- Email: [email protected]
Optimum design of water-utilize systems featuring regeneration re-use for multiple contaminants
1082
Process SystemsEngineering2003 B. Chen and A.W.Westerberg(editors) 9 2003 Publishedby ElsevierScienceB.V.
Optimum design of water-utilize systems featuring regeneration re-use for multiple contaminants Dong-mei Xu a , Yang-dong Hu a'*, Ben Hua b, Xiu-lin Wang a
College of Chemistry and Chemical Engineering, Ocean University of China, Qingdao, 266003, P. R. China. a
blnstitute of Chemical Engineering, South China University of Technology, Guangzhou 510640,P. R.China Abstract The optimum design of the water-utilize systems featuring regeneration re-use for multiple contaminants is addressed. A global automated optimization design method targeting minimum fresh water consumption, which is based on sequential operation model, is presented, and the mathematical formulation is a MINLP, which can be solved by the computer code based on the hybrid genetic algorithm. This computer code consists of the outer and inner loops: the outer loop is used to optimize the operation sequence based on the adaptive values fed back from the inner loop which is used to optimize the design for a population of given operation sequences. One example is solved to demonstrate the advantage and effectiveness of the new method. Keywords design, wastewater minimization, water utilize system, non-linear programming, regeneration reuse, multiple contaminants
I. INTRODUCTION Water pollution and scarcity of water are a life-and-death problem facing with human being, and stricter discharge regulations, has caused the price of freshwater and wastewater treatment facilities to rise. Therefore, for process industries, it is worthy of investigation evolving water utilize networks to reduce both fresh water consumption and wastewater production. Up to now, several related studies for multiple contaminants problem have been published. Wang and Smith tll addressed a targeting and network design procedure based on elaborate special shifting of streams on the concentration-load diagram for multiple contaminants. Liu t21 presented a few interesting heuristic rules. Doyle and Smith TM started solving the multi-component problem using an iterative procedure and then they proposed to solve the NLP multi-component problem based on the forgoing solution. Alva-Argaez et al. t4'51 presented a solution approach for multiple contaminant systems in combination with water "Corresponding author: Yang-DongHu Tel:+86-532-2032141 E-mail: [email protected] College of Chemistryand Chemical Engineering, Ocean Universityof China, Qingdao, 266003, P. R. China.
1083 treatment in which they include piping costs as well as treatment costs, later E61they discussed some trans-shipment models based on several assumptions, some of them restrictive. Benko et al. [7] modeled the example problem proposed by Takama et al. t81 as a non-convex MINLP problem. Bagajewicz et al. t91 used a combination of mathematical programming and necessary conditions of optimality to approach the design of water utilization with multiple contaminants, later, Bagajewicz tl~ (2000) presented a review of the procedures to design and retrofit water networks in which multi-component systems are also addressed. Regeneration re-use is a usual and significant approach to reduce the amount of fresh water that the overall water utilize system needs, however the design of a water utilize system featuring regeneration re-use for multiple contaminants is a complex problem that involves lots of decision variables, which has not been solved satisfactorily, in this work we approach the above problem and present a simultaneously targeting-design, global automated optimization method that targets minimum fresh water consumption for the overall process. 2. A GLOBAL AUTOMATED OPTIMIZATION DESIGN M E T H O D Given are: a set of water-using operations to be considered in the water system, each one is described by its limiting operation data: the maximum inlet, outlet concentration of water profile of each contaminant and mass load of contaminant to be transferred. The inlet and outlet concentration limits account for corrosion, fouling and maximum solubility, etc Eli. The objective is to design a minimum fresh water consumption water network when the intermediate regeneration is allowed. The following assumption will be utilized: (1) It is allowed wastewater being blended with wastewater from other operations and fresh water; (2) All the operations are physical changes without chemical reaction. 2 possibilities to any operation are considered: possibility 1: it is forbidden to split the operation; possibility 2: it is allowed to split the operation by the fraction of mass load. In this paper, the mass load fraction of the operation i in the process before regeneration of a; is introduced, which can be used to distribute all the operations to the process before regeneration and aider regeneration conveniently. And for the performance of the regeneration processes tll, the "removal ratio" specification is adopted. As a consequence of presence of multiple contaminants, sometimes it is difficult to ascertain the optimal operation sequence, which is a variable needs to be optimized and will be discussed later on. First we address the sequential operation model under the condition that the operation sequence is given.
2.1 Sequential operation model constitution for a given operation sequence Step 1: Assume that all the operations in the water utilize system are distributed to the process before regeneration and the process atter regeneration by their fractions a, : for the operations of which the maximum inlet concentration of a certain contaminant equals the value of 0, the aggregation of them is denoted by Ob (these operations can't match with the regenerated water), their fraction a, take the value of 1, i.e. they are assigned to the process before regeneration entirely; for the operations forbidden to be split, the aggregation of them is denoted by Oforbidden, their a, take the value of 0 or 1,but take the value in the region[0,1 ]
1084 for the operations allowed to be split, the aggregation of them is denoted by Oallowed. Step 2: Assume all the outlet water streams from the former operations are the water sources of the latter operation i on condition that the operation sequence is given, i.e. all the precursors of operation i send wastewater to operation i. Then number all the available water sources in the process before regeneration and the process after regeneration: in the process before regeneration, prescribe that outlet wastewater of thejth operation be regarded as thejth water source, and when j takes the value of 0, the fresh water is specified. In the process after regeneration, prescribe that the regenerated water be regarded as the 1st water source except for the fresh water, i.e., j takes 1. Accordingly in turn the outlet wastewater of the lth operation be regarded as the 2 nd water source, the outlet wastewater of the jth operation be regarded as the (j+l)th water source, and w h e n j takes the value of 0, the fresh water is also specified. Step 3: Match each operation with all its available water sources in the process before regeneration and the process after regeneration respectively. Based on the above steps, the sequential operation model of water system with regeneration reuse for multiple contaminants is constituted. It is assumed that all the outlet wastewater streams from the process before regeneration are mixed to form one stream before it is regenerated.
2.2 Mathematical formulation of the design of the multiple contaminant system For a water-utilize system with m contaminants composed of n operations, the design problem can be formulated as a MINLP1 on the basis of the sequential operation model constituted above that is shown as follows. min(min f ws,t) l ~ S
(MINLP1)
(S: aggregation of all possible operation sequences of P~ in number,)
(1)
subject to: a) The process before regeneration
"~ aimi, ,
= (C,,, - C.,,k )fwsl
(k--1,2 ...... m)
(2)
i=!
t,,o,i = fw,,
(3)
i=1
Z C'.j.kL'a.i/j~L'.J.i < Ci,,(max),i,k
(j:O,1.2 ...... i-l, k : l , 2 ...... m)
(4-i)
jewi
C~,i,k < Cou,,(m=)~,k Z Ll,j,i(Cl,i,k -Cl,J,k) = aimi,*
(k=l,2 ...... m) 0=0,1.2 ...... i-l, k=l,2 ...... m)
(5-i) (6-i)
jew~ i-I
L,.j, i < L,,j - E Ll,j.g (J E Wl, Wl-{j [ the water sources of operation i, j=0,1 ...... i-1 } g=j +l
(7-i)
1085 For the constraints (4-i) - (5-i) - (6-i) and (7-i), let i=1,2 ...... n, then the successive constraints for the n operations in the process before regeneration are listed. b) the process after regeneration L <~fwsl
(8)
= f w,,t •;'•L2,o,,
- f wsl
i=!
•Q L2,u = fr
(9)
(10)
i=1
Similar to the constraints (4-i) - (5-i) - (6-i) and (7-i), the constraints ( 1l-i) - (12-i) (13-i) and (14-i) in the process after regeneration can be established by substituting the variables Lj,j,i, CI,j, k for the vafiablesL2,j.i, C2.j,k, and the difference from the process before regeneration is, that the regenerated water as a water source should be considered. Here we also let i= 1,2 ...... n. Besides, the fractions of all the operations a, are subject to the constraints below. a,=l
(i~Ob)
(15)
a, (1- a,) = 1
(i E Oforbidden)
(16)
(i E Oallowed)
(17)
O< oti
Objective 1 is to seek the minimum fresh water flowrate that can be used to solved the problem, which is a nested objective composed of the inner layer and outer layer: the inner layer is for a given operation sequence and the outer layer is for all the possible operation sequences based on the results from inner layer, i.e. the optimization of the overall water utilize system. Equation 2 is the total mass transfer balance for each contaminant in the process before regeneration. Equation 3, 9 and 10 are the total fresh and regenerated water balance in the process before regeneration and the process after regeneration. Inequation 8 represents the constraint to the total regenerated water. Inequation 4-i and 5-i represent the constraints of inlet and outlet condition of each contaminant to operation i. Eequation 6-i is the mass transfer balance for each contaminant of operation i. Inequation 7-i represents the constraints of the flowrate of each water source matched with operation i. The above MINLP 1 is solved by the computer code programmed by our group based on the hybrid genetic algorithm, which takes the crossover of PMX and reversal mutation for the operation sequence code. This computer code consists of the outer and inner loops: the outer loop is used to optimize the operation sequence based on the adaptive values fed back from the inner loop and the inner loop is used to optimize the design by solving the MINLP to come at the adaptive value of each given operation sequence for a population of operation
1086 sequences. In the solving process, the infeasible matches of the water sources and operations can be automated eliminated, then the values of all the variables for the design of water network with regeneration reuse are yielded. However the regenerated water flowrate obtained fr by solving MINLP1 is probably an arbitrary value meeting the constraints instead of the minimum compatible with minimum fresh water flowrate. To solve the problem, We take minimum fresh water flowrate fws,t and the operation sequence l attained above as known quantity, and substitute objective function min(minfws,t) in MINLP1 for objective function min fr to establish MINLP2 under the same constraints, which can be easily solved to determine minimum regenerated water flowrate f,. compatible with minimum flesh water flowrate and other variables again, accordingly the flowsheet of the overall process is acquired. 3. EXAMPLE Consider an example composed of 6 operations, whose limiting operation data is shown in table 1. Here let us suppose an removal ratio R of 0.9 and determine the regeneration reuse network targeting minimum fresh water flowrate by the method presented in this paper in such a case: "it is allowed splitting all the operations by their fraction of mass load", which are among all the possible cases of example 1. The flowsheet is shown in Fig. 1. And the minimum fresh water and the corresponding minimum regenerated water flowrate are 27.960 t/h and 24.746 t/h respectively. 4. CONCLUSION This paper presented a simultaneous targeting-design, automated optimum method based on sequential operation model for the design of multiple-contaminant water-utilize systems featuring regeneration reuse that minimizes fresh water consumption. Two possibilities to any operation and the optimization of operation sequence are considered, therefore it can be used as a general-purpose method for designing water networks. It has the advantages that for the regeneration reuse possibility the optimum design is able to be achieved without knowing the fresh and corresponding regenerated water requirements in advance and the infeasible matches of the water sources and operations can be automated eliminated. Table 1. Limiting operation data No. ContaCin,max minants (ppm) Salts 300 5000 H2S Ammonia 1500 Salts 10 0 H2S Ammonia 0 Salts 10 0 H2S Ammonia 0
Cout,max Load (ppm) 500 11000 3000 200 500 1000 1000 2000 3500
(kg/h) 0.18 0.75 0.1 3.61 0.25 0.8 6 15 10
No. 4 5 6
Contaminants Salts
Cin,max Cout,max Load (ppm) (k#h)
(ppm) 100 50 H2S Ammonia 1000 100 Salts 50 H2S Ammonia 1000 85 Salts 300 H2S Ammonia 200
400 2000 3500 350 1800 3500 350 6500 1000
2 0.8 1
3 1.9 2.1 3.8 1.1 2
1087 fresh water
a 200.00
~18.05t/h ~~176 5 >l[240787s I . . . . I . I a 800.00 [ "-I I l" 7.5t/h "] operation 3 [ b 2000.00> / | c 1333.33 regeneratec~water L.~l.529t/h a 350.00 h2.164t/h a 350.00 1 " ~ I .. .-Ib 251 71 ~ I .. .. Ib 152 49,_ 9 " -[ 7.891 t/h >l operauon :~1c 258.35[.0.29 l t/~"-Ioperauon o] c 207.89 I / 8.373t/h [ 0.144t/h I ~ I .. . la 500.00 "- operauon " 4 h a 400.00 >Ioperauon I Ib 1054.31 ~4.547t/h ' I b 197.40 c334.882 c219.66 Fig. 1. The regeneration reuse design for this example f
NOTATION C :concentration of the contaminant, ppm f : water flowrate that can be used to solve the problem, t/h
L~j.,, L2j,, : the jst water source quantity used by operation i, t/h Lt.j, L2.j : gross amount of
the jst water source, t/h
m;,k : mass load of the kth contaminant of operation i, kg/h Subscripts: ws: fresh water r: regenerated water k: the kth contaminant j: the jst water source 1 and 2: the process before regeneration and the process atter regeneration REFERENCES [ 1] Y. P. Wang, R. Smith, Chem. Engng. Sci., 49(7) (1994a) 981. [2] Z. Liu, American Institute of Chemical Engineering spring meeting, Session 43 (1999) Houston, TX. [3] S. J. Doyle and R. Smith, Transaction of International Chemical Engineering, Part B, 75(3) (1997) 181. [4] A. Alva-Argaez, A. C. Kokossis. and R. Smith, American Institute of Chemical Engineering Annual Meeting (1998a) Paper 13f. Miami, FL. [5] A. Alva-Argaez, A. C. Kokossis. and R. Smith, Comp. chem. Engng., 22(Suppl.) (1998b) $741. [6] A. Alva-Argaez, A. Vallianatos and A. C. Kokossis, Comp. chem. Engng., 23 (1999) 1439. [7] N. Benko, E. Rev, Z. Szitkai and Z. Fonyo, Comp. chem. Engng., 23(Supp.) (1999) $589. [8] N. Takama, T. Kuriyama, K. Shiroko and T. Umeda, Comp. Chem. Engng. 4 (1980) 251. [9] M. J. Bagajewicz, M. Rivas and M. J. Savelski, Comp. chem. Engng., 24(2000) 1461. [10]M. J. Bagajewicz, Comp. chem. Engng., 24 (2000) 2093.
Process SystemsEngineering2003 B. Chen and A.W.Westerberg(editors) 9 2003 Publishedby ElsevierScienceB.V.
Optimum design of water-utilize systems featuring regeneration re-use for multiple contaminants Dong-mei Xu a , Yang-dong Hu a'*, Ben Hua b, Xiu-lin Wang a
College of Chemistry and Chemical Engineering, Ocean University of China, Qingdao, 266003, P. R. China. a
blnstitute of Chemical Engineering, South China University of Technology, Guangzhou 510640,P. R.China Abstract The optimum design of the water-utilize systems featuring regeneration re-use for multiple contaminants is addressed. A global automated optimization design method targeting minimum fresh water consumption, which is based on sequential operation model, is presented, and the mathematical formulation is a MINLP, which can be solved by the computer code based on the hybrid genetic algorithm. This computer code consists of the outer and inner loops: the outer loop is used to optimize the operation sequence based on the adaptive values fed back from the inner loop which is used to optimize the design for a population of given operation sequences. One example is solved to demonstrate the advantage and effectiveness of the new method. Keywords design, wastewater minimization, water utilize system, non-linear programming, regeneration reuse, multiple contaminants
I. INTRODUCTION Water pollution and scarcity of water are a life-and-death problem facing with human being, and stricter discharge regulations, has caused the price of freshwater and wastewater treatment facilities to rise. Therefore, for process industries, it is worthy of investigation evolving water utilize networks to reduce both fresh water consumption and wastewater production. Up to now, several related studies for multiple contaminants problem have been published. Wang and Smith tll addressed a targeting and network design procedure based on elaborate special shifting of streams on the concentration-load diagram for multiple contaminants. Liu t21 presented a few interesting heuristic rules. Doyle and Smith TM started solving the multi-component problem using an iterative procedure and then they proposed to solve the NLP multi-component problem based on the forgoing solution. Alva-Argaez et al. t4'51 presented a solution approach for multiple contaminant systems in combination with water "Corresponding author: Yang-DongHu Tel:+86-532-2032141 E-mail: [email protected] College of Chemistryand Chemical Engineering, Ocean Universityof China, Qingdao, 266003, P. R. China.
1083 treatment in which they include piping costs as well as treatment costs, later E61they discussed some trans-shipment models based on several assumptions, some of them restrictive. Benko et al. [7] modeled the example problem proposed by Takama et al. t81 as a non-convex MINLP problem. Bagajewicz et al. t91 used a combination of mathematical programming and necessary conditions of optimality to approach the design of water utilization with multiple contaminants, later, Bagajewicz tl~ (2000) presented a review of the procedures to design and retrofit water networks in which multi-component systems are also addressed. Regeneration re-use is a usual and significant approach to reduce the amount of fresh water that the overall water utilize system needs, however the design of a water utilize system featuring regeneration re-use for multiple contaminants is a complex problem that involves lots of decision variables, which has not been solved satisfactorily, in this work we approach the above problem and present a simultaneously targeting-design, global automated optimization method that targets minimum fresh water consumption for the overall process. 2. A GLOBAL AUTOMATED OPTIMIZATION DESIGN M E T H O D Given are: a set of water-using operations to be considered in the water system, each one is described by its limiting operation data: the maximum inlet, outlet concentration of water profile of each contaminant and mass load of contaminant to be transferred. The inlet and outlet concentration limits account for corrosion, fouling and maximum solubility, etc Eli. The objective is to design a minimum fresh water consumption water network when the intermediate regeneration is allowed. The following assumption will be utilized: (1) It is allowed wastewater being blended with wastewater from other operations and fresh water; (2) All the operations are physical changes without chemical reaction. 2 possibilities to any operation are considered: possibility 1: it is forbidden to split the operation; possibility 2: it is allowed to split the operation by the fraction of mass load. In this paper, the mass load fraction of the operation i in the process before regeneration of a; is introduced, which can be used to distribute all the operations to the process before regeneration and aider regeneration conveniently. And for the performance of the regeneration processes tll, the "removal ratio" specification is adopted. As a consequence of presence of multiple contaminants, sometimes it is difficult to ascertain the optimal operation sequence, which is a variable needs to be optimized and will be discussed later on. First we address the sequential operation model under the condition that the operation sequence is given.
2.1 Sequential operation model constitution for a given operation sequence Step 1: Assume that all the operations in the water utilize system are distributed to the process before regeneration and the process atter regeneration by their fractions a, : for the operations of which the maximum inlet concentration of a certain contaminant equals the value of 0, the aggregation of them is denoted by Ob (these operations can't match with the regenerated water), their fraction a, take the value of 1, i.e. they are assigned to the process before regeneration entirely; for the operations forbidden to be split, the aggregation of them is denoted by Oforbidden, their a, take the value of 0 or 1,but take the value in the region[0,1 ]
1084 for the operations allowed to be split, the aggregation of them is denoted by Oallowed. Step 2: Assume all the outlet water streams from the former operations are the water sources of the latter operation i on condition that the operation sequence is given, i.e. all the precursors of operation i send wastewater to operation i. Then number all the available water sources in the process before regeneration and the process after regeneration: in the process before regeneration, prescribe that outlet wastewater of thejth operation be regarded as thejth water source, and when j takes the value of 0, the fresh water is specified. In the process after regeneration, prescribe that the regenerated water be regarded as the 1st water source except for the fresh water, i.e., j takes 1. Accordingly in turn the outlet wastewater of the lth operation be regarded as the 2 nd water source, the outlet wastewater of the jth operation be regarded as the (j+l)th water source, and w h e n j takes the value of 0, the fresh water is also specified. Step 3: Match each operation with all its available water sources in the process before regeneration and the process after regeneration respectively. Based on the above steps, the sequential operation model of water system with regeneration reuse for multiple contaminants is constituted. It is assumed that all the outlet wastewater streams from the process before regeneration are mixed to form one stream before it is regenerated.
2.2 Mathematical formulation of the design of the multiple contaminant system For a water-utilize system with m contaminants composed of n operations, the design problem can be formulated as a MINLP1 on the basis of the sequential operation model constituted above that is shown as follows. min(min f ws,t) l ~ S
(MINLP1)
(S: aggregation of all possible operation sequences of P~ in number,)
(1)
subject to: a) The process before regeneration
"~ aimi, ,
= (C,,, - C.,,k )fwsl
(k--1,2 ...... m)
(2)
i=!
t,,o,i = fw,,
(3)
i=1
Z C'.j.kL'a.i/j~L'.J.i < Ci,,(max),i,k
(j:O,1.2 ...... i-l, k : l , 2 ...... m)
(4-i)
jewi
C~,i,k < Cou,,(m=)~,k Z Ll,j,i(Cl,i,k -Cl,J,k) = aimi,*
(k=l,2 ...... m) 0=0,1.2 ...... i-l, k=l,2 ...... m)
(5-i) (6-i)
jew~ i-I
L,.j, i < L,,j - E Ll,j.g (J E Wl, Wl-{j [ the water sources of operation i, j=0,1 ...... i-1 } g=j +l
(7-i)
1085 For the constraints (4-i) - (5-i) - (6-i) and (7-i), let i=1,2 ...... n, then the successive constraints for the n operations in the process before regeneration are listed. b) the process after regeneration L <~fwsl
(8)
= f w,,t •;'•L2,o,,
- f wsl
i=!
•Q L2,u = fr
(9)
(10)
i=1
Similar to the constraints (4-i) - (5-i) - (6-i) and (7-i), the constraints ( 1l-i) - (12-i) (13-i) and (14-i) in the process after regeneration can be established by substituting the variables Lj,j,i, CI,j, k for the vafiablesL2,j.i, C2.j,k, and the difference from the process before regeneration is, that the regenerated water as a water source should be considered. Here we also let i= 1,2 ...... n. Besides, the fractions of all the operations a, are subject to the constraints below. a,=l
(i~Ob)
(15)
a, (1- a,) = 1
(i E Oforbidden)
(16)
(i E Oallowed)
(17)
O< oti
Objective 1 is to seek the minimum fresh water flowrate that can be used to solved the problem, which is a nested objective composed of the inner layer and outer layer: the inner layer is for a given operation sequence and the outer layer is for all the possible operation sequences based on the results from inner layer, i.e. the optimization of the overall water utilize system. Equation 2 is the total mass transfer balance for each contaminant in the process before regeneration. Equation 3, 9 and 10 are the total fresh and regenerated water balance in the process before regeneration and the process after regeneration. Inequation 8 represents the constraint to the total regenerated water. Inequation 4-i and 5-i represent the constraints of inlet and outlet condition of each contaminant to operation i. Eequation 6-i is the mass transfer balance for each contaminant of operation i. Inequation 7-i represents the constraints of the flowrate of each water source matched with operation i. The above MINLP 1 is solved by the computer code programmed by our group based on the hybrid genetic algorithm, which takes the crossover of PMX and reversal mutation for the operation sequence code. This computer code consists of the outer and inner loops: the outer loop is used to optimize the operation sequence based on the adaptive values fed back from the inner loop and the inner loop is used to optimize the design by solving the MINLP to come at the adaptive value of each given operation sequence for a population of operation
1086 sequences. In the solving process, the infeasible matches of the water sources and operations can be automated eliminated, then the values of all the variables for the design of water network with regeneration reuse are yielded. However the regenerated water flowrate obtained fr by solving MINLP1 is probably an arbitrary value meeting the constraints instead of the minimum compatible with minimum fresh water flowrate. To solve the problem, We take minimum fresh water flowrate fws,t and the operation sequence l attained above as known quantity, and substitute objective function min(minfws,t) in MINLP1 for objective function min fr to establish MINLP2 under the same constraints, which can be easily solved to determine minimum regenerated water flowrate f,. compatible with minimum flesh water flowrate and other variables again, accordingly the flowsheet of the overall process is acquired. 3. EXAMPLE Consider an example composed of 6 operations, whose limiting operation data is shown in table 1. Here let us suppose an removal ratio R of 0.9 and determine the regeneration reuse network targeting minimum fresh water flowrate by the method presented in this paper in such a case: "it is allowed splitting all the operations by their fraction of mass load", which are among all the possible cases of example 1. The flowsheet is shown in Fig. 1. And the minimum fresh water and the corresponding minimum regenerated water flowrate are 27.960 t/h and 24.746 t/h respectively. 4. CONCLUSION This paper presented a simultaneous targeting-design, automated optimum method based on sequential operation model for the design of multiple-contaminant water-utilize systems featuring regeneration reuse that minimizes fresh water consumption. Two possibilities to any operation and the optimization of operation sequence are considered, therefore it can be used as a general-purpose method for designing water networks. It has the advantages that for the regeneration reuse possibility the optimum design is able to be achieved without knowing the fresh and corresponding regenerated water requirements in advance and the infeasible matches of the water sources and operations can be automated eliminated. Table 1. Limiting operation data No. ContaCin,max minants (ppm) Salts 300 5000 H2S Ammonia 1500 Salts 10 0 H2S Ammonia 0 Salts 10 0 H2S Ammonia 0
Cout,max Load (ppm) 500 11000 3000 200 500 1000 1000 2000 3500
(kg/h) 0.18 0.75 0.1 3.61 0.25 0.8 6 15 10
No. 4 5 6
Contaminants Salts
Cin,max Cout,max Load (ppm) (k#h)
(ppm) 100 50 H2S Ammonia 1000 100 Salts 50 H2S Ammonia 1000 85 Salts 300 H2S Ammonia 200
400 2000 3500 350 1800 3500 350 6500 1000
2 0.8 1
3 1.9 2.1 3.8 1.1 2
1087 fresh water
a 200.00
~18.05t/h ~~176 5 >l[240787s I . . . . I . I a 800.00 [ "-I I l" 7.5t/h "] operation 3 [ b 2000.00> / | c 1333.33 regeneratec~water L.~l.529t/h a 350.00 h2.164t/h a 350.00 1 " ~ I .. .-Ib 251 71 ~ I .. .. Ib 152 49,_ 9 " -[ 7.891 t/h >l operauon :~1c 258.35[.0.29 l t/~"-Ioperauon o] c 207.89 I / 8.373t/h [ 0.144t/h I ~ I .. . la 500.00 "- operauon " 4 h a 400.00 >Ioperauon I Ib 1054.31 ~4.547t/h ' I b 197.40 c334.882 c219.66 Fig. 1. The regeneration reuse design for this example f
NOTATION C :concentration of the contaminant, ppm f : water flowrate that can be used to solve the problem, t/h
L~j.,, L2j,, : the jst water source quantity used by operation i, t/h Lt.j, L2.j : gross amount of
the jst water source, t/h
m;,k : mass load of the kth contaminant of operation i, kg/h Subscripts: ws: fresh water r: regenerated water k: the kth contaminant j: the jst water source 1 and 2: the process before regeneration and the process atter regeneration REFERENCES [ 1] Y. P. Wang, R. Smith, Chem. Engng. Sci., 49(7) (1994a) 981. [2] Z. Liu, American Institute of Chemical Engineering spring meeting, Session 43 (1999) Houston, TX. [3] S. J. Doyle and R. Smith, Transaction of International Chemical Engineering, Part B, 75(3) (1997) 181. [4] A. Alva-Argaez, A. C. Kokossis. and R. Smith, American Institute of Chemical Engineering Annual Meeting (1998a) Paper 13f. Miami, FL. [5] A. Alva-Argaez, A. C. Kokossis. and R. Smith, Comp. chem. Engng., 22(Suppl.) (1998b) $741. [6] A. Alva-Argaez, A. Vallianatos and A. C. Kokossis, Comp. chem. Engng., 23 (1999) 1439. [7] N. Benko, E. Rev, Z. Szitkai and Z. Fonyo, Comp. chem. Engng., 23(Supp.) (1999) $589. [8] N. Takama, T. Kuriyama, K. Shiroko and T. Umeda, Comp. Chem. Engng. 4 (1980) 251. [9] M. J. Bagajewicz, M. Rivas and M. J. Savelski, Comp. chem. Engng., 24(2000) 1461. [10]M. J. Bagajewicz, Comp. chem. Engng., 24 (2000) 2093.