Cooperative Water Network Systemto Reduce Carbon FootprintS E O N G - R I N L I M † A N DJ O N G M O O N P A R K * , ‡

Department of Bioproducts and Biosystems Engineering,University of Minnesota, 1390 Eckles Avenue, St. Paul,Minnesota 55108, and Advanced EnvironmentalBiotechnology Research Center, Department of ChemicalEngineering, School of Environmental Science andEngineering, Pohang University of Science and Technology,San 31, Hyoja-dong, Pohang 790-784, South Korea

Received January 24, 2008. Revised manuscript receivedMay 18, 2008. Accepted June 12, 2008.

Much effort has been made in reducing the carbon footprintto mitigate climate change. However, water network synthesishas been focused on reducing the consumption and cost offreshwater within each industrial plant. The objective of this studyis to illustrate the necessity of the cooperation of industrialplants to reduce the total carbon footprint of their water supplysystems. A mathematical optimization model to minimizeglobal warming potentials is developed to synthesize (1) acooperative water network system (WNS) integrated over twoplants and (2) an individual WNS consisting of two WNSsseparated for each plant. The cooperative WNS is comparedto the individual WNS. The cooperation reduces their carbonfootprint and is economically feasible and profitable. Astrategy for implementing the cooperation is suggested for thefair distribution of costs and benefits. As a consequence,industrial plants should cooperate with their neighbor plantsto further reduce the carbon footprint.

1. IntroductionClimate change has been a big issue faced by sustainabledevelopment. Much effort has been made to reduce theemissions of greenhouse gases (1, 2). For instance, renewableenergy sources are being developed and used to replace fossilfuels in the electricity and transportation industries (3), andprocesses and systems are being improved to enhance theirenergy efficiency and mitigate the emissions of greenhousegases (4). The Annex I Parties in the Kyoto Protocol shouldcommit to reduce greenhouse gases by an average of 5.2%below the year 1990 levels during the 2008-2012 period (5).

The carbon footprint is used for CO2 abatement effortsas an indicator to quantify the global warming potential(GWP) of total greenhouse gases incurred throughout thelife cycle and associated supply chains. The global warmingintensity of transportation fuels has been measured with thecarbon footprint (6). New technologies, recycling, anddemand management have been analyzed to reduce thecarbon footprint of the whole copper cycle (7). The effectsof international trade and income level on the carbonfootprint of American households have been evaluated todesign the policies needed to reduce consumer impacts on

global warming (8). Companies are trying to minimize thecarbon footprint of products, systems, processes, and servicesin the stages of research and development, design, produc-tion, transportation, use, and disposal in the context ofenvironmental management. Leading retailers in the UnitedKingdom have used the carbon footprint for their business:they measure the carbon footprint of products and informand drive customers to choose a product with a small carbonfootprint, in line with environmental labeling (9). This market-driven force can accelerate the conversion of systems andprocesses used for production into environmentally friendlyones to minimize the carbon footprint of final products.

The carbon footprint of water supply systems hassignificant impacts on the carbon footprint of final products.A water supply system is an essential utility required foroperating systems and processes in almost all industrialplants. The GWPs from all water supply systems associatedwith the supply chains of a final product are accumulatedinto the carbon footprint of the final product, as shown inFigure 1. Moreover, the carbon footprint of a water supplysystem in a plant has wide impacts on those of the variousproducts produced in the plant. Therefore, it should bementioned that water supply systems are important targetsto be improved with a high priority in order to effectivelyreduce the carbon footprint of products, and the carbonfootprint of water supply systems should be minimized toproduce environmentally competitive products.

Traditional water network synthesis has focused onminimizing the freshwater consumption and associated costsof a water supply system via economic aspects. The synthesisis defined as the optimization of a water supply system byconnecting water sources (e.g., wastewater and/or treatedwastewater) to water sinks (e.g., unit processes) for waterreuse in case the properties of the water sources meet thewater quality requirement of the water sinks (10). Mostprevious studies have focused on solving mathematicaloptimization models such as nonlinear programming (NLP)and mixed-integer nonlinear programming with determin-istic approaches to find global optima minimizing freshwaterconsumption or cost, because of nonconvexities from bilinearvariables in the mass balances of the models (10–12). Geneticalgorithms have been applied as a stochastic approach tosearch global optima (13). Pinch analysis technologies havebeen studied to graphically analyze and identify the minimumfreshwater consumption flowrate of a water supply systemand then to heuristically generate a water network system(WNS) (10). A graphical method for pinch analysis has beendeveloped to target the minimum flowrate and design a WNSsimultaneously (14). Life-cycle costing has been employedfor evaluating the economic feasibility and profitability of aWNS (15), and for developing the mathematical optimizationmodel to synthesize an economically friendly WNS (16).

The paradigm of sustainable development has drivenwater network synthesis to minimize environmental impactsas well as economic costs. Multiobjective optimization hasbeen studied to minimize a total annualized cost andenvironmental impacts (17). Life-cycle assessment and life-cycle costing on a WNS have been studied to evaluate theenvironmental and economic performance of the WNS (18).However, the effort to reduce the carbon footprint of a WNShas not been performed in water network synthesis.

Industrial plants should explore neighbor plants andcooperate with one another to find more and better op-portunities to reduce the carbon footprint and economiccosts. Until now, many plants have made much effort withinthe system boundary of each individual plant to reduce

* Corresponding author phone: +82-54-279-2275; fax: +82-54-279-2699; e-mail: jmpark@postech.ac.kr.

† University of Minnesota.‡ Pohang University of Science and Technology.

Environ. Sci. Technol. 2008, 42, 6230–6236

6230 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 16, 2008 10.1021/es800243e CCC: $40.75 2008 American Chemical SocietyPublished on Web 07/19/2008

environmental impacts and economic costs using waternetwork synthesis. However, limits of further improvementhave been reached as a result of the full utilization of all ofthe possible opportunities to increase water reuse withineach plant. It should be mentioned that the opportunitiescan be increased by extending the system boundary of waternetwork synthesis and including more unit processes withvarious water qualities and flowrates. Therefore, the coop-eration of industrial plants for extending the system boundarycan break through the limits by creating new opportunitiesto reduce the carbon footprint and economic costs.

The possibility for successful cooperation through net-working can be increased by determining the fair distributionof economic and environmental costs and benefits for eachparticipating industrial plant. However, the fair distributionon a quantitative basis is impossible because a networksystem for plants is such an integrated one for interdepen-dency that the network system cannot be separated into anindividual part of each plant. The difference between thecosts and benefits of plants from the cooperation can inducethe dissatisfaction of some plants with more costs or lessbenefits. These problems can cause a delay and evennullification of the cooperation because of the failure ofmutual agreement for the distribution of economic andenvironmental costs and benefits. Therefore, a solution ofthese problems is required for the practical implementationof the cooperative WNS to reduce the total carbon footprintof plants.

The objective of this study is to illustrate that thecooperation of industrial plants is needed to reduce the totalcarbon footprint of their water supply systems. Two plantsare employed to demonstrate the effect of the cooperationon the carbon footprint and economic cost of their watersupply systems. A mathematical optimization model tominimize the GWP of a WNS is developed to synthesizeindividual and cooperative WNSs: the individual WNSconsists of two WNSs separated for each plant, and thecooperative WNS is one WNS integrated with the unitprocesses in the two plants. The configuration, designcharacteristics, the GWPs of principal contributors, and thecarbon footprint of the cooperative WNS are compared tothose of the individual WNS. The economic performance ofthe cooperative WNS is evaluated to examine whether the

cooperative WNS is economically feasible and profitable. Astrategy for practically implementing the cooperative WNSis suggested as a solution for the fair distribution of the costsand benefits incurred in constructing and managing thecooperative WNS.

2. Methods2.1. Mathematical Optimization Model. The mathematicaloptimization model to minimize the carbon footprint of aWNS was developed by formulating the GWPs of principalcontributors to its carbon footprint and the mass balancesand constraints based on a superstructure model.

2.1.1. Superstructure Model. A superstructure model wasused to develop the mathematical optimization model tominimize the carbon footprint of a WNS, as shown in Figure2. The superstructure model is used to take into account allof the opportunities affecting the value of the objectivefunction and obtain all of the feasible solutions of themathematical model. This superstructure model includesall possible interconnections between water sources andsinks, such as those from the outlet of a unit process to theinlet of the others, as well as between freshwater sources andunit processes, to fully utilize opportunities of water reuse.However, local recycling from the outlet to the inlet withina unit process was not allowed to avoid excessive electricityconsumption and costs derived from pumping with a highflowrate (15). It was assumed that a mixer combines manystreams into a single stream and that a splitter divides onestream into all possible streams to water sinks.

2.1.2. Principal Contributors and Their Unit GlobalWarming Potentials. The GWPs of the principal contributorsto the carbon footprint of a WNS were formulated in theobjective function of the mathematical optimization modelto minimize its carbon footprint. The principal contributorsare the primary targets, so-called “hot spots”, and are focusedon to effectively minimize the carbon footprint. The em-ployment of principal contributors simplifies a mathematicaloptimization model by excluding minor contributors, whichmakes it easy to apply the model to the real situations ofindustrial plants. The consumptions of freshwater andelectricity were selected as the principal contributors for themodel with respect to the results of a life-cycle assessment(LCA) on a WNS: the proportion of the GWP of freshwater

FIGURE 1. Effect of greenhouse gases (GHG) from water supply systems existing throughout the supply chains on the carbonfootprint of a final product.

VOL. 42, NO. 16, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 6231

consumption to the total GWP of the WNS was found to be68%, and that of electricity consumption was 28% (18).

The unit GWPs of the principal contributors were evalu-ated on the basis of LCA databases to be incorporated intothe mathematical optimization model. The LCA data takeinto account the GWP associated with the supply chains andlife-cycle stages of a principal contributor. The reference flowsof the principal contributors were set to 1 m3 of industrialwater, and 1 kW ·h of electricity in order to calculate theirunit GWPs. The unit GWP of the consumption of industrialwater was 0.302 kg CO2-eq/m3 according to the Ecoinventv.1.2 database (19). These data include the GWP associatedwith infrastructure and energy use for water treatment andtransportation to the end user, as shown in Figure 3a. Theunit GWP of electricity consumption was 0.495 kg CO2-eq/kW ·h according to the LCA results from the PASS (20). Thisdata includes the extraction of raw materials, transportation,electricity generation, and waste treatment, as shown inFigure 3b, and the electricity generation mix consists of 43.3%nuclear, 37.5% coal, 12.3% natural gas, 5.1% oil, and 1.8%hydro.

2.1.3. Mathematical Formulation. The mathematical op-timization model consists of (1) an objective functiongenerated by formulating the total GWP of the principalcontributors and (2) mass balances and constraints neededto represent the superstructure model and real situations inindustrial plants. It should be mentioned that the GWPsassociated with wastewater from a WNS were not taken into

account in this model because the total contaminant loadsin wastewater are not changed through water networksynthesis. In other words, the GWPs from wastewater wereregarded as a baseline for this model.

The objective function is the sum of the GWPs of theprincipal contributors, which is minimized to generate a WNSwith the smallest carbon footprint. The carbon footprint ofa WNS in the model is expressed as the amount of all of theGWPs of the principal contributors generated per hour. Thecarbon footprint of the objective function is minimized byoptimizing the tradeoff between the GWPs from the con-sumption of industrial water and from electricity consump-tion. The GWP of the consumption of industrial water iscalculated from its flowrate and unit GWP. The GWP ofelectricity consumption for pumping is estimated from itsunit GWP and the power requirement calculated from theflowrate of industrial water and the pressure requirement.The pressure is the sum of the head loss through the pipelineand the additional head required to meet water pressure atthe end of the pipeline. The head loss is calculated by usingthe Darcy-Weisbach equation (21). The optimal velocity inthe pipeline is assumed to be proportional to the flowrateto simplify the mathematical model (12). The mass balancesand constraints are formulated on the basis of the super-structure model. The detailed formulation of the math-ematical optimization model is presented in the SupportingInformation.

FIGURE 2. Superstructure model used to synthesize a cooperative water network system (WNS).

FIGURE 3. Concept of life-cycle impact assessment to evaluate the total amount of greenhouse gases (GHGs) associated with areference flow of water and electricity: GHGs are emitted throughout the supply chains of water and electricity.

6232 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 16, 2008

2.2. Water Network Synthesis. Five processes of twoindustrial plants were employed as water sources and sinksfor water network synthesis: three processes for cleaning,cooling, and brushing are used for plant “A” producinggalvanized and aluminized steel sheets, and two cleaningprocesses for copper-coated and tin-coated plates are usedfor plant “B” producing electrolytic steel plates used inmanufacturing bottles and cans. The process limiting datafor water network synthesis are presented in the SupportingInformation, showing operational conditions required foreach unit process. Chemical oxygen demand (COD) andsuspended solid (SS) were used as water quality indicatorsbecause these concentrations represent the amount oforganic contaminants and particles. The distance matrixbetween the water sources and sinks is presented in theSupporting Information, which was used to estimate thelength of pipelines.

The individual and cooperative WNSs were synthesizedusing the mathematical optimization model to minimize thecarbon footprint. The individual WNS consisted of two WNSs,which were individually synthesized by using the unitprocesses within each plant: WNS “A” was networked withthe three processes within plant “A”, and WNS “B” with thetwo processes within plant “B”. The cooperative WNS wassynthesized with the five processes in the two plants toincrease the opportunities of water reuse. The WNSs weregenerated from the optimal solutions to the mathematicaloptimization model. GAMS/MINOS (22) was used as an NLPsolver to find the optimal solutions. The configurations ofthe individual and cooperative WNSs were embodied fromtheir own optimal solutions.

2.3. Estimation of Carbon Footprint and EconomicPerformance. The carbon footprints of the individual andcooperative WNSs were estimated to examine which WNSwas more environmentally friendly in the context of the

prevention of climate change. The GWPs of the principalcontributors during the lifetime were evaluated using equa-tions in the mathematical optimization model. The lifetimewas assumed to be 15 years with respect to the lifetime ofpipelines. The carbon footprint of the individual andcooperative WNSs during the lifetime was calculated bysumming the GWPs of the principal contributors.

The economic performances of the individual and co-operative WNSs were estimated to examine whether thecooperative WNS was economically feasible and profitable.The costs of construction, operations and maintenance, anddisposal of the two WNSs were estimated, and then theincremental costs and benefits of the cooperative WNS wereobtained by subtracting the costs of the cooperative WNSfrom the costs of the individual WNS (15). The net presentvalue of the cooperative WNS was evaluated from theincremental costs and benefits to measure its economicfeasibility, and its payback period and internal rate of returnwere evaluated to measure its economic profitability. Thedetailed equations related to the economic evaluation arepresented in the Supporting Information.

3. Results and Discussion

The configuration, design characteristics, carbon footprint,and economic costs of the cooperative WNS were comparedto those of the individual WNS, and the economic feasibilityand profitability of the cooperative WNS was evaluated, inorder to estimate costs and benefits incurred from thecooperation of the two industrial plants.

3.1. Configurations and Design Characteristics. Thecooperative WNS utilized more opportunities of water reusethan the individual WNS. The configurations of the individualand cooperative WNSs are shown in Figure 4. The totalnumber of interconnections for water reuse in the cooperative

FIGURE 4. Configurations of the individual and cooperative water network systems (WNSs): (a) individual WNS, (b) cooperative WNS(IW, industrial water; WW, wastewater; units, m3/h).

VOL. 42, NO. 16, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 6233

WNS was 40% greater than that in the individual WNS. Itshould be mentioned that the interconnections between plant“A” and plant “B” were generated in the cooperative WNSby creating more opportunities for water reuse, even thoughthe number of the interconnections within plant “A” or plant“B” in the cooperative WNS was less than that in the individualWNS. Pumps for the interconnections between plant “A”and plant “B” were required to be operated in one plant toreduce the consumption of industrial water in the other plant.

The cooperative WNS decreased the consumption ofindustrial water but increased the power requirement forpumping and the total pipe length, when compared to theindividual WNS. The design characteristics of the individualand cooperative WNSs are summarized in Table 1. The totalflowrate of industrial water consumed in the cooperativeWNS was 7.4% less than that in the individual WNS becausethe cooperation of the two plants increased the opportunitiesof water reuse and the flowrate of reused water. Theconsumption of industrial water in plant “A” was decreasedby 51.3%, and that in plant “B” was decreased by 1.3%: thecooperation induced higher benefits in plant “A” than inplant “B”. The total wastewater generation rate in thecooperative WNS was 8.0% less than that in the individualWNS because the consumption of industrial water in thecooperative WNS was less than that in the individual WNS.However, the flowrate and the COD and SS loads ofwastewater in plant “A” were increased by 37.9%, 28.6% and30.8%, respectively, as a result of the cooperation, while thosein plant “B” were decreased by 13.9%, 10.8%, and 13.3%,respectively. The interconnections from plant “B” to plant“A” in the cooperative WNS increased the flowrate andcontaminant loads of wastewater treated in plant “A” butdecreased those treated in plant “B”: the variation of theflowrate and contaminant loads treated in each plantengendered another issue on the fair distribution of costsand benefits incurred from the cooperation of the two plants.The pipe length of the cooperative WNS was 114.8% greaterthan that of the individual WNS because of the intercon-nections between the two plants for water reuse. The powerrequirement for pumping in the cooperative WNS was 3.9%greater than that in the individual WNS because the flowrateof reused water and the pipe length related to head losseswere increased as a result of the cooperation, even thoughthe consumption of industrial water was decreased. It shouldbe mentioned that the costs of pipeline construction andelectricity for pumping cannot be easily distributed to thetwo plants because of the interdependency of the two plants.

In other words, the interconnections were shared by the twoplants, and all of the components of the cooperative WNSwere integrated for the cooperation. Therefore, the fairdistribution of the costs and benefits to the two plants isimportant for successful construction, operation and main-tenance, and disposal of the cooperative WNS. A strategy forthe fair distribution is mentioned later, to take into accountthe practical implementation of the cooperative WNS.

3.2. Carbon Footprint. The cooperation of the two plantsreduced the total carbon footprint of their water supplysystems. Figure 5 shows the GWPs of the principal contribu-tors and the carbon footprint of the individual and coopera-tive WNSs incurred during their lifetimes (15 years). Thecarbon footprint of the cooperative WNS was 6.3% less thanthat of the individual WNS. The decrease of the GWP fromthe reduction of industrial water outweighed the increase ofthe GWP from the increase of electricity consumption,when the cooperative WNS was compared to the individualWNS: the GWP from the consumption of industrial water inthe cooperative WNS was 7.4% less than that in the individualWNS, while the GWP from the electricity consumption in thecooperative WNS was 4.0% greater. This means that thecooperation of industrial plants to construct a cooperative

TABLE 1. Design Characteristics of the Individual and Cooperative Water Network Systems (WNSs)

units individual WNS cooperative WNS

industrial water consumption flowrateplant “A” m3/h 7.6 3.7plant “B” 54.7 54.0total 62.3 57.7

wastewater generation

flowrateplant “A” m3/h 6.6 9.1plant “B” 51.0 43.9total 57.6 53.0

COD loadplant “A” kg/h 1.41 1.82plant “B” 3.71 3.30total 5.12 5.12

SS loadplant “A” kg/h 1.27 1.67plant “B” 3.04 2.64total 4.31 4.31

pipeline construction length total m 1350 2900number of interconnectionsfor water reuse

intraplant plant “A” 3 0plant “B” 2 2

interplant plants “A” and “B” 0 5

total 5 7electricity consumption power requirement total kW 4.11 4.27

FIGURE 5. Global warming potentials (GWPs) of the principalcontributors, that is, the consumptions of industrial water andelectricity, and carbon footprint, that is, the sum of the GWPsof the principal contributors, of the water network systems(WNSs) during their lifetimes (15 years).

6234 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 16, 2008

WNS can be extended until the increase of GWP fromelectricity consumption for water reuse equals the decreaseof GWP from the reduction of industrial water, in terms ofthe system boundary of a cooperative WNS.

3.3. Economic Evaluation of the Cooperative WaterNetwork System. The cooperative WNS was more economi-cally feasible and profitable than the individual WNS. Figure6 shows the economic evaluation of the cooperative WNSs:the cost comparison between the individual and cooperativeWNSs in their life cycle stages, the incremental costs andbenefits of the cooperative WNS, and the economic feasibilityand profitability of the cooperative WNS. The cooperationof the two plants decreased the annual operations andmaintenance cost by 6.9% because of the reduction ofindustrial water but increased the construction and disposalcost by 101.5% and 102.8%, respectively, because of theinterconnections between plant “A” and plant “B”, when thecooperative WNS was compared to the individual WNS. Theincremental benefits of the cooperative WNS were incurredfrom the reduction of the operations and maintenance cost,and its incremental costs were incurred from the increase ofthe construction and disposal costs. The net present valueof the cooperative WNS was 132,000 USD during its lifetime,which shows that the cooperative WNS is economicallyfeasible. The payback period was 3 years, and the internalrate of return was 42.3%, which shows that the cooperativeWNS is highly profitable. This means that the cooperationof industrial plants to reduce the carbon footprint cancontribute to an increase in their economic benefits, whichis in line with the concept of environmental management.

3.4. Strategy for Practical Implementation. The fairdistribution of costs and benefits plays a significant role inthe practical implementation of the cooperative WNS: thecosts and benefits include the economic costs and benefitsfrom its construction, operations and maintenance, anddisposal, and the environmental costs and benefits such asthe decrease of the carbon footprint. However, the costs andbenefits cannot be easily distributed into the two plantsbecause the cooperative WNS is integrated with the unit

processes in the two plants and so cannot be decomposedto parts of each plant on a quantitative basis.

A strategy for the practical implementation of thecooperative WNS is to establish a joint venture companyinvested by the two plants for the fair distribution of costsand benefits, in order to deal with the cooperative WNS asa unit to construct, operate, and manage the system, and totreat wastewater from the system. The costs that would beneeded for the individual WNS can be invested in the jointventure company, and the additional costs required for thecooperative WNS can be invested from the two plants withthe aim to obtain economic and environmental benefitsincurred from the cooperative WNS. All benefits from thejoint venture company can be fairly distributed to the twoplants on a quantitative basis of the share of their additionalinvestments. Wastewater from the two plants in the coop-erative WNS can be treated by the joint venture company,irrespective of the origins of wastewater. The costs that wouldbe needed to treat wastewater in the case of the individualWNS can be periodically paid to the joint venture company.Increased or reduced costs for wastewater treatment in thecooperative WNS affect the profit of the joint venturecompany, which will be returned to the two plants.

AcknowledgmentsThis work was supported by the Korea Research FoundationGrant (KRF-2007-357-D00150) funded by the Korean Gov-ernment (MOEHRD). This work was also supported in partby the Korean Science and Engineering Foundation (R11-2003-006) through the Advanced Environmental Biotech-nology Research Center at Pohang University of Science andTechnology and in part by the program for advancededucation of chemical engineers (second stage of BK21).

Supporting Information AvailableFormulation of the mathematical optimization model, equa-tions for economic evaluation and their related nomenclature,process limiting data, distance matrix, and freshwater

FIGURE 6. Economic evaluation of the cooperative water network system (WNS): (a) construction, operations and maintenance, anddisposal costs of the WNSs; (b) incremental costs and benefits of the cooperative WNS; (c) net present value (NPV) and paybackperiod of the cooperative WNS; and (d) internal rate of return (IRR) of the cooperative WNS.

VOL. 42, NO. 16, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 6235

concentrations. These materials are available free of chargevia the Internet at http://pubs.acs.org.

Literature Cited(1) Fiaschi, D.; Carta, R. CO2 abatement by co-firing of natural gas

and biomass-derived gas in a gas turbine. Energy 2007, 32, 549–567.

(2) Otto, V.; Reilly, J. Directed technical change and the adoptionof CO2 abatement technology: The case of CO2 capture andstorage; Energy Econ. In press.

(3) Goyal, H. B.; Seal, D.; Saxena, R. C. Bio-fuels from thermo-chemical conversion of renewable resources: A review. Renew-able Sustainable Energy Rev. 2008, 12, 504–517.

(4) Blesl, M.; Das, A.; Fahl, U.; Remme, U. Role of energy efficiencystandards in reducing CO2 emissions in Germany: An assessmentwith TIMES. Energy Policy 2007, 35, 772–785.

(5) Chen, W. Carbon quota price and CDM potentials afterMarrakesh. Energy Policy 2003, 31, 709–719.

(6) Christen, K. The carbon footprint of transportation fuels.Environ. Sci. Technol. 2007, 41, 6636–6636.

(7) Giurco, D.; Petrie, J. G. Strategies for reducing the carbonfootprint of copper: New technologies, more recycling ordemand management. Miner. Eng. 2007, 20, 842–853.

(8) Weber, C.; Matthews, H. S. Quantifying the global and distri-butional aspects of American household carbon footprint. Ecol.Econ. , In Press. DOI: 10.1016/j.ecolecon.2007.09.021.

(9) Senior, K. UK business embraces carbon footprint reduction.Fron. Ecol. Environ. 2007, 5, 288–288.

(10) Mann, J. G.; Liu, Y. A. Industrial Water Reuse and WastewaterMinimization; McGraw-Hill: New York, 1999.

(11) Bagajewicz, M. A review of recent design procedures for waternetworks in refineries and process plants. Comput. Chem. Eng.

2000, 24, 2093–2113.(12) Gunaratnam, M.; Alva-Argaez, A.; Kokossis, A.; Kim, J.-K.; Smith,

R. Automated design of total water systems. Ind. Eng. Chem.Res. 2005, 44, 588–599.

(13) Prackotpol, D.; Srinophakun, T. GAPinch: genetic algorithmtoolbox for water pinch technology. Chem. Eng. Process. 2004,43, 203–217.

(14) Alwi, S. R. W.; Manan, Z. A. Generic graphical technique forsimultaneous targeting and design of water networks. Ind. Eng.Chem. Res. 2008, 47, 2762–2777.

(15) Lim, S.-R.; Park, D.; Lee, D. S.; Park, J. M. Economic evaluationof a water network system through the net present value methodbased on cost and benefit estimations. Ind. Eng. Chem. Res.2006, 45, 7710–7718.

(16) Lim, S.-R.; Park, D.; Park, J. M. Synthesis of an EconomicallyFriendly Water Network System by Maximizing Net PresentValue. Ind. Eng. Chem. Res. 2007, 46, 6936–6943.

(17) Erol, P.; Thoming, J. ECO-design of reuse and recycling networksby multi-objective optimization. J. Clean. Prod. 2005, 13, 1492–1503.

(18) Lim, S.-R.; Park, J. M. Environmental and Economic Analysis ofa Water Network System using LCA and LCC. AIChE J. 2007, 53,3253–3262.

(19) Ecoinvent data, version 1.2; Swiss Center for Life CycleInventories: Dubendorf, Switzerland, 2005.

(20) PASS, version 3.1.3; Korea Accreditation Board: Seoul, Republicof Korea, 2006.

(21) McGhee, T. J. Water Supply and Sewerage; McGraw-Hill: NewYork, 1991.

(22) GAMS, A User Guide; GAMS Development Corporation: Wash-ington DC, 2005.

ES800243E

6236 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 16, 2008