Article

pubs.acs.org/IECR

Eco-Design of a Wastewater Treatment System Based

on Process Integration
Donghee Park,† Dae Sung Lee,*,‡ and Seong-Rin Lim*,§
†
Department of Environmental Engineering, Yonsei University, Wonju 220-710, South Korea
‡
Department of Environmental Engineering, Kyungpook National University, Daegu, 702-701, South Korea
§
Department of Environmental Engineering, College of Engineering, Kangwon National University, Chuncheon, Gangwon, 200-701,
South Korea
*
S Supporting Information

ABSTRACT: Because almost all industrial plants have wastewater treatment systems, eco-design of the systems is an effective
way to reduce environmental impacts and economic costs of industry sectors. The eco-design using biobjective optimization has,
however, a limitation due to the subjective weighting on the two objectives. The objective of this study is to eco-design existing
wastewater treatment systems based on process integration by converting biobjective to single objective problems. For the
mathematical optimization model, an objective function is formulated by monetizing environmental impacts to external costs and
summing the external and economic costs. Mass balances and constraints are formulated to reflect the superstructure model and
real situations. Two case studies are performed to verify the developed model. The eco-design outcomes are compared to their
respective economic cost- and environmental impact-minimized designs. This comparison shows that the developed model
optimizes the trade-offs between the biobjectives. This study can be applied to reduce environmental impacts and economic costs
of other process systems.

1. INTRODUCTION an increase in other ones in other life cycle stages.11 Therefore,

Because wastewater treatment systems consume considerable eco-design should be employed to optimize the trade-offs among
amounts of electricity and chemicals, the systems are one of the all impacts and costs in all life cycle stages.
principal targets to be focused on to reduce economic costs in Although eco-design frequently employs biobjective optimi-
industrial plants. For economical design of the systems, many zation to reduce environmental impacts and economic costs,12,13
studies have been performed to integrate wastewater treatment this approach has a limitation due to the decision makers’ sub-
systems distributed in industrial plants.1−5 This integration jective weighting on each objective in the objective function.
approach is to minimize the flow rate of the wastewater treated in Depending on the preference of decision makers, different optimum
distributed systems and to maximize the flow rate of the waste- solutions can be obtained from a biobjective optimization problem.
water bypassing the distributed systems. This can reduce capital As an approach to overcome this limitation, it should be mentioned
investment and operating costs for wastewater treatment because that, if environmental impacts and economic costs are in a
these costs are proportional to the flow rate of the wastewater unidimensional unit, this enables free trade-offs between the two
treated in distributed systems.1,2,6 Therefore, the integration is a objectives and thus generates a single optimum solution, irrespec-
cost-effective strategy to comprehensively restructure existing tive of who decision makers are. Therefore, eco-design needs to seek
wastewater treatment systems. a way to monetize environmental impacts to economic costs.
In addition to the economic aspects of wastewater treatment The objective of this study is to eco-design existing wastewater
systems, their environmental impacts need to be reduced to treatment systems based on process integration by monetizing
increase the sustainability of industrial plants. Since the systems environmental impacts to economic costs in biobjective optimi-
directly and indirectly incur environmental impacts,6,7 the systems zation problems. For the eco-designs, a mathematical optimiza-
should be optimized and improved to proactively prevent environ- tion model is developed to integrate distributed and terminal
mental pollution. For instance, since electricity and chemicals used wastewater treatment systems in an industrial plant and to sim-
in the systems are produced by generating greenhouse gases and air ultaneously minimize environmental impacts and economic
pollutants in power plants, the systems need to be designed and costs. Environmental impacts are monetized by using the EPS
operated to reduce the electricity and chemicals. Therefore, existing 2000 methodology.12 The model is applied for two case studies.
wastewater treatment systems should be restructured to enhance In each case study, an eco-design outcome is compared to two
environmental and economic performance at the same time. other design alternatives (environmental impact-minimized and
Eco-design can be used to improve environmental and eco- economic cost-minimized designs) to examine the characteristics
nomic performance of existing wastewater treatment systems.8,9
In general, eco-design takes into account life cycle stages and Received: July 17, 2012
supply chains associated with systems in reducing environmental Revised: October 30, 2012
impacts and economic costs.10 This is because a design can lead Accepted: January 11, 2013
to a decrease in an impact/cost in a stage but simultaneously to Published: January 11, 2013

© 2013 American Chemical Society 2379 dx.doi.org/10.1021/ie3018934 | Ind. Eng. Chem. Res. 2013, 52, 2379−2388
Industrial & Engineering Chemistry Research Article

Figure 1. Superstructure model used to integrate distributed and terminal wastewater treatment systems (WW: wastewater, DS: distributed system, TS:
terminal system).

of the eco-design. The developed model can be widely used in 2.3. Mathematical Optimization Model. Based on the
other industrial plants to obtain a single optimum solution for a superstructure model, a mixed integer nonlinear programming
generally acceptable eco-design. (MINLP) model was developed by formulating environmental
impacts and economic costs. This model includes (i) binary variables
2. METHODS to formulate whether a stream exists or whether a wastewater
The eco-designs were performed with existing distributed and treatment system is operated and (ii) bilinear variables to for-
terminal wastewater treatment systems in an industrial plant (for mulate the mass balances based on the superstructure model.
case study I, five distributed and one terminal systems; and for case 2.3.1. Monetization of Environmental Impacts. This
study II, seven distributed and one terminal systems). The systems monetization was performed by using the EPS 2000 methodology12
were integrated based on a mathematical optimization model that to evaluate external costs of principal contributors. The EPS
was developed by using a superstructure model and by monetizing methodology is based on the willingness-to-pay method to valuate
environmental impacts to economic costs, so-called external environmental impacts in a monetary value (i.e., EUR).15,16 This is
costs.13−16 The solution to the developed mathematical model generally used in environmental economics. The monetization was
was schematically embodied into an eco-design outcome. This eco- performed by using the Gabi 4.0 software17 in order to evaluate the
design was compared to the designs generated by minimizing unit external cost for piping, electricity consumption, and pipe
environmental impacts (hereinafter, this design is called “envi- recycling. In this study, the principal contributors are (i) piping in
design”) and economic costs (hereinafter, “cost-design”), respectively. the construction stage, (ii) electricity for pumping and for
2.1. Data. Data used for two case studies are presented in the wastewater treatment, and maintenance and repair (M&R) in the
Supporting Information (Tables S1−S6): characteristics of waste- operation and maintenance (O&M) stage, and (iii) pipe scrap
water sources; design and operational conditions of distributed and recycling in the disposal stage. The external costs are not discounted
terminal systems; and distance matrix between the wastewater to present values to take into account future generations.18
sources and sinks. The raw wastewaters are primarily treated in the 2.3.2. Economic Costs. The economic costs consist of (i) pipe
distributed systems and then secondarily in the terminal system. material cost, labor cost, construction expense, overhead, and
These systems consist of physical and chemical treatment processes profit in the construction stage, (ii) electricity consumption for
(i.e., coagulation, flocculation, and sedimentation). The required pumping and for wastewater treatment, labor, and M&R in the
discharge quality of the terminal system was set at 20 mg/L for O&M stage, and (iii) decommissioning cost, construction expense,
CODcr, 5 mg/L for SS, and 8 mg/L for F−: it is assumed that overhead, and profit in the disposal stage. These costs are dis-
the treated wastewater meeting these requirements has negligible counted to present values by using interest and escalation rates.
environmental impact potential. 2.3.3. Formulation. For the eco-design, the objective function
2.2. Superstructure Model. Figure 1 shows the super- of the mathematical model is formulated as the sum of the
structure model used to integrate the distributed and terminal external costs and economic costs of the integrated wastewater
systems. This model takes into account all possible intercon- treatment system. Mass balances are formulated based on the
nections:5,6 from wastewater sources to the inlets of distributed superstructure model: overall integrated system, wastewater source
and terminal systems; and from the outlets of distributed systems splitters; and mixers and splitters of distributed and terminal
to their inlets and to the inlets of terminal systems. The mixer systems. Also, constraints are formulated to take into account real
nodes imply the combination of all possible streams into a stream, situations in industrial plants: flow rates and loads of distributed
and the splitter nodes imply the division of a given stream into all and terminal systems, flow rates of streams, and concentrations
possible streams. of the discharge quality in terminal systems.
2380 dx.doi.org/10.1021/ie3018934 | Ind. Eng. Chem. Res. 2013, 52, 2379−2388
Industrial & Engineering Chemistry Research Article

Figure 2. Configurations of the three designs for caste study I: (a) envi-design; (b) cost-design; and (c) eco-design (WW: wastewater, DS: distributed
system, TS: terminal system).

2.3.4. Mathematical Optimization Models for Comparative Minimize ECO = ENVI + COST (1)
Alternatives. For the envi-design and cost-design, the objective ENVI is the sum of external costs incurred in the stages of con-
functions were formulated with the total external cost and eco- struction, O&M, and disposal.
nomic cost of the integrated system, respectively.
2.4. MINLP Solver. GAMS19 is used to obtain the optimal solu- t

tions to the three models: CPLEX for linear programming, MINOS ENVI = EXCcon + ∑ EXCOt &M + EXCdis
for nonlinear programming (NLP), and DICOPT for MINLP (note t=1 (2)
that these NLP and MINLP solvers do not always guarantee global For the construction stage, external cost is derived from piping,
optimum solutions). The optimum solutions are embodied to the which is estimated with a regression equation from the relation-
schematic configurations of the three designs, respectively. ship between the cross-sectional area of the pipe and the unit
external cost of piping. The cross-sectional area is determined by
3. MATHEMATICAL OPTIMIZATION MODEL FOR THE an optimum velocity, which varies depending on the flow rate.20
ECO-DESIGN
3.1. Objective Function. The objective function for the eco-
EXCcon = ∑ ∑ EPi ,j
design is to minimize ECO, which is the sum of the total external i j (3)
cost (hereinafter called “ENVI”) and total economic cost (hereinafter
EPi , j = (apeAi , j + bpeli , jBi , j )lijBij
called “COST”) of the integrated wastewater treatment system. (4)

2381 dx.doi.org/10.1021/ie3018934 | Ind. Eng. Chem. Res. 2013, 52, 2379−2388

Industrial & Engineering Chemistry Research Article

Fi , j Table 1. Quantitative Characteristics of the Envi-Design,

Ai , j = Cost-Design, and Eco-Design for Case Study I
vi , j (5)
envi- cost- eco-
vi , j = aopFi , j + bop item unit design design design
(6)
pipe length m 1,650 750 1,410
For the O&M stage, external costs are incurred from electricity distributed number (in use) - 2 2 2
consumption for pumping and for wastewater treatment, and systems flow rate of bypassing m3/h 146.0 85.2 144.4
M&R. wastewater
terminal flow rate of treated m3/h 178.9 178.9 178.9
EXCOt & M = EEpt + EEwt
t
+ EMRC t (7) systems wastewater
power electricity (pumping) kW 19.6 24.0 19.8
The external cost of electricity consumption for pumping is requirement electricity (wastewater kW 2.7 2.7 2.7
estimated by calculating pumping power based on the Darcy− treatment)
Weisbach equation.21
EEpt = (∑ ∑ Pi , j)UEetop Table 2. External Costs Due to Environmental Impacts from
i j (8) the Envi-Design, Cost-Design, and Eco-Design for Case Study
I (Unit: EUR in Thousands)
ρFi , jg (HLi , j + Ha) 1
Pi , j = envi- eco-
ηpηm 1000 (9)
life cycle item design cost-design design
construction piping 1,848 953 1,638

li , j ⎛ π ⎞ vi , j
0.5 2 operations & electricity for pumping 43,872 53,605 44,221
maintenance during a year
HLi , j = f ⎜ ⎟ Bi , j
g ⎜⎝ Ai , j ⎟⎠ 4
(O&M) electricity for 13,403 13,403 13,403
(10) wastewater treatment
during a year
The external cost of electricity consumption for wastewater treat- maintenance & repair 55 29 49
ment is estimated based on the power requirement of agitators, (M&R) during a year
flocculators, and sludge scrapers in the distributed and terminal subtotal during service 859,950 1,005,540 865,094
time (15 years)
systems.
disposal pipe recycling 2,693 1,389 2,387
t
EEwt = ∑ PsysBsysUEetop total during the life cycle 864,491 1,007,882 869,119
sys (11)
The external cost of M&R is assumed to be proportional to that equation from the relationship between the cross-sectional area
of the piping in the construction stage. of the pipe and the unit economic costs.
EMRC t = δEXCcon (12) TDPpiping = ∑ ∑ DPi ,j
For the disposal stage, the external cost is estimated based on i j (17)
the recycling of pipelines, which is calculated by using a DPi , j = (adpAi , j + bdp)li , jBi , j (18)
regression equation from the relationship between the cross-
sectional area of the pipeline and the unit external cost of the
recycling.
TDLpiping = ∑ ∑ DLi ,j
i j (19)
EXCdis = ∑ ∑ EDi ,j DLi , j = (adlAi , j + bdl)li , jBi , j
i j (13) (20)
The construction expenses are assumed to be proportional to the
EDi , j = (adeAi , j + bde)li , jBi , j (14) sum of the material and labor costs.
For economic costs, COST is the sum of discounted economic EXPpiping = α(TDPpiping + TDLpiping ) (21)
costs in the stages of construction, O&M, and disposal.
The overheads are assumed to be proportional to the sum of the
⎛ t
COt & M(1 + e)t C t (1 + e)t ⎞ material and labor costs, and the construction expenses.
COST = ⎜⎜Ccon + ∑ + dis ⎟⎟CCF
⎝ t=1
(1 + i)t (1 + i)t ⎠ OHpiping = β(TDPpiping + TDLpiping + EXPpiping ) (22)
(15)
The profits are assumed to be proportional to the sum of the
The construction cost consists of pipe material costs, labor costs, labor cost, construction expenses, and overheads.
construction expenses, and the contractor’s overheads and profits.
PROpiping = γ(TDLpiping + EXPpiping + OHpiping ) (23)
Ccon = TDPpiping + TDLpiping + EXPpiping + OHpiping
For the O&M stage, the economic costs of electricity con-
+ PROpiping sumption and labor are calculated based on their unit economic
costs. The M&R cost is assumed to be proportional to the
(16) construction cost for piping.
The pipe material and labor costs are estimated by using their
COt & M = ECpt + ECwt
t
+ LC t + MRC t (24)
unit economic costs, which are calculated by using the regression
2382 dx.doi.org/10.1021/ie3018934 | Ind. Eng. Chem. Res. 2013, 52, 2379−2388
Industrial & Engineering Chemistry Research Article

Table 3. Economic Costs of the Envi-Design, Cost-Design, and Eco-Design for Case Study I (Unit: EUR in Thousands)
life cycle stage item envi-design cost-design eco-design
construction piping 38,065 21,296 34,593
construction expenses 7,613 4,259 6,919
overhead 2,284 1,278 2,076
contractor’s profit 4,102 2,267 3,714
subtotal 52,065 29,100 47,301
operations & maintenance (O&M) electricity for pumping during a year 9,319 11,386 9,393
electricity for wastewater treatment during a year 2,847 2,847 2,847
labor during a year 426,650 426,650 426,650
maintenance & repair (M&R) during a year 1,562 873 1,419
subtotal during service time (15 years, discounted) 5,404,115 5,421,031 5,403,272
disposal pipe decommissioning 12,451 6,853 11,257
construction expenses 2,490 1,371 2,251
overhead 747 411 675
contractor’s profit 1,569 863 1,418
subtotal (discounted) 11,706 6,442 10,583
total 5,467,886 5,456,573 5,461,156

ECpt = ∑ ∑ Pi ,jUCetop Table 4. Total Economic and External Costs of the

i j (25) Envi-Design, Cost-Design, and Eco-Design for Caste
Study I (Unit: EUR in Thousands)
t
ECwt = ∑ PsysBsysUCetop life cycle stage envi-design cost-design eco-design
sys (26) construction 53,913 30,053 48,939
t operations & maintenance (O&M) 6,264,065 6,426,570 6,268,366
LC = ∑ UCl ,sysBsys disposal 14,399 7,831 12,970
sys (27) total 6,332,378 6,464,455 6,330,275

MRC t = δCcon (28)

For the disposal stage, the economic cost consists of pipe de- For the contaminant mass balances of the distributed and
commissioning costs, expenses, overheads, and profits. terminal systems
t FC
i c , i(1 − R c , sys) − FjCc , j = 0 (37)
Cdis = TDLdecom + EXPdecom + OHdecom + PROdecom (29)
The pipe decommissioning cost is assumed to be proportional to For the constraints of flow rates and loads on the distributed and
the labor cost for piping. terminal systems
max
TDLdecom = εTDLpiping 0 ≤ Fsys ≤ Fsys Bsys (38)
(30)
max
The expenses, overheads, and profits in the disposal stage are FsysCc , sys − Lsys ≤0 (39)
calculated as in the construction stage.
For the constraints of flow rates on all the streams
EXPdecom = αTDLdecom (31)
Fi , j − Fimax
, j Bi , j ≤ 0 (40)
OHdecom = β(TDLdecom + EXPdecom) (32)
Fi , j − Fimin
, j Bi , j ≥ 0 (41)
PROdecom = γ(TDLdecom + EXPdecom + OHdecom) (33)
For the constraint of concentrations on the discharge quality of
The pipe decommissioning cost for the existing system is not the terminal system required to meet the discharge limits of
taken into account in the formulation because all existing pipes environmental regulations
are assumed to be at end-of-life: the cost is the same for the three Cc , tsout − Ccmax
, tsout ≤ 0 (42)
designs and thus does not affect the total cost of a new integrated
system. The eco-design was performed by using eqs 1−42. For the
3.2. Mass Balances. For the mass balances of the splitters comparison, the envi-design and cost-design were generated by
using eqs 2−14 and 34−42 and eqs 5−6 and 15−42, respectively.
Fspl = ∑ Fspl ,j
j (34) 4. RESULTS AND DISCUSSION
For the mass balances of the mixers 4.1. Case Study I. Figure 2 shows the configurations of the
envi-design, cost-design, and eco-design. The three designs have
Fmix = ∑ Fi ,mix similar configurations. For instance, in all the designs, the three
i (35) distributed systems (i.e., DSs 2, 4, and 5) were not used be-
cause the other distributed and terminal systems (i.e., DSs 1 and
FmixCc , mix = ∑ Fi ,mixCc ,i 3, and TS) can treat all the wastewaters generated in the plant
i (36) by fully utilizing their design capacities. This is because all the
2383 dx.doi.org/10.1021/ie3018934 | Ind. Eng. Chem. Res. 2013, 52, 2379−2388
Industrial & Engineering Chemistry Research Article

Figure 3. Configurations of the three designs for caste study II: (a) envi-design; (b) cost-design; and (c) eco-design (WW: wastewater, DS: distributed
system, TS: terminal system).

objective functions for the three designs have common targets later. It is noted that the eco-design seems to be generated by
(for instance, piping and electricity consumption) in reducing combining the characteristics of the envi-design and cost-design
environmental impacts and/or economic costs, as discussed (see Table 1).
2384 dx.doi.org/10.1021/ie3018934 | Ind. Eng. Chem. Res. 2013, 52, 2379−2388
Industrial & Engineering Chemistry Research Article

The eco-design exhibits a little higher total external cost than the most significant contributor to the total external costs, the
the envi-design but significantly lower than the cost-design (see envi-design and eco-design seem to be focused on reducing
Table 2). Since the envi-design was generated by minimizing pumping power with the first priority. This can be supported by
environmental impacts, the envi-design leads to the lowest total the similar configurations of the eco-design and envi-design.
external cost. It is interesting that the values of each external cost For economic cost as well, the eco-design lies between the
for the eco-design and the envi-design are in similar ranges, even envi-design and cost-design. Table 3 shows the economic costs of
though the eco-design was intended to minimize environmental the three designs. Like the results for external costs, economic
impacts and economic costs at the same time. For the con- costs of the eco-design and envi-design are in similar ranges. The
tributors to external costs, the eco-design exhibits lower external eco-design optimization was performed by reducing external
costs from electricity consumption for pumping than the cost- costs rather than economic costs. The cost-design has the lowest
design but higher external costs from piping, pipe recycling, and total economic cost. The cost-design exhibits higher O&M cost
M&R. It is noted that the cost difference for the pumping than the eco-design and envi-design but lower construction and
outweighed that for the piping, pipe recycling, and M&R; thus, disposal costs. The cost difference for the construction and
disposal is greater than for the O&M. It should be noted that,
the eco-design has lower external cost than the cost-design. It is
although the principal contributor to economic cost is the costs
noted that, because the electricity consumption for pumping is
for labor and electricity consumption in the O&M stage, the
highest contributors affecting the extent order of the total
Table 5. Quantitative Characteristics of the Envi-Design,
economic costs are pipe-related costs in the construction and
Cost-Design, and Eco-Design for Case Study II
disposal stages. This is because in this study three designs use the
envi- cost- eco- same distributed and terminal systems, respectively.
item unit design design design The eco-design exhibits the lowest total external and economic
pipe length m 89,570 59,310 66,790 cost due to the optimization of the trade-off between external and
distributed number (in use) - 7 5 6 economic costs in the life cycle stages. Table 4 shows the total
systems flow rate of m3/h 148.6 144.2 147.7 external and economic costs of the three designs. In the
bypassing
wastewater construction and disposal stages, the eco-design exhibits lower
terminal flow rate of treated m3/h 464.5 464.5 464.5 economic and external costs than the envi-design but higher than
systems wastewater the cost-design. In contrast, in the O&M stage the eco-design has
power electricity kW 1093.5 1331.1 1177.7 higher total economic and external cost than the envi-design but
requirement (pumping)
lower than the cost-design. Due to this trade-off, the eco-design is
electricity kW 17.6 16.4 17.0
(wastewater the most environmentally and economically friendly. These
treatment) results demonstrated that the developed mathematical model

Table 6. External Costs Due to Environmental Impacts from the Envi-Design, Cost-Design, and Eco-Design for Case Study II
(Unit: EUR in Thousands)
life cycle item envi-design cost-design eco-design
construction piping 93,418 25,685 29,727
operations & maintenance (O&M) electricity for pumping during a year 2,442,770 2,973,452 2,630,719
electricity for wastewater treatment during a year 39,315 36,634 37,974
maintenance & repair (M&R) during a year 2,803 771 892
subtotal during service time (15 years) 37,273,319 45,162,855 40,043,775
disposal pipe recycling 136,138 37,437 43,327
total during the life cycle 37,502,875 45,225,977 40,116,829

Table 7. Economic Costs of the Envi-Design, Cost-Design, and Eco-Design for Case Study II (Unit: EUR in Thousands)
life cycle stage item envi-design cost-design eco-design
construction piping 1,967,343 595,289 676,297
construction expenses 393,469 119,058 135,259
overhead 118,041 35,717 40,578
contractor’s profit 211,292 63,031 71,799
subtotal 2,690,145 813,095 923,933
operations & maintenance (O&M) electricity for pumping during a year 518,887 631,613 558,810
electricity for wastewater treatment during a year 8,351 7,782 8,066
labor during a year 826,663 666,664 746,664
maintenance & repair (M&R) during a year 80,704 24,393 27,718
subtotal during service time (15 years, discounted) 17,604,853 16,326,722 16,459,328
disposal pipe decommissioning 640,565 190,212 216,862
construction expenses 127,113 38,042 43,372
overhead 38,434 11,413 13,012
contractor’s profit 80,711 23,967 27,325
subtotal (discounted) 602,228 178,828 300,571
total 20,897,226 17,318,645 17,683,832

2385 dx.doi.org/10.1021/ie3018934 | Ind. Eng. Chem. Res. 2013, 52, 2379−2388

Industrial & Engineering Chemistry Research

■
Article

optimized the trade-offs between economic and external costs to ASSOCIATED CONTENT
simultaneously reduce the environmental impacts and economic *
S Supporting Information
costs of the integrated system. In addition, it is noted that the Additional calculations and tables as noted in the text. This
envi-design is more environmentally and economically friendly material is available free of charge via the Internet at http://pubs.
than the cost-design. acs.org.

■
4.2. Case Study II. Although the configurations of the three
designs for case study II look different due primarily to the AUTHOR INFORMATION
number of distributed systems (see Figure 3), overall, case study Corresponding Author
II showed almost the same consequences as case study I. For pipe *Tel.: +82-33-250-6358. Fax: +82-33-254-6357. E-mail: srlim@
length and pump power as principal contributors to economic kangwon.ac.kr (S.-R.L.). Tel: +82-53-950-7286. Fax: +82-53-
and external costs, respectively, the eco-design was between the 950-6579. E-mail: [email protected] (D.S.L.).
envi-design and cost-design (see Table 5). Thus, the eco-design
Notes
had the lowest economic and external costs (detailed results are The authors declare no competing financial interest.

■
presented in Tables 6, 7, and 8). Consequently, the developed
ACKNOWLEDGMENTS
Table 8. Total Economic and External Costs of the
This work was financially supported by Basic Science Research
Envi-Design, Cost-Design, and Eco-Design for Case Study II
Program through the National Research Foundation of Korea
(Unit: EUR in Thousands)
(NRF) funded by the Ministry of Education, Science and
life cycle stage envi-design cost-design eco-design Technology (MEST) (2011-0008373) and by 2011 Research
construction 2,783,563 838,780 953,660 Grant from Kangwon National University. This work was also
operations & maintenance 54,878,172 61,489,577 56,503,103 supported by the Priority Research Centers Program through the
(O&M) NRF grant funded by MEST (2010-0028301).
disposal
total
738,366
58,400,101
216,265
62,544,622
343,898
57,800,661 ■
Sets
NOMENCLATURE

mathematical model was verified again. In addition, it is noted C = {c|c is a contaminant in wastewater}, c = 1, 2, ..., Nc
that the cost-design utilizes the five distributed systems excluding WW = {ww|ww is a wastewater source}, ww = 1, 2, ..., Nm
DSs 4 and 5 to minimize labor cost significantly affecting the total SYS = {sys|sys is a distributed or terminal system}, sys = 1, 2, ...,
of economic cost, while the eco-design utilizes the six distributed Nn and 1, 2, ..., Nk
systems excluding DS 2, and the envi-design utilizes all the SPL = {spl|spl is a splitter}, spl = 1, 2, ..., Ns
systems. MIX = {mix|mix is a mixer}, mix = 1, 2, ..., Nt
4.3. Implications from the Two Case Studies. Eco-design Variables
for integrated wastewater treatment systems should be focused Ai,j cross-sectional area of a pipe from i to j
on optimizing the flow rates of streams by pump in order to Bi,j binary variable for the existence of a pipe from i to j
minimize the total of economic and external costs. Lower flow Bsys binary variable for the existence of a distributed or
rates lead to a decrease in external cost from pumping because of terminal system
a decrease in head loss (eqs 5, 6, 9, and 10) but to an increase in Cc,i concentration of a stream i
pipe-related cost because of an increase in the number (i.e., Cc,j concentration of a stream j
length) of pipelines (eqs 5 and 6); and vice versa. In the two case Cc,mix concentration at the outlet of a mixer
studies, the eco-designs were generated by optimizing the flow Cc,sys concentration at the inlet of a distributed or terminal
rates of streams, whereas the cost-designs and envi-designs were system
generated by increasing and decreasing flow rates, respectively. Cc,tsout concentration at the outlet of a terminal system
Next to flow rates, labor cost significantly affects the eco-design DLi,j labor cost for piping from i to j (i.e., WW to DS; WW
of integrated wastewater treatment systems due to its high con- to TS; DS to DS; and DS to TS)
tribution to economic cost. DPi,j pipe material cost for piping from i to j (i.e., WW to
DS; WW to TS; DS to DS; and DS to TS)
5. CONCLUSIONS ECCcon economic cost in the construction stage
Existing wastewater treatment systems were eco-designed based ECCtO&M economic cost in the O&M stage
on process integration. For these eco-designs, a mathematical ECCtdis economic cost in the disposal stage
optimization model was developed by monetizing environmental ECtp electricity cost for pumping
impacts to external costs and thus converting biobjective to ECtwt electricity cost for wastewater treatment
single objective problems. The developed model optimized the EXPdecom construction expenses for pipe decommissioning
trade-offs between environmental impacts and economic costs of EXPpiping construction expenses for piping
the integrated wastewater treatment systems. Thus, this eco- EDi,j external cost of pipe recycling from i to j (i.e., WW to
design strategy can be used to eliminate the limitation from DS; WW to TS; DS to DS; and DS to TS)
subjective weighting on environmental impacts and economic EEtp external cost of electricity consumption for pumping
costs in biobjective optimization and to generate a single design EEtwt external cost of electricity consumption for wastewater
outcome, regardless of who decision makers are. Since industrial treatment
plants have been trying to reduce freshwater and wastewater by EXCcon external cost in the construction stage
applying water reuse and recycling, the developed model can be EXCtO&M external cost in the O&M stage
widely used to eco-design existing wastewater treatment systems EXCdis external cost in the disposal stage
in various industrial plants. EMRCt external cost of M&R
2386 dx.doi.org/10.1021/ie3018934 | Ind. Eng. Chem. Res. 2013, 52, 2379−2388
Industrial & Engineering Chemistry Research Article

EPi,j external cost of piping from i to j (i.e., WW to DS; WW li,j pipe length from i to j (i.e., WW to DS; WW to TS; DS
to TS; DS to DS; and DS to TS) to DS; and DS to TS)
Fi,j flow rate from i to j (i.e., WW to DS; WW to TS; DS to Lmax
sys maximum contaminant load of a distributed or terminal
DS; and DS to TS) system
Fi,mix flow rate from i to a mixer ρ density of wastewater
Fmix flow rate from a mixer Rc,sys removal efficiency of a distributed or terminal system
Fspl flow rate to a splitter t service lifetime (15 year)
Fspl,j flow rate from a splitter to j top operating hours per annum (8,760 h)
Fsys flow rate at the inlet of a distributed or terminal system UCe unit economic cost of electricity (0.065 USD per kWh)
HLi,j head loss through a pipe from i to j (i.e., WW to DS; UCl,sys unit economic cost of the labor to operate a distributed
WW to TS; and DS to DS) or terminal system (96,000 USD and 320,000 USD per
COST life cycle cost year, respectively)
LCt labor cost UE unit external cost of electricity consumption (0.255
MRCt M&R cost EUR per kWh)

■
OHdecom contractor’s overhead for pipe decommissioning
OHpiping contractor’s overhead for piping
REFERENCES
Pi,j power requirement for pumping wastewater from i to j
(i.e., WW to DS; WW to TS; and DS to DS) (1) Wang, Y.-P.; Smith, R. Design of distributed effluent treatment
Psys power requirement for wastewater treatment in a systems. Chem. Eng. Sci. 1994, 49 (18), 3127.
distributed or terminal system (2) Galan, B.; Grossmann, I. E. Optimal design of distributed
wastewater treatment networks. Ind. Eng. Chem. Res. 1998, 37 (10),
PROdecom contractor’s profits for pipe decommissioning 4036.
PROpiping contractor’s profits for piping (3) Hernandez-Suarez, R.; Castellanos-Fernandez, J.; Zamora, J. M.
TDPpiping total direct pipe material cost Superstructure decomposition and parametric optimization approach
TDLdecom total direct labor cost for pipe decommissioning for the synthesis of distributed wastewater treatment networks. Ind. Eng.
TDLpiping total direct labor cost for piping Chem. Res. 2004, 43 (9), 2175−2191.
TEC total external cost during the life cycle (4) Dzhygyrey, I.; Jezowski, J.; Kvitka, O.; Statyukha, G. Distributed
vi,j optimum velocity in a pipe from i to j (i.e., WW to DS; wastewater treatment network design with detailed models of processes.
WW to TS; and DS to DS) Comput.-Aided Chem. Eng. 2009, 26, 853−858.
(5) Lim, S. R.; Lee, H.; Park, J. M. Life cycle cost minimization of a total
Parameters wastewater treatment network system. Ind. Eng. Chem. Res. 2009, 48 (6),
ade, bde regression parameter for an external cost of pipe 2965−2971.
recycling (ade = 96.955, bde = 1.3418) (6) Lim, S.-R.; Park, D.; Park, J. M. Environmental and economic
adl, bdl regression parameter for a labor cost for piping (adl = feasibility study of a total wastewater treatment network system. J.
2106.3, bdl = 16.326) Environ. Manage. 2008, 88 (3), 564−575.
adp, bdp regression parameter for a pipe material cost (adp = (7) Lim, S. R.; Park, J. M. Environmental impact minimization of a total
714.15, bdp = 2.9053) wastewater treatment network system from a life cycle perspective. J.
aop, bop regression parameter for an optimum velocity (aop = Environ. Manage. 2009, 90 (3), 1454−1462.
0.0297, bop = 0.6173) (8) Erol, P.; Thoming, J. ECO-design of reuse and recycling networks
by multi-objective optimization. J. Cleaner Prod. 2005, 13 (15), 1492−
ape, bpe regression parameter for an external cost of piping (ape = 1503.
66.457, bpe = 0.921) (9) Donnelly, K.; Beckett-Furnell, Z.; Traeger, S.; Okrasinski, T.;
α coefficient for construction expenses (0.2) Holman, S. Eco-design implemented through a product-based environ-
β coefficient for contractor’s overhead (0.05) mental management system. J. Cleaner Prod. 2006, 14 (15−16), 1357−
CCF currency conversion factor (0.8 for USD-to-EUR) 1367.
Cmax
c,tsout maximum concentration at the outlet of a terminal (10) Spitzley, D. V.; Grande, D. E.; Keoleian, G. A.; Kim, H. C. Life
system cycle optimization of ownership costs and emissions reduction in US
Cc,ww concentration of a wastewater source vehicle retirement decisions. Transp. Res. Part D-Transp. Environ. 2005,
γ coefficient for a contractor’s profits (0.1) 10 (2), 161−175.
δ coefficient for the external and economic costs of M&R (11) Swarr, T. E. Life cycle management and life cycle thinking: Putting
a price on sustainability. Int. J. Life Cycle Assess. 2006, 11 (4), 217−218.
(0.3) (12) Steen, B. A Systematic Approach to Environmental Priority Strategies
e escalation rate (0.03) in Product Development (EPS). Version 2000 - Models and Data of the
ε coefficient for the total direct labor cost of pipe Default Method; Center for Environmental Assessment of Products and
decommissioning (0.4) Material Systems (CPM): Stockholm, 1999.
f friction factor (0.02) (13) Matthews, H. S.; Lave, L. B. Applications of environmental
Fmax
i,j maximum flow rate from i to j (i.e., WW to DS; WW to valuation for determining externality costs. Environ. Sci. Technol. 2000,
TS; DS to DS; and DS to TS) 34 (8), 1390−1395.
Fmax
sys maximum flow rate at the inlet of a distributed or (14) Gibbons, E.; O’Mahony, M. External cost internalisation of urban
terminal system transport: A case study of Dublin. J. Environ. Manage. 2002, 64 (4), 401−
Fmin minimum flow rate from i to j (i.e., WW to DS; WW to 410.
i,j
(15) Rey-Martinez, F. J.; Velasco-Gomez, E.; Martin-Gil, J.; Gracia, L.
TS; DS to DS; and DS to TS) M. N.; Navarro, S. H. Life cycle analysis of a thermal solar installation at a
g acceleration of gravity (9.8 m/sec2) rural house in Valladolid (Spain). Environ. Eng. Sci. 2008, 25 (5), 713−
Ha additional head for the elevation of systems (10 m H2O) 723.
i interest rate (0.057) (16) Swanstrom, L.; Reiss, H.; Troitsky, O. Y. Environmental balances
ηm motor efficiency (0.85) of thermal superinsulations. Int. J. Thermophys. 2007, 28 (5), 1653−
ηp pump efficiency (0.6) 1667.

2387 dx.doi.org/10.1021/ie3018934 | Ind. Eng. Chem. Res. 2013, 52, 2379−2388

Industrial & Engineering Chemistry Research Article

(17) IKP PE Europe GMBH. Gabi 4.0 Software; 2004.

(18) Lim, S.-R.; Park, J. M. Environmental indicators for
communication of life cycle impact assessment results and their
applications. J. Environ. Manage. 2009, 90 (11), 3305−3312.
(19) GAMS Development Corporation. GAMS, A User Guide;
Washington, DC, 2005.
(20) Lim, S. R.; Park, J. M. Synthesis of an environmentally friendly
water network system. Ind. Eng. Chem. Res. 2008, 47 (6), 1988−1994.
(21) McGhee, T. J. Water Supply and Sewerage; McGraw-Hill: New
York, 1991.

2388 dx.doi.org/10.1021/ie3018934 | Ind. Eng. Chem. Res. 2013, 52, 2379−2388