Water Management in Process Industries ... - ACS Publications

Ind. Eng. Chem. Res. 2008, 47, 1111-1119

1111

Water Management in Process Industries Incorporating Regeneration and Recycle through a Single Treatment Unit Santanu Bandyopadhyay* Energy Systems Engineering, Indian Institute of Technology, Bombay, Powai, Mumbai 400076, India

Calin-Cristian Cormos Department of Chemical Engineering, Faculty of Chemistry and Chemical Engineering, UniVersity “Babes¸ -Bolyai”, Arany Janos 11, Cluj-Napoca 400028, Romania

Process water management has gained increased attention recently, because of awareness of the danger to the environment that is caused by the overextraction of water. At the same time, the imposition of stricter discharge regulations has caused an increase in the effluent treatment costs, which implies capital expenditure with little or no return. Process water management can be divided into two distinct activities: minimization of freshwater requirement and optimal treatment of wastewater generated from the process. Applying the concept of regeneration and recycling of wastewater in the process, freshwater requirement can be reduced significantly while satisfying environmental regulations. Graphical representation, as well as analytical algorithms, are proposed to address integrated process water management issues that involve regeneration and recycle through a single treatment unit. 1. Introduction Because of environmental regulations, water management has become an important issue for process engineers. The cost of treating wastewater streams is increasing steadily as environmental regulations are becoming more and more stringent, requiring also capital expenditure with little or no influence on economic benefits. Principles of process integration, which were first developed for the efficient use of energy, may be applied to address the issues related to water management in a process industry. Process water management may be classified into two distinct activities: optimum allocation of reusable water to minimize the freshwater requirement, and optimal treatment of wastewater generated to meet the environmental regulations. Bandyopadhyay et al.1 developed a procedure to target effluent treatment simultaneously with the freshwater targeting. Freshwater requirement in a process industry can be significantly reduced using different techniques:2 reusing wastewater that is generated by some processes for other processes that require lower water quality; recirculation of the wastewater after partial or total regeneration; and through regeneration and recycle of wastewater. In this paper, a methodology has been proposed to address issues related to process water management with regeneration and recycle while satisfying environmental regulations. El-Halwagi and Manousiouthiakis3 proposed a methodology to target mass-exchange networks. Wang and Smith4 proposed a graphical method for targeting the freshwater requirement through water reuse. These methods are applicable for units, which can be modeled as mass-transfer units (e.g., washing, scrubbing, etc.), with water being used as a mass-separating agent (the water comes into contact with process materials and becomes contaminated). These operations have fixed contaminant loads, and the maximum allowable inlet and outlet concentrations are specified. These methods cannot be applied * To whom correspondence should be addressed. Tel.: +91-2225767894. Fax: +91-22-25726875. E-mail address: santanu@ me.iitb.ac.in.

for process applications such as cooling towers and boilers, because these units cannot be modeled as mass-transfer operations.5 These operations have specified inlet and outlet flow rates, which may not necessarily be equal and, therefore, can account for water losses or gains inside a process. These units may be modeled with outlet streams leaving at a specified concentration, while the inlet streams have a maximum allowable concentration.6 Wang and Smith7 extended the previous method for problems with fixed flow-rate operations through local recycling and splitting of water-using operations to meet the flow-rate constraints. The targeting methodology proposed by Dhole et al.5 involves separate plots for source and demand composites, with the flow rate as the horizontal axis and the contaminant concentration as the vertical axis. However, the procedure developed by Dhole et al.5 does not give true targets, because the mixing of two streams can relax the pinch and, therefore, decreases the freshwater requirement.8 Sorin and Bedard9 and Polley and Polley10 proposed a set of rules for solving fixed flow-rate problems. Hallale8 developed a new graphical iterative approach called a water surplus diagram, for targeting the minimum freshwater requirement for fixed contaminant-load problems as well as fixed flow-rate problems. Manan et al.11 developed a tabular approach to ease the exercise of graphical iterations. A rigorous graphical technique, on the cumulativecontaminant-load-cumulative-flow-rate diagram, was developed to target the freshwater requirement.12,13 Mathematical optimization techniques were also applied to solve the freshwater minimization problem.14-16 Pillai and Bandyopadhyay6 developed a mathematically rigorous procedure to target the minimum resource requirement. The proposed methodology of Pillai and Bandyopadhyay6 effectively combines the mathematical rigor of mathematical optimizations with an algorithmic approach of process integration. Water targeting for batch processes was developed by Foo et al.17 The design and treatment of wastewater streams that contain contaminants in large quantities is an important part of the water management process, to guarantee environmental limits. Wang

10.1021/ie0712232 CCC: $40.75 © 2008 American Chemical Society Published on Web 01/18/2008

1112

Ind. Eng. Chem. Res., Vol. 47, No. 4, 2008

and Smith18 developed an approach for the design of distributed effluent treatment systems. Kuo and Smith19 extended this procedure for multiple treatment processes. Mathematical optimization techniques were also used to design a distributed effluent treatment system.20,21 Overall water management in a process industry is usually performed sequentially. The designs of water-using processes are addressed first and subsequently, based on the designed water reuse network, the distributed effluent treatment system is designed. This sequential procedure may lead to a sub-optimal solution for the distributed effluent treatment system.1 Takama et al.22 solved the complete water management problem using nonlinear optimization techniques. Kuo and Smith23 presented a methodology to discuss the interaction between operations that use water and effluent treatment systems. Bandyopadhyay et al.1 introduced a source composite curve-based approach for simultaneously targeting a distributed effluent treatment system and the minimum freshwater requirement. The proposed methodology is applicable to problems that involve both fixed contaminated load and fixed flow rate. Bandyopadhyay24 generalized the concept of the source composite to reduce waste production for a variety of applications by applying the unified algorithmic procedure, called the waste targeting algorithm (WTA), to water management, hydrogen management, and material reuse/recycle in a metal degreasing process. Shenoy and Bandyopadhyay25 applied the concept of source composite curve to target multiple resources for optimal resource allocation. Wastewater can be regenerated by removing contaminants that have accumulated and then the water may be recycled within the process. A significant amount of freshwater can be saved using wastewater regeneration.3,4,22 In case the treated water is allowed to re-enter processes in which it has previously been used, further conservation in the freshwater requirement is possible. Feng et al.26 presented a sequential graphical methodology for problems that involve fixed contaminant load, to reduce the freshwater requirement and wastewater generation to the maximum extent through regeneration and recycle. Similarly, in most of the methodologies developed to study the effect of regeneration and reuse or regeneration and recycle, the environmental regulations have been neglected. In the case of regeneration and recycling, a treatment unit can simultaneously also act as a regeneration unit. Applying the concept of regeneration and recycling of wastewater in the process, freshwater requirement can be reduced significantly while satisfying environmental regulations. Using the concept of regeneration and recycling, it is even possible to design processes with the minimum discharge of wastewater.27 There exist two extreme limits. In the limiting case of a minimum discharge of wastewater, a significant amount of wastewater must be recycled after regenerating in a treatment unit. The operating cost is typically high for such a case, because of the high cost of effluent treatment. For the other limiting case (no effluent treatment), a significant amount of freshwater is required in the process, such that no treatment is necessary. The operating cost, in such a case, is high, because of the high cost of freshwater. Based on the relative costs of freshwater and effluent treatment, there exists an optimum operating cost of the overall process. A methodology has been proposed in this paper to target the minimum flow rate of effluent water to be treated in a treatment unit to simultaneously satisfy environmental regulations and a given freshwater supply for a process plant with several water-using units. Graphical representations, as well as analytical algorithms, are proposed to address integrated process water management issues that involve regeneration and recycle.

The conceptual approach presented in this paper is restricted to a single contaminant. 2. Problem Statement The general problem of water management in a process industry that considers regeneration and recycle may be mathematically stated as follows. In a process, a set of Ns internal sources of wastewater is given. Each source produces a flow rate Fsi with a given concentration of Csi. A set of Nd internal demands of water is also given. Each demand accepts a flow rate Fdj with a concentration that must be less than a predetermined maximum limit Cdj. There is an external freshwater source with a concentration Crs, and there is a treatment unit of wastewater. There is also an external demand, called wastewater, which can be discharged to the environment and it must be less than a predetermined maximum limit of Ce, governed by the local environmental regulations. There is no flow rate limitation associated with the resource and the waste. The costs of freshwater and effluent treatment per unit of mass are given as κrs and κT, respectively. Wastewater may be treated in the treatment unit and may be recycled in the process. The objective of this work is to develop an algorithmic technique with graphical representation that will identify a cost optimum strategy for integrating wastewater sources and water demands, incorporating regeneration and recycle while satisfying the environmental limit. 3. Targeting Procedure Bandyopadhyay et al.1 proposed a novel limiting composite curve, called a source composite curve, to simultaneously target the minimum freshwater requirement, maximum water reuse, minimum wastewater generation, and minimum effluent to be treated to meet environmental norms. The source composite curve is plotted on a diagram showing contaminant load (m) versus concentration (C). However, the proposed methodology does not consider regeneration and recycle. In the following section, a methodology is presented to target the minimum effluent to be treated in the treatment unit to meet environmental norms, for a given supply of freshwater. Generation of the source composite curve for a given freshwater supply has been developed first. 3.1. Source Composite Curve for a Given Freshwater Flow Rate. Every fixed contaminant load problem may be converted to a fixed flow rate problem at the targeting stage, with the limiting water flow rate being the specified flow rate for each process.6 Therefore, a water allocation problem may be modeled with outlet streams leaving at a specified concentration and flow rate (sources of wastewater), while the inlet streams have a maximum allowable concentration and a specified flow rate. The outlet of any process may be viewed as a source of wastewater, whereas the inlet to any process represents the scope of reusing wastewater or demands of wastewater. The source composite curve represents the maximum contaminant load at different contaminant concentrations. Physically, the source composite curve (as defined by Bandyopadhyay et al.1) is equivalent to the grand composite curve in heat exchanger network synthesis28,29 and invariant rectifying and stripping curves in distillation.30-34 An algebraic procedure for generation of source composite curve for a given freshwater supply is given below. Numerical results, calculated in each step, may be tabulated in the form of a table for ready reference. For better understanding an example, consisting four water-using processes, has been considered. The limiting process data for this example are given in Table 1.1,10

Ind. Eng. Chem. Res., Vol. 47, No. 4, 2008 1113 Table 1. Limiting Process Data for Example 1a Inlet/Demand

Outlet/Source

process

contaminant concentration (ppm)

flow rate (t/h)


flow rate (t/h)

P1 P2 P3 P4

20 50 100 200

50 100 80 70

50 100 150 250

50 100 70 60

a Conditions: concentration of contaminant in freshwater, C ) 0 ppm; fw environmental limit for discharge concentration, Ce ) 50 ppm; removal ratio of the treatment unit, r ) 0.95; maximum inlet concentration to the treatment unit, not specified; cost of freshwater, κrs ) $1.5/t; and cost of effluent treatment, κT ) $1.5/t.

Step 1: For the given problem with the given freshwater supply, determine the amount of wastewater to be discharged to the environment, using the overall mass balance. The overall flow balance across the process can be simplified as Ns

Frs +

∑ i)1

Nd

Fsi )

Fdj + Fe ∑ j)1

(1)

where Frs and Fe denote the flow rate of freshwater supply and wastewater to be discharged to the environment. Equation 1 may be rewritten as

Fe ) Frs + ∆

(2)

where ∆ is a constant for a given process and is defined as Ns

∑ i)1

Figure 1. Source composite curve and the targeting minimum treatment flow rate for Example 1 with a freshwater supply of 20 t/h.

Table 2. Generation of Source Composite Curve for the Four-Process Example concentration, C (ppm)

flow, F (t/h)

cumulative flow, Q (t/h)

mass load, P (kg/h)

cumulative load, ∆m (kg/h)

250 200 150 100 50 20 0

60 -70 70 20 -50 -50 20

60 -10 60 80 30 -20 0

0 3 -0.5 3 4 0.9 -0.4

0 3 2.5 5.5 9.5 10.4 10

demands has been adopted. The net flow rates of water, which correspond to each concentration for Example 1, are tabulated in the second column of Table 2. Step 4: The corresponding entries in the third column of Table 2 are cumulative flow rates given by the formula

Nd

Fsi -

Fdj ∑ j)1

∆ signifies the overall water loss/gain in the system. A positive ∆ signifies that there is water generation in the system and a negative ∆ implies overall water loss. For a given freshwater supply, flow rate of the wastewater to be discharged can be determined using eq 2. Consider the extended problem with freshwater as a source and wastewater as a demand.Note that ∆ for the extended problem should be zero, because it satisfies the overall mass balance. For Example 1, ∆ ) -20 t/h, which suggests that there is an overall water loss of 20 t/h. Therefore, 20 t/h of freshwater supply does not produce any wastewater. It may be noted that the ∆ value for the extended problem with 20 t/h of freshwater is identical to zero. Step 2: Arrange all the distinct contaminant concentrations (of freshwater, demands, and sources together) in descending order in the first column. Without any loss of generality, the entries of the first column are given as C1 > C2 > C3 > ...> Cn. The last entry of this column should be zero (Cn ) 0). Note that the concentration of the freshwater (Crs) is not required to be zero. It may be possible that, for a particular problem, there exists no stream with a zero concentration. However, the algorithm demands that the last entry should be zero (Cn ) 0). To satisfy this condition, a pseudo-stream with zero concentration and zero flow may be assumed. Distinct concentrations of Example 1 are arranged in descending order in the first column of Table 2. Step 3: For each concentration Ci (in the first column), put the corresponding net flow Fi in the second column. The net flow Fi is calculated by taking the algebraic sum of flow rates corresponding to a given concentration Ci. The convention of positive flow rates for sources and negative flow rates for

i

Qi )

Fj ∑ j)1

(3)

The last entry of this column should again be zero (Qn ) 0), because the wastewater flow rate has been determined based on the overall mass balance of the entire problem. For Example 1, cumulative flow rates, calculated using eq 3, are reported in the third column in Table 2. Step 5: Calculate the entries of the fourth column, using the formula

Pi ) Qi-1(Ci-1 - Ci)

(4)

Note that the value of Q0 is assumed to be zero. The fourth column of Table 2 consists of entries calculated using eq 4. Step 6: The fifth column of Table 2 contains the cumulative contaminant mass load,

∆mi )

i Pj ∑j)1

(5)

Now, the fifth column in Table 2 (cumulative mass load) may be plotted against the first column (concentration) to obtain the source composite curve for a given freshwater flow rate. The last entry (mn) in the fifth column of Table 2 signifies the total mass load of the entire process. For the regeneration and recycle of wastewater, the same must be thrown out of the system through treatment unit. The fifth column in Table 2 contains the cumulative contaminant load for Example 1. The last entry of the fifth column in Table 2 suggests that 10 kg/h of contaminant must be removed from the process (mn ) 10). The source composite curve is shown in Figure 1. Step 7: Draw a vertical line on the diagram of cumulative mass load versus concentration diagram at mn and determine

1114


In combination with eq 7, the equation of the treatment line is given as

Table 3. Targeting Minimum Effluent Treatment Flow Rate concentration, C (ppm)


treatment flow rate, fT (t/h)

250 200 150 100 50 33.33

0 3 2.5 5.5 9.5 10

42.11 37.63 53.51 50.26 20.53 15.79

∆m )

the concentration at which it cuts the source composite curve. If the vertical line intersects the source composite curve at several points, consider the point with the lowest concentration. Physically, the source composite curve, generated using the first six steps of the proposed algorithm, is equivalent to the grand composite curve in heat exchanger network synthesis.28,29 Similar to a pocket in a grand composite curve, a pocket is produced in the source composite curve. A pocket represents a local deficit and a local surplus that are in perfect balance. By drawing the vertical line, the pocket is removed from the source composite curve and only the active portion of the source composite curve is generated. The vertical line drawn on the diagram of the cumulative mass load versus concentration at mn removes the pocket from the source composite curve. For Example 1, a vertical line at mn ) 10 intersects the source composite curve at 33.33 ppm. Step 8: Construct another table of concentration and cumulative mass load, up to the intersection point. This may be defined as the active source composite curve, for a given freshwater supply. Table 3 represents the contaminant concentration and cumulative mass load up to the intersection point or the data for the active source composite curve for Example 1. 3.2. Targeting Minimum Effluent Flow Rate to be Treated. After generating the source composite curve for a given freshwater flow rate, the next step is to determine the minimum effluent treatment flow rate necessary to satisfy the environmental regulations. A treatment unit may be modeled in two ways: with a constant outlet concentration and with a constant removal factor. Furthermore, there may be an added constraint on the maximum inlet concentration to a treatment unit. For a treatment unit with a constant removal factor, the outlet concentration (CTout) of the treated effluent is dependent on the inlet concentration (CTin). The removal ratio (r) of the contaminant load from the system is defined as

r)

fTCTin - fTCTout CTin - CTout ∆mn ) ) fTCTin CTin ∆mT

(6)

To remove a quantity ∆mn (in kg/h) of contaminant load in the effluent treatment unit, the input load for the treatment plant must be ∆mT kg/h. Equation 6 may be rewritten to calculate the input contaminant load to the treatment unit:

∆mT )

∆mn r

(7)

Any treatment line (i.e., the line that represents the removal of contaminant load on a contaminant load versus concentration (∆m-C) diagram) must pass through the point (∆mT,0) for a fixed removal ratio. Any straight line on a ∆m-C diagram passing through the point (∆mT,0) with a flow rate of fT may be represented as follows:

∆m ) ∆mT - fTC

∆mn - fT C r

(8)

The minimum treatment flow rate can be targeted by rotating the treatment line with point (∆mn/r,0) as the pivot point, such that it just touches the active source composite curve. The point at which the treatment line touches the active source composite curve is called the treatment pinch point. Note that the point at which it touches the concentration axis corresponds to the inlet concentration to the treatment unit. In case the specified maximum inlet concentration to the treatment unit is less than the inlet concentration obtained using the previous method, the effluent treatment flow rate must be increased to satisfy this constraint. In such a case, the minimum effluent flow rate to be treated may be calculated directly using eq 8:

fT )

∆mn rCTin,max

(9)

On the other hand, for a treatment unit with a constant outlet concentration, any treatment line on a ∆m-C diagram must pass through the point (∆mn,CTout). The equation of the treatment line, for a treatment unit with a fixed outlet concentration, is given as

∆m ) ∆mn - fT(C - CTout)

(10)

Similar to the previous procedure, the minimum treatment flow rate can be targeted by rotating the treatment line with point (∆mn,CTout) as the pivot point, such that it just touches the active source composite curve. The point at which the treatment line touches the active source composite curve represents the treatment pinch point and the point at which it touches the concentration axis represents the inlet concentration to the treatment unit. If the specified maximum inlet concentration to the treatment unit is less than the inlet concentration obtained using rotating the treatment line, the minimum effluent flow rate to be treated may be calculated directly using the following equation:

fT )

∆mn CTin,max - CTout

(11)

Algorithmic steps, in continuation of the previous steps, are described below, to target the minimum effluent treatment flow rate. Step 9: Based on the type of the treatment unit, determine the pivot point. For a treatment unit with a constant removal ratio, the pivot point is (∆mn/r,0), and for a treatment unit with a fixed outlet concentration, the pivot point is (∆mn,CTout). To generalize the applicability of the proposed algorithm, let us denote the pivot point as (∆m*,C*). For Example 1, the treatment unit has a fixed removal ratio, without any limit on the inlet concentration. The pivot point for the treatment unit is calculated to be (10.53,0). This signifies that the input mass load to the treatment unit must be 10.53 kg/h (10.0/0.95 ) 10.53). Step 10: In the third column of the table that contains the active source composite curve (obtained using Step 8), the treatment flow rate that corresponds to each entry is calculated.

Ind. Eng. Chem. Res., Vol. 47, No. 4, 2008 1115

The corresponding treatment flow rate may be calculated using the following formula:

fTi )

∆m* - ∆mi Ci - C*

(12)

The maximum entry in this column defines the minimum effluent flow rate to be treated (f RT ) without any specified maximum inlet concentration. This methodology is equivalent to the method based on rotating the treatment line until it touches the active source composite curve. The last column in Table 3 represents the effluent flow rate to be treated, as calculated using eq 12. The minimum effluent flow rate to be treated in the treatment unit is calculated to be 53.51 t/h (the maximum entry in the third column of Table 3), which corresponds to a treatment pinch of 150 ppm. Step 11: For a given maximum inlet concentration to the treatment unit, the flow rate of the effluent to be treated in the treatment unit may be calculated as

f LT )

∆m* CTout,max - C*

(13)

The minimum effluent treatment flow rate (FT) to be treated in the treatment unit is the maximum of the terms f RT and f LT. By applying the proposed algorithm, the minimum effluent treatment flow rate (FT) to be treated in the treatment unit can be targeted for a given freshwater flow rate (Frs). Therefore, the operating cost can be calculated as

Φ ) Frsκrs + FTκT

(14)

Now, the same algorithm can be repeated for different freshwater flow rates and the optimum operating cost can be determined. Typically, the operating cost of a treatment unit is dependent on the effluent flow rate. However, in some cases, the operating cost of a treatment unit is dependent on the mass load, volumetric flow rate, outlet concentration, etc. In such cases, eq 14 should be modified appropriately. 4. Illustrative Examples Application of the proposed procedure is demonstrated through illustrative examples. 4.1. Example 1: Four-Process Example with a Treatment Unit of Fixed Removal Ratio. The limiting process data for this example, which consists of four water-using processes, are given in Table 1.1,10 The costs of both the freshwater and the effluent treatment are assumed to be $1.5/t. For this example, ∆ is -20 t/h, which suggests that there is an overall water loss of 20 t/h. If 220 t/h of freshwater is used in this example, it will produce 200 t/h of wastewater. The concentration of the wastewater is 50 ppm, and it may be discharged directly into the environment without treatment. The operating cost is $330/h. Reusing water from other processes, the minimum freshwater requirement can be calculated to be 70 t/h with a corresponding effluent flow rate of 50 t/h and the minimum effluent treatment flow rate is targeted to be 35.96 t/h.1 Application of the proposed methodology confirms these results (not shown for brevity), which implies that, if 70 t/h of freshwater is used, there is no requirement of regeneration. The corresponding operating cost is calculated to be $158.9/h. Considering regeneration and recycle, the freshwater requirement can be reduced further.

Figure 2. Variations in the minimum effluent treatment flow rate and the operating cost for different freshwater supplies for Example 1.

In the limiting case of zero discharge of wastewater, the freshwater requirement is 20 t/h. The proposed algorithm may be applied to determine the minimum effluent treatment flow rate for the limiting case of zero discharge. The first seven steps of the proposed algorithm generate the source composite curve for an example with a freshwater supply of 20 t/h, and the numerical values are tabulated in Table 2. The source composite curve is shown in Figure 1. In this example, the treatment unit has a fixed removal ratio, without any limit on the inlet concentration. The pivot point for the treatment unit is calculated to be (10.53,0). This signifies that the input mass load to the treatment unit must be 10.53 kg/h (10.0/0.95 ) 10.53). The last column in Table 3 represents the effluent flow rate to be treated. The minimum effluent flow rate to be treated in the treatment unit is calculated to be 53.51 t/h, which corresponds to a treatment pinch of 150 ppm. The treatment line and the treatment pinch point are also shown in Figure 1. The inlet concentration of the wastewater sent to the treatment unit may be read from Figure 1 (the point at which the treatment line intersects the concentration axis) or determined using eq 10. In this example, the inlet concentration to the treatment unit is calculated to be 196.72 ppm. The corresponding outlet concentration is 9.84 ppm. The regenerated water available at 9.84 ppm is recycled back into the process, to reduce the freshwater requirement and the corresponding reduction in wastewater generation. The operating cost for the case of zero discharge is $110.3/h. The same algorithm is repeated for different freshwater flow rates and corresponding minimum treatment flow rates, and the operating costs are determined. Variations of the minimum treatment flow rate and the operating cost for different freshwater supplies are shown in Figure 2. The minimum effluent flow rate decreases as the freshwater flow rate increases. This curve is piecewise linear. The slope of the treatment flow rate curve changes at a freshwater supply of ∼101.3 t/h as the treatment pinch jumps from 150 ppm to 250 ppm. Note that the zero discharge case with a freshwater requirement of 20 t/h represents the cost optimal solution. One of the water allocation networks for this example is shown in Figure 3. This network is designed using the principles of the nearest-neighborhood algorithm that was proposed by Prakash and Shenoy.12 4.2. Example 2: Specialty Chemical Plant with a Treatment Unit of Fixed Removal Ratio. The limiting process data for a specialty chemical plant that consists of five water-using processes are given in Table 4.1,7 Similar to the previous example, the costs of both the freshwater and the effluent treatment are assumed to be $1.5/t. For this example, ∆ is -40 t/h, which suggests that there is an overall water loss of 40 t/h. If 869 t/h of freshwater is used in this specialty chemical plant, it will produce 829 t/h of

1116


Figure 4. Source composite curve and the targeting minimum treatment flow rate for Example 2 with a freshwater supply of 90.64 t/h. Figure 3. One possible water allocation networks for Example 1 corresponds to the minimum operating cost. (The values show the flow rate in units of t/h, with contaminant concentrations (in ppm) given within brackets.) Table 4. Limiting Process Data for Example 2a Inlet/Demand

Outlet/Source

process


flow rate (t/h)


flow rate (t/h)

reactor/thickener cyclone filtration steam system cooling system

100 200 0 0 10

80 50 10 10 15

1000 700 100 10 100

20 50 40 10 5

Conditions: concentration of contaminant in freshwater, Cfw ) 0 ppm; environmental limit for discharge concentration, Ce ) 50 ppm; removal ratio of the treatment unit, r ) 0.9; maximum inlet concentration to the treatment unit, not specified; cost of freshwater, κrs ) $1.5/t; and cost of effluent treatment, κT ) $1.5/t. a

wastewater. The concentration of the wastewater is 50 ppm, and it may be discharged directly into the environment without treatment. The corresponding operating cost is $1303.5/h. By reusing water from other processes, the minimum freshwater requirement can be calculated to be 90.64 t/h, with a corresponding wastewater flow rate of 50.64 t/h.1 The minimum removal ratio for this plant is 0.939.1 Therefore, a single treatment unit with a fixed removal ratio of 0.9 is not sufficient to satisfy the environmental norm directly. Bandyopadhyay et al.1 suggested three different ways to overcome this: by adding freshwater to dilute the effluent, by recycling across the treatment unit, or by installing additional treatment unit(s). Bandyopadhyay et al.1 determined that 32.26 t/h of additional freshwater is required to satisfy the environmental norm with a single treatment unit. Therefore, a total of 122.9 t/h freshwater produces 82.9 t/h of wastewater and 50.64 t/h of wastewater must be treated to satisfy the environmental norm. Application of the proposed methodology with a given freshwater of 122.9 t/h confirms these results (not shown for brevity). The corresponding operating cost of the overall system is calculated to be $260.32/h. The treatment unit reduces the concentration of the effluent stream. If some amount of the treated effluent is recycled across the treatment unit, the inlet concentration of the combined effluent may be reduced to meet the environmental norms. Bandyopadhyay et al.1 determined that 35.85 t/h of wastewater may be recycled across the treatment unit. The freshwater requirement remains 90.64 t/h, and the corresponding operating cost is calculated to be $265.7/h. The environmental discharge limit may also be satisfied by installing an additional treatment unit in series. For this example, note that two treatment units in series is sufficient.1 The target for the minimum treatment

Table 5. Generation of Source Composite Curve for Example 2 with a Freshwater Supply of 90.64 t/h concentration, C (ppm)

flow, F (t/h)


mass load, P (kg/h)


1000 700 200 100 50 10 0

20 50 -50 -35 -50.64 -5 70.64

20 70 20 -15 -65.64 -70.64 0

0 6 35 2 -0.75 -2.625 -0.707

0 6 41 43 42.250 39.625 38.918

Table 6. Targeting Minimum Effluent Treatment Flow Rate for Example 2 with a Freshwater Supply of 90.64 t/h concentration, C (ppm)



1000 700 229.74

0 6 38.918

43.24 53.20 18.82

flow rate is 17.94 t/h of effluent to be treated in the first unit, and the combined effluent of 50.64 t/h will be treated in the next treatment unit. The corresponding operating cost of the overall system is calculated to be $238.84/h. Note that the investment cost of the second treatment unit is not considered. Considering regeneration and recycle for this example with a freshwater supply of 90.64 t/h, the requirement for the minimum effluent flow rate to be treated can be reduced further. The first seven steps of the proposed algorithm generate the source composite curve for this example with a freshwater supply of 90.64 t/h, and the numerical values are tabulated in Table 5. The last entry of the fifth column of the table suggests that 38.92 kg/h of contaminant must be removed from the process (mn ) 38.92). The source composite curve is shown in Figure 4. A vertical line at mn ) 38.92 intersects the source composite curve at 229.74 ppm. Table 6 represents the contaminant concentration and cumulative mass load up to the intersection point or the data for the active source composite curve. Similar to the previous example, the treatment unit has a fixed removal ratio, without any limit on the inlet concentration. The pivot point for the treatment unit is calculated to be (43.242,0). The last column in Table 6 represents the effluent flow rate to be treated. The minimum effluent flow rate to be treated in the treatment unit is calculated to be 53.20 t/h, which corresponds to a treatment pinch of 700 ppm. The treatment line and the treatment pinch point are also shown in Figure 4. The operating cost for this case is $215.77/h, which is less than both options presented by Bandyopadhyay et al.1 The same algorithm is repeated for different freshwater flow rates and corresponding minimum treatment flow rates, and the operating cost are determined. Variations of the minimum treatment flow rate and the operating cost for different freshwater supplies are shown in Figure 5. The minimum effluent flow rate decreases as the freshwater flow rate increases. The

Ind. Eng. Chem. Res., Vol. 47, No. 4, 2008 1117 Table 7. Targeting Minimum Effluent Treatment Flow Rate for Example 2 That Has a Treatment Unit with a Constant Outlet Concentration of 60 ppm concentration, C (ppm)



1000 700 229.74

0 6 38.918

41.40 51.43 0

Table 8. Limiting Process Data for Example 3 Inlet/Demand

Figure 5. Variations in the minimum effluent treatment flow rate and the operating cost for different freshwater supplies for Example 2.

Outlet/Source

process


flow rate (t/h)


flow rate (t/h)

P1 P2 P3 P4

0 50 50 400

20 100 40 10

100 100 800 800

20 100 40 10

a Conditions: concentration of contaminant in freshwater, C ) 0 ppm; fw environmental limit for discharge concentration, Ce ) 20 ppm; removal ratio of the treatment unit, r ) 0.9; maximum inlet concentration to the treatment unit, CTin,max ) 350 ppm; cost of freshwater, κrs ) $1.5/t; and cost of effluent treatment, κT ) $1.0/t.

Table 9. Generation of Source Composite Curve for Example 3 with a Freshwater Supply of 90 t/ha

Figure 6. Source composite curve and the targeting minimum treatment flow rate for Example 2 that has a treatment unit with a constant outlet concentration of 60 ppm.


slope of the treatment flow rate curve changes at a freshwater supply of ∼508.9 t/h as the treatment pinch jumps from 700 ppm to 1000 ppm. Note that the zero discharge case with a freshwater requirement of 40 t/h represents the cost optimal solution. The effect of changing the regenerator from a fixed removal ratio to a constant outlet concentration type has also been studied. It has been assumed that another treatment unit of constant outlet concentration of 60 ppm is available. The costs of both the freshwater and the effluent treatment are assumed to be the same. For comparison with the results obtained for the treatment unit with a fixed removal ratio, a fresh water supply of 90.64 t/h has been assumed. Changing the type of treatment unit does not change the source composite curve, as tabulated in Table 5. Because the treatment unit is of the fixed outlet concentration type, the pivot point for the treatment unit is calculated to be (38.92,60). Table 7 represents the active source composite curve and the effluent flow rate to be treated. The minimum effluent flow rate to be treated in the treatment unit is calculated to be 51.43 t/h, which corresponds to a treatment pinch of 700 ppm. The treatment line and the treatment pinch point are shown in Figure 6. It has been observed that the zero discharge case with a freshwater requirement of 40 t/h

concentration, C (ppm)

flow, F (t/h)


mass load, P (kg/h)


800 400 100 50 20 0

50 -10 120 -140 -90 70

50 40 160 20 -70 0

0 20 12 8 0.6 -1.4

0 20 32 40 40.6 39.2

also represents the cost optimal solution for this type of treatment unit. The variation of the operating cost with different freshwater supply has not been shown, for the sake of brevity. 4.3. Example 3: Fixed Contaminant-Load Example with a Treatment Unit of Fixed Removal Ratio. The limiting process data for this example, which consists of five waterusing processes, are given in Table 8.4 The maximum discharge limit for the effluent water is assumed to be 20 ppm. The removal ratio of the treatment unit is assumed to be 0.9, with a maximum allowable inlet concentration to the treatment unit of 350 ppm. The costs of the freshwater and the effluent treatment are assumed to be $1.5/t, and $1.0/t, respectively. This is a fixed contaminant-load problem. Therefore, there is no overall water gain or loss and, hence, ∆ is zero. If 2050 t/h of freshwater is used in this plant, it will produce 2050 t/h of wastewater with a concentration of 20 ppm. The wastewater may be discharged directly into the environment without treatment. The corresponding operating cost is $3075/h. Reusing water from other processes, the minimum freshwater requirement can be calculated to be 90 t/h. This confirms the result reported by Wang and Smith.4 The corresponding effluent flow rate is 90 t/h (because ∆ is zero) and the effluent is available at a contaminant concentration of 455.6 ppm. The proposed algorithm is applied to determine the minimum effluent treatment flow for a supply of 90 t/h of freshwater. The first seven steps of the proposed algorithm generate the source composite curve for an example with a supply of 90 t/h of freshwater, and the numerical values are tabulated in Table 9. The last entry of the fifth column suggests that that 39.2 kg/h of contaminant must be removed from the process (mn ) 39.2). The source composite curve is shown in Figure 7. A vertical line at mn ) 39.2 intersects the source composite curve at 55 ppm. Table 10 represents the contaminant concentration and

1118


Table 10. Targeting Minimum Effluent Treatment Flow Rate for Example 3 with a Freshwater Supply of 90 t/h, without Considering the Limit on Inlet Concentration to the Treatment Unit concentration, C (ppm)



800 400 100 55

0 20 32 39.2

54.44 58.89 115.56 79.19

Table 11. Generation of Source Composite Curve for Example 3 with a Freshwater Supply of 20 t/h concentration, C (ppm)

flow, F (t/h)


mass load, P (kg/h)


800 400 100 50 20 0

50 -10 120 -140 -20 0

50 40 160 20 0 0

0 20 12 8 0.6 0

0 20 32 40 40.6 40.6

cumulative mass load up to the intersection point or the data for the active source composite curve. In this example, the treatment unit has a fixed removal ratio of 0.9, and the maximum allowable inlet concentration to the treatment unit is 350 ppm. The pivot point for the treatment unit is calculated to be (43.56,0). This signifies that the input mass load to the treatment unit must be 43.56 kg/h (39.2/0.9 ) 43.56). The maximum entry in the last column of Table 10 represents the effluent flow rate to be treated without considering the limitation on the maximum inlet concentration to the treatment unit. This corresponds to an effluent flow rate of 115.56 t/h, and the treatment pinch is observed at 100 ppm. This corresponds to an effluent flow rate of 115.56 t/h, and the inlet concentration to the treatment unit may be calculated as 376.92 ppm. This is higher than the permissible limit of 350 ppm. Hence, the effluent flow rate, which is calculated using eq 13, is revised to be 124.77 t/h to satisfy the inlet concentration limit. The treatment line is shown in Figure 7. Note that the treatment line does not pass through the treatment pinch point of 100 ppm. The operating cost for this case is $259.77/h, which is considerably less than the option of no treatment unit. In this example, process P1 requires 20 t/h of freshwater (at 0 ppm). Therefore, the freshwater requirement cannot be reduced below 20 t/h. Hence, the absolute minimum wastewater discharge is also 20 t/h (because ∆ is zero). In such a problem, it is not possible to achieve the zero-discharge condition. The proposed algorithm is applied to determine the minimum effluent treatment flow for a supply of 20 t/h of freshwater. The first seven steps of the proposed algorithm generate the source composite curve for this example with a supply of 20 t/h of freshwater, and the numerical values are tabulated in Table 11. The last entry of the fifth column of the table suggests that 40.6 kg/h of contaminant must be removed from the process (mn ) 40.6). The source composite curve is shown in Figure 8. A vertical line at mn ) 40.6 intersects the source composite curve at 20 ppm. Table 12 represents the contaminant concentration and cumulative mass load up to the intersection point or the data for the active source composite curve. The pivot point for the treatment unit is calculated to be (45.11,0). The maximum entry in the last column of Table 12 represents the effluent flow rate to be treated, without considering the limitation on the maximum inlet concentration to the treatment unit. This corresponds to an effluent flow rate of 225.56 t/h, and the treatment pinch is observed at 20 ppm. This corresponds to an effluent flow rate of 225.56 t/h, and the inlet concentration to the treatment unit may be calculated as 200 ppm, which is well


Figure 9. Variations in the minimum effluent treatment flow rate and the operating cost for different freshwater supplies for Example 3. Table 12. Targeting Minimum Effluent Treatment Flow Rate for Example 3 with a Freshwater Supply of 20 t/h, without Considering the Limit on Inlet Concentration to the Treatment Unit concentration, C (ppm)



800 400 100 50 20

0 20 32 40 40.6

56.39 62.78 131.11 102.22 225.56

within the limit of 350 ppm. The treatment line and the treatment pinch point are shown in Figure 8. The operating cost for this case is $255.56/h. The same algorithm is repeated for different freshwater flow rates and corresponding minimum treatment flow rates, and the operating costs are determined. Variations of the minimum treatment flow rate and the operating cost for different freshwater supplies are shown in Figure 9. The minimum effluent flow rate decreases as the freshwater flow rate increases. Interestingly, note that the minimum discharge case with a freshwater requirement of 20 t/h does not represent the cost optimal solution. The cost optimal solution for this problem corresponds to a freshwater flow rate of 34.91 t/h, with a minimum operating cost of 180.64 $/h. Note that the operating cost curve goes through a minimum not only because of the different cost coefficients assumed for the freshwater and the treatment, but also because of the sharp change in the effluent treatment flow rate (see Figure 9). This is primarily because of the jump of the treatment pinch point. Conclusion Water management in a process industry consists of optimal allocation of reusable water to reduce freshwater requirement and optimal design of effluent treatment units to meet environmental norms. Applying the concept of regeneration and recycling of wastewater in the process, the freshwater require-

Ind. Eng. Chem. Res., Vol. 47, No. 4, 2008 1119

ment can be reduced significantly while satisfying environmental regulations. In this paper, a graphical representation and an analytical algorithm have been proposed to address integrated process water management issues that involve regeneration and recycle. In the proposed methodology, a single treatment unit is utilized to produce regenerated water, as well as to satisfy the environmental regulations. Proposed methodologies and algorithms are applicable for systems with a single contaminant. This is demonstrated through three illustrative examples that consider a single contaminant. In many multicontaminant problems, there exists a limiting component and other components are consistent with the limiting component. In such problems, the proposed methodology can be applied based on the limiting contaminant. Present research is directed toward a general approach for general multicomponent problems. Literature Cited (1) Bandyopadhyay, S.; Ghanekar, M. D.; Pillai, H. K. Process Water Management. Ind. Eng. Chem. Res. 2006, 45, 5287. (2) Smith, R. Chemical Process Design and Integration; Wiley: Sussex, U.K., 2005. (3) El-Halwagi, M. M.; Manousiouthakis, V. Synthesis of Mass Exchange Networks. AIChEJ. 1989, 8, 1233. (4) Wang, Y. P.; Smith, R. Wastewater Minimization. Chem. Eng. Sci. 1994, 49, 981. (5) Dhole, V. R.; Ramchandani, N.; Tanish, R. A.; Wasilewski, M. Make Your Process Water Pay for Itself. Chem. Eng. 1996, 103, 100. (6) Pillai, H. K.; Bandyopadhyay, S. A Rigorous Targeting Algorithm for Resource Allocation Networks. Chem. Eng. Sci. 2007, 62, 6212. (7) Wang, Y. P.; Smith, R. Wastewater Minimization with Flowrate Constraints. Trans. Inst. Chem. Eng. 1995, 73A, 889. (8) Hallale, N. A New Graphical Targeting Method for Water Minimisation. AdV. EnViron. Res. 2002, 6, 377. (9) Sorin, M.; Bedard, S. The Global Pinch Point in Water Reuse Network. Trans. Inst. Chem. Eng. 1999, 77B, 305. (10) Polley, G. T.; Polley, H. L. Design Better Water Networks. Chem. Eng. Progress, 2000, 96, 47. (11) Manan, Z. A.; Tan, Y. L.; Foo, D. C. Y. Targeting the Minimum Water Flowrate Using Water Cascade Analysis Technique. AIChE J. 2004, 50, 3169. (12) Prakash, R.; Shenoy, U. V. Targeting and Design of Water Networks for Fixed Flowrate and Fixed Contaminant Load Operations. Chem. Eng. Sci. 2005, 60, 255. (13) El-Halwagi, M. M.; Gabriel, F.; Harell, D. Rigorous Graphical Targeting for Resource Conservation via Material Recycle/Reuse Networks. Ind. Eng. Chem. Res. 2003, 42, 4019. (14) Alva-Argeaz, A.; Vallianatos, A.; Kokossis, C. A Multi-Contaminant Transshipment Model for Mass Exchange Networks and Wastewater Minimisation Problems. Comput. Chem. Eng. 1999, 23, 1439. (15) Savelski, M.; Bagajewicz, M. On the Optimality Conditions of Water Utilization Systems in Process Plants wWith Single Contaminants. Chem. Eng. Sci. 2000, 55, 5035.

(16) Benko, N.; Rev, E.; Fonyo, Z. The Use of Nonlinear Programming to Optimal Water Allocation. Chem. Eng. Commun. 2000, 178, 67. (17) Foo, C. Y.; Manan, Z. A.; Yunus, R. M.; Aziz, R. A. Synthesis of Mass Exchange Network For Batch ProcessessPart I: Utility Targeting. Chem. Eng. Sci. 2004, 59, 1009. (18) Wang, Y. P.; Smith, R. Design of Distributed Effluent Treatment Systems. Chem. Eng. Sci. 1994, 49, 3127. (19) Kuo, W. C. J.; Smith, R. Effluent Treatment System Design. Chem. Eng. Sci. 1997, 52, 4273. (20) Galan, B.; Grossmann, I. E. Optimal Design of Distributed Wastewater Treatment Networks. Ind. Eng. Chem. Res. 1998, 37, 4036. (21) Herna´ndez-Sua´rez, R.; Castellanos-Ferna´ndez, J.; Zamora, J. M. Superstructure Decomposition and Parametric Optimization Approach for the Synthesis of Distributed Wastewater Treatment Networks. Ind. Eng. Chem. Res. 2004, 43, 2175. (22) Takama, N.; Kuriyama, T.; Shiroko, K.; Umeda, T. Optimal Water Allocation in Petroleum Refinery. Comput. Chem. Eng. 1980, 4, 251. (23) Kuo, W. C. J.; Smith, R. Designing for the Interactions Between Water Use And Effluent Treatment. Trans. Inst. Chem. Eng. 1998, 76A, 287. (24) Bandyopadhyay, S. Source Composite Curve for Waste Reduction. Chem. Eng. J. 2006, 125, 99. (25) Shenoy, U. V.; Bandyopadhyay, S. Targeting for Multiple Resources. Ind. Eng. Chem. Res. 2007, 46, 3698. (26) Feng, X.; Bai, J.; Zheng, X. On the Use of Graphical Method To Determine the Targets of Single Contaminant Regeneration Recycling Water System. Chem. Eng. Sci. 2007, 26, 2127. (27) Cormos, C. C.; Bandyopadhyay, S. Process Water Management with Regeneration and Recycle. In Proceedings of the 17th European Symposium on Computer Aided Process Engineering (ESCAPE 17), Bucharest, Romania, May 27-30, 2007; Plesu, V.; Agachi, P. S., Eds.; Computer-Aided Chemical Engineering Series; Elsevier Science B.V. Publishers: Amsterdam, 2007; p 1343. (28) Shenoy, U. V.; Sinha, A.; Bandyopadhyay, S. Multiple Utilities Targeting For Heat Exchanger Networks. Trans. Inst. Chem. Eng.: Chem. Eng. Res. Des. 1998, 76A, 259. (29) Varghese, J.; Bandyopadhyay, S. Targeting for Energy Integration of Multiple Fired Heaters. Ind. Eng. Chem. Res. 2007, 46, 5631. (30) Bandyopadhyay, S.; Malik, R. K.; Shenoy, U. V. Invariant Rectifying-Stripping Curves for Targeting Minimum Energy and Feed Location in Distillation. Comput. Chem. Eng. 1999, 23, 1109. (31) Bandyopadhyay, S. Thermal Integration of a Distillation Column through Side-exchangers. Trans. Inst. Chem. Eng.: Chem. Eng. Res. Des. 2007, 85A, 155. (32) Bandyopadhyay, S.; Mishra, M.; Shenoy, U. V. Energy-Based Targets for Multiple-Feed Distillation Columns. AIChE J. 2004, 50, 1837. (33) Bandyopadhyay, S.; Malik, R. K.; Shenoy, U. V. Feed Preconditioning Targets for Distillation through Invariant Rectifying-Stripping Curves. Ind. Eng. Chem. Res. 2003, 42, 6851. (34) Bandyopadhyay, S. Effect of Feed on Optimal Thermodynamic Performance of a Distillation Column. Chem. Eng. J. 2002, 88, 175.

ReceiVed for reView September 11, 2007 ReVised manuscript receiVed November 9, 2007 Accepted November 19, 2007 IE0712232

Water Management in Process Industries ... - ACS Publications

Recommend Documents