameter is dependent on steam availability. • The specific heat transfer area decreases with an increase in the heating steam temperature and the temperature difference of the heating steam and boiling brine. • The specific flow rate of cooling water decreases at higher heating steam temperatures and lower temperature difference between the heating steam and boiling brine.
F. Mandani et al. / Desalination 128 (2000) 161-176 In s u m m a r y , selection o f the o p t i m u m design and o p e r a t i n g c o n d i t i o n s s h o u l d take into consideration attractive features for s y s t e m operation
173
Subscripts a ab
---
Absorber H e a t i n g steam
ac
--
absorber Deceleration component
area. B o t h factors result in c o n s i d e r a b l e savings in the capital and p r o d u c t i o n cost.
b c cw d
-----
Brine C o n d e n s a t e or c o n d e n s e r C o o l i n g water Distillate v a p o r
8. S y m b o l s
e ev f g
-----
Evaporator Entrained v a p o r Feed seawater G e n e r a t o r or c o n c e n t r a t e d solution
H e a t capacity, kJ/kg °C C o n v e r s i o n ratio, defined as M J M j Distillate f l o w rate generated in the evaporator, kg/s E n t h a l p y o f liquid phase, kJ/kg E n t h a l p y o f v a p o r phase, kJ/kg T h i c k n e s s o f d e m i s t e r pad, m m L o g a r i t h m i c m e a n t e m p e r a t u r e dif-
gr m o p r s t v
---------
Gravitational c o m p o n e n t M o t i v e steam in heat p u m p Dilute L i B r solution Demister Frictional c o m p o n e n t H e a t i n g steam T r a n s m i s s i o n line Vapor
ference, °C M a s s f l o w rate, kg/s Pressure, k P a P e r f o r m a n c e ratio, defined as M J Mm
y
--
Static head
at h i g h e r t e m p e r a t u r e s . A t these conditions, a drastic r e d u c t i o n o c c u r s in the specific f l o w rate o f c o o l i n g w a t e r and the specific heat transfer
A BPE C
---
Cp CR D
----
H H" L LMTD
-----
M P PR
----
- -
AP s sA
----
SMcw AT T U
-----
v
H e a t transfer area, m 2 B o i l i n g p o i n t e l e v a t i o n , °C M a s s fraction o f L i B r in solution
generated
in the
LiBr
R e f e r e n c e s
P r e s s u r e drop, kPa Salt c o n c e n t r a t i o n , mg/l Specific heat transfer surface area, m2/(kg/s) Specific f l o w rate o f c o o l i n g water T e m p e r a t u r e drop, °C T e m p e r a t u r e , °C Overall heat transfer coefficient, W/m2K
[1] H.M. Ettouney, H.T. El-Dessouky and I. Alatiqi, Chem. Eng. Prog., 95 (1999) 43. [2] G.F. Leitner, Int. Desalination Water Reuse Q., 7 (1998) 10. [3] J. Bednarski, M. Minamide and O.J. Morin, Proc., IDA World Congress on Desalination and Water Sciences, Madrid, 1 (1997) 227. [4] M.A.DarwishandH. E1-Dessouky, Applied Thermal Engineering, 18 (1996) 523.
--
V a p o r specific v o l u m e , m3/kg
V X
---
V a p o r velocity, m/s Salinity, p p m
Greek P )~
--
Density, kg/m 3 L a t e n t heat, k J / k g
[5] F. A1-Juwayhel, H.T. EI-Dessouky and H.M. Ettouney, Desalination, 114 (1997) 253. [6] H.T. E1-Dessouky and H.M. Ettouney, Simulation of combined multiple effect evaporation-vapor compression desalination processes, 1st IDA Int. Desalination Conference, Cairo, September, 1997. [7] H.T. EI-Dessouky, H.M. Ettouney, H. A1-Fulaij and F. Mandani, Chem. Eng. Proc., in press.
- -
174
F. Mandani et al. / Desalination 128 (2000) 161 176
[8] H.M. Ettouney, H.T. E1-Dessoukyand Y. AI-Roumi, Int. J. Energy Res., 23 (1999) 431. [9] J.M. Veza, Desalination, 101 (1995) I. [10] H.T. EI-Dessouky and H.M. Ettouney, Heat Transfer Eng., 20 (1999)52. [11] C. Temstet, G. Canton, J. Laborie and A. Durante, Desalination, 105 (1996) 109. [12] J. Weinberg and A. Ophir, Ashdod experience and other dual-purpose desalination plants based on multi-effect desalination with aluminum tubes, Symp., Desalination of Seawater with Nuclear Energy, Taejon, Republic of Korea, 1997. [13] Y. Michels, Desalination, 93 (1993) 111. [14] G. Kronenberg, Proc., IDA World Congress on Desalination and Water Science, Abu Dhabi, 3 (1995) 459. [15] H.T. E1-Dessouky, H.M. Ettouney and F. AIJuwayhel, Trans. I. Chem. E., in press, [16] J.L. Yhrelkeld, Thermal Environmental Engineering, 2nd ed., Prentice-Hall, Englewood, N J, USA, 1972. [17] J. Weinberg, A. Ophir and U. Fisher, Proc., 7th Int. Symposium on Fresh Water from the Sea, 1 (1980) 283. [18] G. Alefeld, and F. Ziegler, ASHRAE Tech. Data Bull., June 1985, pp. 11-24. [19] S.E. Aly, Desalination, 68 (1988) 57. [20] K. Fatbalah and S.E. Aly, Energy Convers., 31 (1991) 529. [21] S. Yanniotis and P.A. Pilavachi, Chem. Eng. Technol., 19 (1996) 448. [22] H.T. El-Dessouky, I. Alatiqi, S. BingulacandH.M. Ettouney, Chem. Eng. Yech., 21 (1998) 15. [23] O.A. Hamed, A.M. Zamamiri, S. Aly and N. Lior, Energy Convers. Mgmt, 37 (1996) 379. [24] I. Alatiqi, H.M. Ettouney and H.T. EI-Dessouky, Desalination, 126 (1999) 33. [25] H.T. E1-Dessouky, I.M. Alatiqi, H.M. Ettouney and N.S. AI-Deffeeri, Chem. Eng. Process., 39 (2000) 129. [26] H.M. Hellmann and G. Grossman, ASHRAE Trans Symp., 1996, pp. 980-997.
Appendix A: Model correlations The correlation for the demister pressure drop was developed by E1-Dessouky et ai. [25]. The correlation was conducted over the following parameter range, 1 _
(27)
where 9p is the demister pad density, Lp is the pad thickness, V is the vapor velocity in the demister pad. The BPE is obtained as a function of the brine salinity and temperature. The value of BPE is obtained from the following empirical correlation, which is valid for 20,000 _
(28)
with B =-(6.71 +6.34 x 10 2 T + 9.74 × 10-5 T2) 10-3
C =(22.238+9.59x10 3T+9.42× 10-5T2) 10 8
where the BPE is in °C. The seawater specific heat, Cp, is given by the following correlation: C = (A + B T + C T 2 +DT3)x l0 -3 P
(29)
The variables A, B, C and D are evaluated as a function of the water salinity as follows: A = 4206.8 - 6.6197S+1.2288 × 10-2S e
175
F. Mandani et al. / Desalination 128 (2000) 161-176
B = - 1 . 1 2 6 2 + 5.4178x10 -2S-2.2719x10 4S 2 C=1.2026x10 -2- 5.3566×10 4S+ 1.8906x10-6S2 D -- 6.8777 x 10 -7 + 1.517 x 10 -6 S- 4.4268 × 10 9 ~/2 where C o is in kJ/kg°C, Tin °C, and S is the water salinity in gm/kg. The above correlation is valid over salinity and temperature ranges of 20,000 < X _<160,000ppm and 20_< T_< 180°C, respectively, The correlation for the water vapor saturation pressure is given by
where v is in m3/kg and T is in °C. The above correlation is valid over a temperature range of 25-110°C. The percentage errors for the calculated vs. the steam table values are less than 10%. The latent heat correlation for the water vapor is )~ = 2589.583 + 0.9156 T-4.8343×10 2 T 2
(33)
where T is in °C and ~. is in kJ/kg. The above correlation is valid over a temperature range of 10-140 °C with errors less than 0.4% for the calculated and the steam table values.
P = 10.17246-0.6167302(7) + 1.832249 The density of seawater is given by x 10 -2 (Z) 2 - 1 . 7 7 3 7 6 x 10 -4 (T) 3
(30)
+ 1.47068× 10 6(7)4
9=103(A,Fl+A2F2+A3F3+A4F4)
(34)
The parameters in the above equation are given by where P is in kPa and T is in °C. The above correlation is developed over a temperature range of 10-110°C with percentage errors less than 2% for the calculated and the steam table values. The saturation temperature correlation is given by 3892.7 T=
42.6776- [ln(P/l~.48654])
) |
B = [(2)(s)/1000- 150]/150 G1 = 0.5 G~ = B G3 = 2B2 - 1
(31)
A~ = 4 . 0 3 2 2 1 9 G j + 0 . 1 1 5 3 1 3 G 2 + 3 . 2 6 x 1 0 4 G 3
-273.15 A2 = -0.108199 G1 + 1.571 x 10 -3 G2- 4.23 × 10 -4 G3 where P is in kPa and T is in °C. The above correlation is valid for the calculated saturation
A3 = - 0 . 0 1 2 2 4 7 G ~ + l . 7 4 x 1 0 -3 G2-9x10-6G3
temperature over a pressure range o f 10-1750 kPa.
A 4 = 6.92x10 -4 G~ -8.7×10 5 G2 _5.3×10-5 G3
The percentage errors for the calculated vs. the steam table values are less than 0.1%. The correlation for the saturation water volume
A -- [(2)(7) - 200]/160
is given by
F~ = 0.5, F 2 =A, F3 = 2 A 2 - 1 , F 4 = 4A3 - 3 A
v = 163.3453019 - 8.041421773 T +0.171021164 T 2 -.001.878124x10 3T3 (32) + 1"03842x10-5 T4 -2"28215x10-8 T5
where 9 is in kg/m 3, s is in gm/kg, and Tis in °C. This correlation is valid over 0_
1~ Mandani et al. / Desalination 128 (2000) 161-176
176
The v a p o r enthalpy o f pure water is given by
c~ = 40.2847, c2~ = 39.9142, c3~ = 33.3572
% = 13.1032, co2 = - 0 . 1 8 6 0 5 1 , c12 = - 0 . 1 9 1 1 9 8
H " = 2 5 0 0 . 1 5 2 + 1.947036 T - 1.945387 (35)
× 1 0 -3 T 2
c22 = 0.199213, c32 = - 0.178258, c42 = - 0.0775101 with a range o f 0.01 - 145°C and R z = 0.9999. The liquid enthalpy o f pure water is given
co3 = - 7.51277 E - 6 , c~3 = 0, c23 = 0, c33 = 0, c43
=
0
by H = 0.5802129 + 4.151904 T + 3 . 5 3 6 6 5 9 (36)
x 1 0 -4 T 2
with a range o f 0 . 0 1 - 1 4 5 °C and R 2 = 0.9999. H e l l m a n n and G r o s s m a n [26] developed the e n t h a l p y correlation for saturated LiBr solution, w h i c h is given by 4
~o lo T (P, C) -- ~ a~X ~+ T w ~ b~C ~ i~o i-o
(38)
2
H ( C , T) = X~_, a, T ~+(1 -C) ~ i-0
H e l l m a n n and G r o s s m a n [26] developed the boiling temperature o f the LiBr solution, which is given by
b~ T ~
i=o
In Eq. (38) Tw is the saturation temperature o f pure (3 7)
4 3 + X ( 1 - X ) ~ ~ cij(2C-1)~T j 0 ~-o
water at pressure P, C is the mass fraction o f L i B r in the solution, P is the pressure, and T is the boiling temperature o f the LiBr solution. The constants in the above relation are as follows:
w h e r e C is the m a s s fraction o f LiBr and T i s the solution temperature. The constants in the above relation are as follows:
a0 = 0, al = 16.634856, a 2 = - 5 5 3 . 3 8 1 6 9
ao = 508.668, a~ = 18.6241, a2 = 0.0985946,
a 6 = - 2 1 1 1 2 5 6 . 7 , a7 = 4,385,190.1
a3 = - 2 . 5 0 0 9 7 9 E - 5 , a4 = 4.15801 E - 8
a8 = -5,409,811.5, a 9 = 3,626,674.2
bl = 1.617155702, b 2 = 4 . 1 0 1 8 7 4 8 5 ,
alo = -1,015,305.9, bo = 1, bl = - 0 . 0 6 8 2 4 2 8 2 1
b3 = 0.000717667
b2 = 5.873619, b3 = - 102.78186, b4 = 930.32374
Coo -- - 1021.61, C~o = - 533.08, C2o = 483.628
b5 = - 4822.394,
C3o = 1155.13, C4o = 640.622, col = 36.8773
b8 = 34100.528,
a3 = 11228.338, a4 = - 110,283.9, a5 = 621,094.64,
b 6=
b9=
15,189.038,
b 7=
-
29,412.863
- 2 1 , 6 7 1 . 4 8 , blo = 5799.56