Carrier-Ga Enhanced Atmopheric Preure Dealination (Dewvaporation): Economic Analyi and Comparion to Revere Omoi and Thermal Evaporation Noah Abba and Kehinde Adeoye Chemical Engineering Undergraduate Univerity of Oklahoma 4/30/2007
Abtract With frehwater reource tretched thin, Jame Beckman of Ariona State Univerity developed carrier ga-enhanced atmopheric preure dealination, or imply dewvaporation, which preent a viable option to eaing water demand. Dewvaporation work by evaporating pure water out of eawater with dry air. Thi now humid air condene the pure vapor while donating it heat to eawater aiding in evaporation for the next cycle. Thi work to recycle heat and therefore give it an advantage over common thermal eparation. A mathematical model compoed of differential equation made poible a decription of the proce a well a an economic analyi. The reult of thi analyi predict a fixed annual cot of about $867 for a unit producing 00 gal/day. Thi correpond to a cot of about $2.9/000gallon. However, thi cot conider uing team to heat the air tream acro the top of the tower. Cheaper method may exit that utilie olar power or wate heat from an exiting plant.
Introduction Dealination i the proce of obtaining water from a olution of alt and water. Thi i an important proce for the future becaue of the high demand of pure water in the world today. The very nature of our exitence i dependent on the availability of water. Making up two-third of the earth urface, the ocean i the main ource water available. Unfortunately eawater we need i not found fit for human conumption and ha to be acquired one of everal technologie. The method developed today till do not remove alt from the water perfectly. The water till ha ome alt in it while the wate product (brine) ha ome water in it. The current method ued for dealination can be claified into two major categorie. Thee are membrane and thermal method. Baically, thermal method are thoe in which heat-driven evaporation i the primary ource of eparation while membrane method are thoe in which a emi-permeable membrane i the main piece of eparation equipment. There are two common example of thee method which are very much ued in different part of the world. Thee are evaporation and revere omoi. A more detailed explanation of thee two follow. World wide ue of thee method are equal at 0% 2 each. Thee method, becaue of their common ue, are good tandard for comparion with dewvaporation. Everyday, reearch and analyi are being carried out to create or evaluate new and exiting technologie for more economical reult. Mot of thee new method are till generally temming from already exiting technologie. For example, dewvaporation ha deep root in the current thermal method. Generally, mot new procee are geared toward improvement in all area of dealination.
Exiting Technologie Membrane method - Revere omoi Membrane method of dealination rely on the paing of alt water feed through a emi permeable membrane in order to eparate alt from water. The mot common membrane method i Revere Omoi. Revere omoi i a method of dealination where the feed olution i forced through a emi-permeable membrane reulting in the paage of high water content olution and leaving behind a highly concentrated alt olution. Phyically, olution flow from a high to low concentration region (omoi), but the application of preure yield the oppoite and water flow from the olution through the membrane (revere omoi). The applied preure ha to be greater than the normal omotic preure of the alt water. Omotic preure i a the preure a olution exert on a membrane in an encloed pace due to the difference in olute concentration between both ide of the membrane. A mentioned earlier, thi proce doe not yield totally pure water, the water and alt travel at different rate through the membrane. A chematic of thi proce i hown below. Membrane Module Pure Saline Feed Pretreatment Pot Water Treatment Pump Brine Fig. : A typical revere omoi plant
Uually, the proce conit of four major part. The firt i the pretreatment tage. Thi i neceary for mot membrane method becaue of the ettling of microorganim and different particle in the feed water olution on the membrane cauing fouling. Variou method of pretreatment include addition of chemical uch a chlorine to kill thee microorganim or ultra-filtration to eparate them out. Pretreatment alo involve the addition of anti-calant to prevent caling on the membrane urface. Scaling i the forming of alt precipitate uch a calcium carbonate on the urface of the membrane due to the concentration of alt ion at the membrane urface reulting in aturation and then precipitation. Thi reduce the ability of the membrane to eparate the alt and water. Example of anti-calant include ulfuric acid. The econd i the pumping tage where the aline feed preure i raied to the calculated optimum for that plant operation. It mut be larger than the omotic preure of the aline feed. The third i the actual eparation proce in the membrane module. During eparation, not only one membrane i preent, there are everal layer of membrane and everal unit of thee membrane giving them the term module. A tream of pure water and another of brine leave thi ection. The lat tage i the pot treatment tage. Pot treatment in revere omoi dealination involve the ue of alt uch a calcium and odium to alleviate the effect of the acid added in the pretreatment tage. Calculated cot aociated with revere omoi include the following: Table : Revere Omoi Cot,6 Dealination Expected Operating cot Capital cot Energy requirement method recovery (%) ($/000gal of product) ($/gal-day) (KWh/000gal of product) Revere omoi 0 2.0 4.00 4.00-0.00 26
The major ource of expenditure i the membrane ued. Thi i a complex web of polymer deigned to allow the paage of water while preventing the paage of alt to an extent. Variou deign have being uggeted to increae the electivity of thee membrane but thee deign increae in price a they get better. Alo becaue of fouling and caling, membrane deign are being evaluated for their performance level alo increaing the cot of a membrane. A membrane lifetime i a maximum of two year and uually need changing after that. Thermal method Multi Stage Flah Dealination Thee are procee involving the heating of alt water to evaporate the pure water and condene it. Thi baic proce ha being in exitence for centurie and i the firt method of dealination to be ued in earlier form uch a boiling. Over the year, modification of thi dealination proce have being in effect and currently the mot ued type i the Multi-Stage Flah Ditillation (MSF). Currently, 80% 7 of the world thermal dealination product are from MSF dealination. In thi proce, the alt water i firt heated at high preure and then tranferred into a chamber at a reduced preure. Due to the drop in preure at the beginning of each chamber, the boiling temperature of water reduce cauing it to evaporate. Thi proce of quick evaporation i uually referred to a flahing. Condenation of the vapor take place on the tube of a heat exchanger that pae through the chamber. The heat exchanger ue the feed water to cool the vapor before the feed enter the chamber. The heat of vaporiation during condenation i then tranferred back to the feed tream and the cycle continue. The flah chamber uually are between and 2 2. The heat
releaed through vaporiation make thi proce able to ue low heat ource to heat up the feed at the tart of the proce. A chematic of the proce can be een below. Condenate Tray Heat Exchanger Saline Feed Heat Vapor Pure Water Pump Wate Brine Fig 2: Schematic of Multi Stage Flah Dealination Advantage of thi proce include the low cot of energy with availability of wate heat and the inignificance of the quality of the aline feed reducing pretreatment cot. Problem aociated with the MSF method are high operating cot with no wate heat available, corroion and eroion. Corroion and eroion are caued by the fact that water i being paed through metal pipe, and alo becaue of the peed at which the water move. The ue of metal in thi proce i inevitable a high conductivity of heat i epecially needed for high yield. The ue of tainle teel i very common to prevent corroion but thi i very cot intenive. MSF plant are uually found in the Middle Eat becaue of the large availability of cheap energy. The major money guler for thi proce i the heat required to heat the alt
water. Thi i very much the main drive of the proce and directly impact the production capacity. The following are ome tatitic about the MSF proce.. Expected recovery 20-30% 2. Aociated cot $2-$4/000gal of product 4 3. Energy requirement 6KWh/000gal of product A general decription of the proportion in which thee method are being ued in the world compared to other membrane method or thermal method can be een in the following chart. Ditribution of dealination proce in the world Other (0% membrane) MSF (84% 6 thermal) RO (90% 3 membrane) Other (6% thermal) Fig.3: Ditribution of dealination procee ued in the world For thee two major procee, the lifepan of a typical plant hould be well looked at. Even though thermal procee may eem to be more expenive than membrane procee, their lifetime are more than twice the lifetime of a membrane.
Wate Dipoal Technique Wate dipoal method in dealination are very particular to the ytem being ued and the urrounding in which they are ued. In thi paper, only the ytem ued will be ued to evaluate different dipoal method a no particular plant i in conideration. Wate from dealination plant can generally be divided into three part. Thee are the pre-treatment, actual proce and pot-treatment wate. The pretreatment method dicued above for the revere omoi plant i very imilar to thoe of the MSF method except for the concern of microorganim and upended particle fouling any part of the ytem. The ultra filtration of thee require correct dipoal. Mot pretreatment wate are olid although anti-calant and chemical addition can contitute liquid wate. The actual proce wate varie depending on whether it i a thermal or membrane proce. For thermal procee, the wate include concentrated brine and dried alt. Some method of dipoal are returning the brine to it ource, brine evaporation and deep-well injection. Some gae may appear, example of which are carbon dioxide, oxygen and nitrogen. Thee require no pecial dipoal method. Membrane procee on the other hand produce le concentrated brine wate but have membrane module to dipoe off a olid wate. The brine wate can be dipoed in a imilar manner a the thermal method. The membrane module have no health haard or rik therefore they are placed in and fill. Pot treatment wate are imilar to thoe obtained from pre treatment wate and cane be dipoed of in the ame way. Dewvaporation Thi i a proce invented recently by Jame Beckmann, an aociate profeor of
chemical engineering at Ariona State Univerity. It ue novel idea to implement alt water dealination in the direction of thermal method. The whole proce i carried out in atmopheric preure and it ue air a a carrier ga for the water vapor. The heat ource can come from wate heat from an exiting plant poibly bringing the heat cot to ero. Thi proce promie lower cot for altwater dealination through thi ability to cut down on energy cot. The proce flow diagram can be eeing below in Fig. 4 and a decription given.
Added heat (Q boiler ) P u r e w a t e r S a l i n e f e e d Ambient Air Outlet Air Inlet Air Blower Air Evaporating water Heat Condening water Saline feed Fig. 4: The flow diagram of a typical dewvaporation proce
Proce decription In thi proce, ambient air of known temperature and humidity i pumped at a contant rate into the bottom of the evaporation chamber of the dewvaporation tower. A air flow up, imultaneouly, aline feed i flowing down from the top of the tower on the ame ide along the heat tranfer wall. Heat tranfer occur between the wall and the aline feed and water evaporate from the olution into the ambient air erving a a carrier ga. Concentrated alt water exit the tower from the bottom of the evaporation ide. The air, which ha now increaed in temperature and humidity flow from the top of the tower. An external heat ource i ued to raie the temperature of the air before the air flow down the dewformation ide of the tower. Uing team a a heat ource increae the humidity of the air that goe into the dewvaporation ide allowing more water to condene. The increaed temperature of the air from heating allow for thi ide to be lightly hotter than the evaporation ide initiating condenation and tranfer of condenate heat to the evaporation ide through the wall. Both Pure water and air at reduced temperature flow from the dewformation ide of the tower. The air can be recycled into the incoming air tream and pumped back into the ytem. Economic analyi The main objective in developing new technologie for dealination i to reduce cot. For mot dealination proce, the two mot important cot component calculated are the operating and capital cot. The operating cot i the cot of running the proce daily. It i uually expreed a cot per unit energy or cot per unit of production. The capital cot i that which come from the equipment ued in production. The capital cot etimation uually come from the calculation of ome property of the equipment uch a
area or capacity while the operating cot come from a parameter that directly impact the production uch a heat or preure. In thi cae the heat added to the ytem i ued. The addition of thee cot give the total annual cot of the proce. The firt tep in thee cot etimation wa the derivation of ma and heat balance. Tower Mathematical Model The firt tep in deigning the tower i generating the neceary operating parameter uch a temperature and flow rate etting. In the abence of a pilot plant, a mathematical model give the bet etimate of thee parameter and alo make ome optimiation poible. For the dewformation tower, differential equation are adequate to decribe the heat and ma tranfer. Building a model of differential equation for the tower require defining a differential portion of the tower. The following diagram diplay the boundarie of the region: 2 3 4 T T + T T Air V d G d dw d T 2 q dw e T Air V e +d T 2 + T + G FD FB Fig. : Differential Set Up
Thi tower i divided region one through five. The firt region i the humid air containing the pure water vapor. A it dump off heat to region two, it water vapor alo condene into thi region (dw d ). Region two i therefore the pure water product (FD). The heat of condenation of the water on the wall then move over to the fourth region. The fourth region i the eawater feed (FB) running along the wall on the evaporation ide. It receive the heat and i expoed to the dry air of region five. Thi force the evaporation of water (dw e ) into region five. V refer to the ga loading in unit of mole of water vapor/ mole of air, while V refer to the aturation ga loading. G i the flow rate of air minu the water vapor it contain and i contant throughout the apparatu. Integrating the ma and heat balance on thi diagram decribe the ytem from top to bottom. The ma and heat balance tarted with the following analyi: Region Ma balance Heat balance GV GV + + dw Gh T ) + GVh ( T ) Gh ( T + ) + GVh ( T + ) + h dw + hl( T T ) d d d 2 d d a ( v a V vap d 2 FD + dw FD + FD hw ( T2 ) + hvdwd + hl( T T2 ) d FD+ dhw ( T2 2 ) + q 3 0 q UL( T2 T4 ) d FB FB + + dw FB h T ) q FB h ( T ) + h dw + hl( T ) 4 d e d e w ( 4 + + d w 4 v e 4 T GV GV + + dw Gh T + ) + GVh ( T + ) + h dw + hl( T T ) d Gh ( T ) GVh ( ) a ( V vap e 4 a + v T Table 2. Ma and Heat Balance For example, the ma balance on the firt region ha term accounting for the water vapor input and output to the region along with the mall amount of water dumped off into the pure water product flow. The ame quantity of air (G) goe into and out of the region uch that the term cancel. The heat balance ha term accounting for the enthalpy of air (Gh a (T )) and water vapor (GVh v (T )) that enter the region. Exiting the region i the new enthalpy of the air and water vapor a well a the heat leaving with the
condened vapor ( h vap dw d ) and alo the heat given to region two due to the temperature difference (hl(t -T 2 )d). The next tep i to convert thee equation into differential that can be olved in a tep wie fahion. The derivation ran a follow: Region Ma balance GV GV + dw Z Z + D dw d GV GV+ d dw d G dv d Heat balance Gha ( T ) + GVhv ( T ) Gha ( T ) + GVhV ( T ) + hvapdwd + hl( T T2 ) d Gh T ) Gh ( T ) + GVh ( T ) GVh ( T ) h dw + hl( T T ) d a ( a v V vap d 2 GCpair + GV Cpv hvapdwd + hl( T T2 ) d Subtituting the ma balance, dv GCp + GV Cpv hvapg + hl T T air ( 2) GCp air hl( T + GV T ) d Cp v 2 h vap G dv Note: V < V then h h, dv 0 and T T 3 d Region 2 Ma balance FD + dwd FD+ Note: ma balance 0 until d V > V Heat balance
FD hw T ) + hvdwd + hl( T T ) d FD+ dhw ( T ) + q ( 2 2 2 2 Rearranging and ubtituting with the ma balance FD h T ) ( FD + dw ) h ( T ) + h dw + hl( T T ) d FD FD w( 2 d w 2 2 v d 2 q hw ( T2 ) FDhw ( T2 2 ) dwd hw ( T2 2 ) + hvdwd + hl( T T2 ) d q UL( T2 T4 Cpw2 dwd ( hw ( T2 2 ) + hv ) + hl( T T2 ) d UL( T2 T4 ) d Region 4 2 dw h ( T ) + h ( d w 2 2 vap 4 2 FD Cp w ) UL( T2 T + FDCp w ) hl( T T FD Cp w ) d ) d Ma balance FB FB + dw + d e Heat balance FB h T ) + q FB + h ( T ) + h dw + q w ( 3 in d w 3 Subtituting ma balance FB h T ) + q FB + h ( T ) + h dw + q w ( 3 in d w 3 v e out w ( T3 ) ( FB dwe ) hw ( T + 3 ) + qin hvdwe qout w3 + dwehw ( T + 3 ) hvdwe + qin qout w3 + dwehw ( T + 3 ) hvdwe + h24l( T2 T4 ) d h4l( T4 T h4l( T4 T ) d h24 L( T2 T4 ) d + hvdwe dwehw ( T + 4 ) FB h + FB Cp FB Cp 4 FB Cp w v e out ) d Region Ma balance GV GV + dw dw + d e G dv e Heat balance Gha ( T ) + GVhV ( T ) + hvapdwe + hl( T4 T ) d Gha ( T ) + GVhv ( T ) Gh T ) Gh ( T ) + GVh ( T ) GVh ( T ) + h dw hl( T T ) d a ( a v V vap e 4 GCpair + GV Cpair + hvapdwe hl( T4 T ) d
dv GCpair + GV Cpair + hvapg hl( T4 T ) d GCp air hl( T + GV 4 Cp T ) d v + h vap G dv The reult are ummaried in the following table: Table 3. Reult of Ma and Heat Balance Region Ma Balance Heat Balance dw d G dv GCp air hl( T T ) d + GV Cp h v 2 vap dv G 2 FD + dwd FD+ d 4 FB FB+ d + dwe dwd hw ( T2 2 ) + h 2 FDCp 4 ( vap 2 h 4 L w ) ULT ( 2 T4 ) hlt ( T ) + d FDCp w FDCp w ( T4 T ) d h24l( T2 T4 ) d + hvdwe dwehw ( T + 4 FB Cp w ) dw e dv G GCp air hl( T + GV 4 Cp T ) d v + h vap G dv In certain cae thee equation make take different form. For example, the reult of the heat balance on region one only take the form hown in Table 3 when V d exceed V d. Thi indicate that the air i currently holding more water than i poible given the aturation ga loading. It mut therefore dump off water to region two, reulting in bulk heat tranfer. When it i lower, the lat term in the denominator i ero becaue no water moved acro the boundary of region one and two in that differential tep. Thi may be the cae when hot humid air firt enter the top of the tower on the
dewformation ide and ha yet to experience ignificant cooling. However, when the heating i done via adding team, the air tream remain aturated even though it i at a higher temperature. Thi i the cae in the model preented here. The ame ituation arie in region five regarding the lat term in it denominator and V e relationhip to V e. The heat and ma balance on region three are trivial and not diplayed. The only affect region three ha on the heat tranfer i to alter the heat tranfer coefficient between region one and two. When V d exceed V d, no water product ha formed and therefore region two doe not exit. The reulting heat tranfer coefficient i the reult of air forced convection with metal rather than air forced convection on water. Solving the Model Microoft Excel put the model to the tet. The preadheet take the equation and integrate them from the top of the tower to the bottom. Starting at the top of the tower, the uer ha everal variable that he may manipulate a etting for column operation. Thee include Q,, Z, G, FB, L, T top, and T4 top. Q i the heat added to the air tream a it croe over the top of the column from the evaporation ide to the dewformation ide. Becaue the heat goe in a team, the air i till aturated a it enter the other ide even though it temperature i lightly higher. The amount of water it hold alo increae becaue of the water added a team. i the incremental height of the column and therefore refer to the tep ie of each calculation a the integration move down the tower. It ummation give the height of the tower. G i the air flow up the evaporation ide and down the dewformation ide and i contant throughout the tower. FB i the feed flow rate of the eawater/brine, L i the
width of the tower, T if the temperature of the air a it exit the top of the evaporation ide, and T4 i the temperature of the feed. T4 i not very flexible ince the uer i retricted to uing ocean temperature near hi plant. The reult of thi integration include the heat required a well a the neceary heat tranfer area of the column. Thee are the two main cot parameter in deigning the column. To tart the integration the preadheet had the following heading: Table 4. Spreadheet Heading and Layout* *Note: Jame Beckman number for the aturation humidity did not agree with thoe found in Perry Chemical Engineer Handbook. The reult reported here are uing aturation calculation from Perry. Otherwie the preadheet give no ignificant product flow rate. The preadheet keep track of all relevant parameter and variable needed in calculating them a it integrate down the column. It ha a column for the temperature change, ma exchange, and humidity baed on temperature down the column. Reult The firt reult to conider from the preadheet i the temperature profile. When the temperature follow expected trend and match exiting data, the model gain credibility. Figure 6 diplay the temperature profile down the column:
Temperature Down the Tower 80.00 70.00 Temperature (deg. C) 60.00 0.00 40.00 30.00 20.00 T Air Dewformation Side T Pure Water Product T Seawater/Brine T Air Evaporation Side 0.00 0.00 0 00 200 300 400 00 600 700 800 Ditance from Tower Top (cm) Fig. 6: Temperature profile down the tower. Thi graph conciely diplay the trend in temperature. The air on the dewformation ide decreae in temperature a it give off heat to the ret of the column a expected. The pure water product alo looe more heat than it gain, o it temperature alo decline. The brine warm lightly a it evaporate. Mot importantly, the air on the evaporation ide ee it temperature fall to roughly that of ambient air. Thi i critical becaue the entry temperature of the air on the evaporation ide i not adjutable by the uer. Lowering thi temperature to ambient air temperature i what etablihe the end of the column. Thi figure ugget that the column mut be about 7.2m high, which i number that may change a operating variable change. Once the temperature profile lend validity to the model and the current etting, the calculation preent other relevant production number. Thee include the flow rate of product and brine that mut be dipoed a well a the air flow throughout the column.
Air flow eemed to have the mot dramatic influence on the production flow rate, column height, and heat required to run the tower. The reult of varying air flow are diplayed in the following table and figure: Table : Tower Reult Deign G mol/h Qboiler J/hour FD gal/day FB gal/day FAC $ Cot $/000gallon 000 290700 87.8 94.77 $,7.4 $2.9 2 2000 8400 272.06 870.6 $,66.9 $.92 3 3000 87200 62.0 80.6 $,72.26 $0.96 4 4000 62800 749.33 393.29 $,773.3 $0.74 000 4300 936.60 206.0 $,830.24 $0.6 6 6000 744200 23.76 8.8 $,867.0 $0.2 Product Flow v. Air Flow 200 000 Product Flow gal/day 800 600 400 200 0 0 000 2000 3000 4000 000 6000 7000 Air Flow Mol/hr Fig. 7: Product Flow v. Air Flow
FAC v. Air Flow 2000.00 900.00 800.00 700.00 600.00 FAC $ 00.00 400.00 300.00 200.00 00.00 000.00 0 000 2000 3000 4000 000 6000 7000 Air Flow mol/hr Fig. 8: FAC v. Air Flow An intereting obervation i that the model predict very little change in flow rate (Le than a half a gallon/day for each C) when altering the temperature increae of the air acro the top of the tower. Thi i probably becaue under additional heating the humid air on the dewformation ide mut experience more cooling to drop off the ame amount of water. However, the air flow reult in large change becaue expoure to more air force more drying. Upping the air flow alo increae flow rate of product. Becaue the flow rate of the product i higher, the brine flow rate i lower thu leading to aving on brine dipoal. Therefore, the benefit of increaing the flow rate, that i, a higher product flow and le brine dipoal, far outpace the downide, which include higher energy cot and larger proce equipment.
Another important factor in deign conideration i the concentration of the brine at the tower exit. If thi concentration hould fall below the olubility of alt in water (374 g/l), crytal will form on the inide of the tower. Thi i obviouly undeirable becaue it can caue blockage and therefore the flow may back up. The following table conider thi important apect of each deign. Table 6. Deign Feaiblity Deign Seawater Flow L/hr Total Salt g/hr Brine Flow L/hr Brine Concentration g/l 80 406 3.9 2 80 406 37 39.4 3 80 406 92 9.0 4 80 406 62 87.2 80 406 32 66.4 6 80 406 3 88.2 The finding in Table 6 unfortunately indicate that deign ix i not a poibility becaue the concentration exceed the olubility limit of alt in water. The model predict a price per 000 gallon of about $2.9 when conidering heat requirement and brine dipoal. Thi fall within the range of $3.70 to $.70 predicted by Jame Beckman pilot plant 0 and i competitive with exiting method uch a revere omoi and evaporation. Dicuion and Concluion Dewvaporation i one of everal olution to growing water demand. Further reearch into the implementation of olar energy to make the proce even cheaper can make the proce more appealing to communitie looking to add to their current water production. According to thi mathematical model, air flow i the mot important production parameter when conidering a dewvaporation plant. Definition of Notation and Variable
dw d dw e FB FD G T V V Q Z Differential Amount of Water Added to FD Differential Amount of Water Evaporated Seawater Flow Pure Water Air Flow Temperature Ga Loading (mole water vapor/ mole of air) Ga Loading at aturation humidity Heat Column Height Reference. Miller, Jame E. Review of Water Reource and Dealination Technology. March 2003. Albuquerque, NM. http://www.andia.gov/water/doc/millersand2003_0800.pdf 2. Krihna, Hari J. Introduction to Dealination Technology. http://www.twdb.tate.tx.u/dealination/the%20future%20of%20dealination %20in%20Texa%20-%20Volume%202/document/C.pdf 3. Beckman, Jame. Carrier Ga-Enhanced Atmopheric Dealination: Final Report. Ariona State Univerity. October 2002. 4. Chaudhry, Shauhid. Unit Cot of Dealination. http://www.owue.water.ca.gov/recycle/deal/doc/unitcotofdealination.doc. R. W. Baker et al. Membrane Separation Sytem. 99, Noye Data Corporation. http://www.knovel.com/knovel2/show_text.jp?setid90749&spaceid0&v erticalid0&bookid32&nodeid8827923&searchtype0&searchmode true&htmlfale&textid2&random472436 6. US Army Corp of Engineer. Wate Dipoal. http://www.uace.army.mil/publication/armytm/tm-83-8/c-0.pdf (wate dipoal) 7. http://www.hwdealination.com/multi%20tage%20flah%20dealination.html 8. Frenkel, Val. Dealination Method, Technology, and Economic. Kennedy and Jenk Conultant http://www.idwater.com/common/paper/paper_90/dealination%20method,% 20Technology,%20and%20Economic.htm (diagram)
9. Beckman, Jame and Baem.M. Hamieh. Seawater Dealination Uing Dewvaporation Technique: Experimental and Enhancement Work with Economic Analyi. Ariona State Univerity. September 2003. http://www.deline.com/articoli/6604.pdf (Dewvaporation) 0. Dealination, Multitage Flah. http://www.uwec.edu/piercech/dealination/msf.htm. Peter, Max S. and Klau Timmerhau. Plant Deign and Economic for Chemical Engineer. McGraw-Hill 2003. 2. Winnick, Jack. Chemical Engineering Thermodynamic. John Wiley & Son. 997. 3. Welty, Jame R. et al. Fundamental of Momentum, Heat, and Ma Tranfer. 4 th edition. John Wiley & Son 200. 4. Perry, R. H. Editor. Perry Chemical Engineer Handbook. 7 th Edition. McGraw-Hill 997.. EPA. Drinking Water Adviory: Conumer Acceptability Advice and Health Effect Analyi on Sodium. February 2003. 6. Patnaik, Pradyot. Handbook of Inorganic Chemical. McGraw-Hill. November