Science of the Total Environment 387 (2007) 42 – 53

www.elsevier.com/locate/scitotenv

Evaluation of environmental burdens caused by changes of food

waste management systems in Seoul, Korea
Suk-Hui Lee a,b , Ki-In Choi b,⁎, Masahiro Osako b , Jong-In Dong a
a
Department of Environmental Engineering, University of Seoul, 90 Jeonnong, Dongdaemun, Seoul 130-743, Korea
b
Research Center for Material Cycles and Waste Management, National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba,
Ibaraki 305-8506, Japan
Received 5 February 2007; received in revised form 16 June 2007; accepted 19 June 2007
Available online 29 August 2007

Abstract

During the last decade, there have been remarkable changes in food waste management in Korea following a ban on direct
landfilling. To evaluate the environmental impacts of food waste management systems, we examined individual treatment systems
with the LCA approach – landfill, incineration, composting, and feed manufacturing – and estimated the change from 1997 to
2005. The efficient system was different in each impact category, but it was evaluated that landfill is the main contributor to human
toxicity and global warming (based on fossil CO2). In contrast, due to the increase of food waste recycling, acidification,
eutrophication, and fresh water aquatic ecotoxicity impact was increased. Especially, the high energy consumption and generated
residue in recycling systems caused the large burdens in toxicity categories.
© 2007 Elsevier B.V. All rights reserved.

Keywords: Food waste; Landfill; Incineration; Composting; Feed manufacturing; Environmental burden; Life cycle assessment

1. Introduction charged food waste used to be directly dumped with

Food waste is a main component of municipal solid
waste (MSW), which is known to induce putrefaction in
waste collection and transportation; lower the efficiency
of storing, conveying, shredding and separating; intro-
duce moisture and Cl in the incineration process; lead to
the emission of odorous compounds; adversely affect
the quality of leachate from landfill, and so on (Durlak
et al., 1997; Wang et al., 1999). In Korea, food waste
accounts for about 30% of total MSW generation, and
its proper management has become one of the most
actively debated issues in the last decade. Most dis-
⁎ Corresponding author. Tel.: +81 29 850 2727; fax: +81 29 850
2830.
E-mail address: choi.ki-in@nies.go.jp (K.-I. Choi).
0048-9697/$ - see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.scitotenv.2007.06.037
MSW, but it continuously provoked complaints from
those living near landfills. Therefore, not only to utilize
available resources but also to reduce the environmental
impact of landfills, in 1997 the Korean government
announced several policies for effective food waste
management. These included food waste reduction and
separate collection from sources, the enlargement of
recycling, and especially the prohibition of direct dis-
posal in landfills from 2005. Pursuant to these policies,
food waste management has attracted growing attention
in recent years, and it has greatly changed in a short
period of time, with a rapid increase of recycling and an
abrupt decrease of landfilling.
Despite increasing interest in studying individual
technologies, however, there has been little research on
the comprehensive evaluation of the Korean food waste
 S.-H. Lee et al. / Science of the Total Environment 387 (2007) 42–53
management system in the environmental view so far. In
this study, based on operating parameters of actual
facilities, we evaluated each fundamental food waste
management technology, considering the treatment of
one ton of food waste in systems as a unit. Based on the
calculated results as well as statistical data, we estimated
the change of environmental impacts of the food waste
management system in Seoul from 1997 to 2005 using
the LCA (life cycle assessment) method.
LCA is an objective and systematic tool for evalu-
ating environmental aspects throughout the life cycle of
a product or service. Though LCA aims to be science-
based, it involves a number of technical assumptions
and value choices. In addition, the environmental im-
pacts obtained from LCA are often described as poten-
tial impacts, because they are not specified in time and
space (Guinée, 2002). In spite of its limitations, the
growing importance of applying LCA to waste manage-
ment systems has been shown by many studies. Already
there have been some studies to compare a few biowaste
treatment systems (Hirai et al., 2000; Lundie and Peters,
2005; Güereca et al., 2006) as well as on the biological
treatments of organic material as a part of integrated
MSW management (Mendes et al., 2004; Eriksson et al.,
2005; Bovea and Powell, 2006; Özeler et al., 2006;
Hong et al., 2006) investigated by the LCA approach, to
the best of our knowledge Korea is the only country that
has implemented a series of strict food waste manage-
ment policies. Accordingly this study may help clarify
the substantial effect of separation and recycling of food
waste.
2. Current situation of food waste management
in Seoul
Seoul is the capital city of Korea, with a population
of approximately 10 million. Table 1 shows the state of
food waste generation and treatment in Seoul. Due to a
series of strict policies, e.g., separate food waste collec-
tion and volume-based charge system, generation grad-
ually decreased from 1998, but in contrast to the
government's expectations, it has been steadily increas-
Table 1
Food waste generation and its treatment in Seoul (million tons yr− 1)
1997 1998 1999
Generation 1.01 1.08 1.06
Landfill 0.88 0.82 0.70
Incineration 0.07 0.07 0.04
Recycling 0.06 0.19 0.32
a
Amounts estimated from data of waste treatment facilities in Seoul.
43
ing since 2003. Several reasons for its rebound have
been suggested, e.g., discrepancies of statistical defini-
tions in waste censuses, underestimation of the moisture
content of food waste with MSW, increasing nonrecycl-
ables in collected food waste, and so on. Especially, one
interesting feature is that the proportion of sorting resi-
dues in collected food waste destined for recycling, e.g.,
plastic chips, toothpicks, caps, and other miscellaneous
items, has increased since direct landfilling was banned
in 2005.
Landfill accounted for more than 90% of generated
food waste just a few years ago, but the same proportion
is now recycled. Officially, it was stated that all food
waste was completely recycled in 2005, but some reports
show that some food waste still flowed into incinerators
or landfill with MSW. Based on the treated amount of
MSW and its composition in each facility, it was esti-
mated that food waste in landfill and incinerators totaled
0.11 and 0.03 million tons in 2005, respectively, as
shown in Table 1. The collected food waste is transported
to recycling facilities in and around Seoul, but recycling
facilities in Seoul can handle less than 20% of the amount
generated, because most of them are located in suburbs
or rural areas largely as a result of the “not in my back
yard” (NIMBY) syndrome. Though the Korean govern-
ment has strived to diversify recycling technologies, e.g.,
anaerobic digestion, carbonization, co-treatment with
sludge or sewage, the main technologies are feed
manufacturing and composting, which account for
58% and 39% of the total recycled amount, respectively,
due to their attractive advantages of lower initial invest-
ment cost and higher source accessibility. The produced
feed and compost are supplied to farmlands, stock farms
or commercial feed makers at a low price or for free.
3. Materials and methods
3.1. Processes descriptions and assumptions
To estimate the environmental burdens from food
waste management in Seoul, we compared two systems:
the landfill-based waste management system for 1997,
2000 2001 2002 2003 2004 2005 a
0.93 0.93 0.92 0.95 0.97 1.30
0.41 0.27 0.26 0.20 0.09 0.11
0.05 0.03 0.02 0.00 0.00 0.03
0.47 0.63 0.64 0.75 0.88 1.16
44 S.-H. Lee et al. / Science of the Total Environment 387 (2007) 42–53

and the recycling-based one for 2005. Two types of food DOC Degradable organic carbon (fraction)(ton C/ton
waste were considered: separate collection for recycling, waste)
and collection mixed with other MSW for incineration DOCf Fraction DOC dissimilated
and landfill, because their physical and chemical F Fraction by volume of CH4 in landfill gas
compositions are quite different from each other due
to the large difference in moisture content, which According to the measured data at landfill S (SLC,
strongly influences the life cycle inventory (LCI) re- 2006), LFG generated in Landfill S contained 54%
sults. Table 2 shows the fundamental characteristics of (volume) CH4. Accordingly, it was calculated to be
the food wastes used in this study. 163 m3 per ton of food waste from the above equation:
To obtain the process burdens in food waste treat- 0.16 as the degradable organic carbon (DOC) fraction,
ment, four main treatment processes were considered: 0.54 as the methane fraction in LFG, 1.0 as the methane
landfill, incineration, composting, and feed manufactur- correction factor, and 0.55 as the dissimilated fraction of
ing, and three substitute production processes: conven- DOC. On the other hand, McDougall et al. (2001)
tional electricity, chemical fertilizer, and soybean. All assumed that 100 m3 of LFG was generated per ton of
operating parameters were taken from existing facilities biological recycling residue but that there was no
treating food waste from Seoul, however, there are no contribution from incineration ash, and these assump-
landfills and incinerators only for food waste in Korea, tions were used in this study. Based on the capacity of
operating parameters for those processes, e.g., energy the flaring facilities of the site, 90% of generated LFG
consumption and process performance, were taken from was assumed to be flared onsite, and then in comparison
MSW treatment conditions (Lim, 2003; MOE, 2006). with those in 1997 and 2005, two types of landfill were
Detailed process descriptions and emissions of each considered: process without power generation, and that
process are as follows. using 12% of the collected LFG for power generation,
Landfill: A large-scale sanitary landfill site (Landfill based on operation data of Landfill S.
S), which is located about 20 km west of Seoul, was Leachate generated by food waste was taken from
considered. Most MSW generated in Seoul had been that of Landfill S because leachate is known to strongly
dumped in Landfill S since 1992, which includes the depend upon precipitation. Based on the statistics for the
advanced leachate treatment process as well as a landfill closed site of Landfill S (SLC, 2006), leachate genera-
gas (LFG) flaring facility. Most of the generated LFG tion was estimated as 0.31 m3/t from the accumulated
had been directly flared at the gas extraction wells or gas generations and the estimations during the 20-year
collection facility, but some of the collected LFG has aftercare period (after 2005, we assumed it to be 81% of
been used for electricity generation since 2000. Esti- the former year's generation). Generated leachate was
mation of LFG generation from food waste was based treated on site with a removal efficiency of over 90%
on the IPCC guideline (IPCC, 2000): (based on CODcr) except ammonium ion which has
been removed at the advanced nitrogen removal process
Lo ¼ MCF⁎DOC⁎DOCf ⁎F ⁎16=12 since 2000. In this study, we assumed the whole leachate
was treated with a removal efficiency of 90%, but 99%
Lo Methane generation potential (ton CH4/ton removal of ammonium in 2005. LFG and leachate
waste) qualities used in this study are shown in Table 3.
MCF Methane correction factor (fraction) Incineration: The selected incinerator is a MSW
incinerator of 900 t/d capacity, which is one of the four
large-scale incinerators in Seoul. It features a series of
Table 2 general incineration processes, e.g., a heat recovery
Characteristics of food waste (FW) in Seoul
facility, advanced air pollution control devices (semi dry
Component Unit Separated FW FW in MSW reactor for acidic pollutants, fabric filter for particles,
Moisture % 79.7 56.7 and selective catalytic reduction for NOx and dioxins),
Ash % 4.6 10.3 wastewater treatment system, and so on. The recovered
Combustible % 15.7 32.9
heat is not only used onsite, but also sold for district
C % 8.3 15.8
H % 1.3 2.2 heating. In this study, however it was assumed that all
O % 5.4 13.6 recovered heat was used only for electricity generation
N % 0.7 1.3 with a generation efficiency of 11%, which was the
S % 0.0 0.0 gross figure for Korean MSW incinerators. CO2 emis-
LHV KJ/kg 1748 3525
sion was calculated to be 579 kg/t of food waste from the
 S.-H. Lee et al. / Science of the Total Environment 387 (2007) 42–53 45

Table 3 emission was estimated to be 6 kg/t of food waste as-

LFG and leachate quality used at landfill process suming that all degraded N obtained from the nitrogen
LFG Flared LFGa Raw leachate b balance of the process became NH3. On this result, we
(mg/m3) (mg/m3) (mg/l) assumed that 90% of NH3 was removed by biological
CO 30 BOD 1984 odor treatment at the facility. Wastewater was recircu-
CO2 746,429 a 1,807,143 c COD 3572 lated onsite without discharge.
CH4 385,714 a SS 218
Feed manufacturing: Food waste itself can be used as
NOx as NO2 173 NH+4 1854
SOx as SO2 24 As 3.6.E− 02 feed for stock farming, but such usage is hindered by
HCl 65 d 1.5 Cd 8.3.E− 03 problems like decay, odor during transportation and
HF 13 d Cr 7.4.E− 02 storage, and contamination. In Korea, it is common to
H2S 200 d Cu 7.7.E− 02 make feed or feed ingredients from food waste through a
HC 2000 d CN 6.3.E− 03
few mechanical or biological operations. The selected
Chlorinated HC 35 d Pb 6.1.E− 02
NH3 0.2 Hg 1.0.E− 03 feed manufacturing facility is the dry-type process,
As PO3−
4 37.6 which consists of only a few mechanical operations, e.g.,
Cd 5.6.E−03 d screening, shredding, dewatering, and drying process
Cr 6.6.E−04 d including odor treatment. In view of the characteristics of
Cu
the process, it is assumed that there are no air emissions
Pb 5.1.E−03 d
Hg 4.1.E−04 d from the process. On the other hand, according to the
Ni surveyed facility data it generated a relatively large
Zn 7.5.E−02 d amount of wastewater with a high concentration, which
a
Measured figure at LFG recovery facility in 2005 (SLC, 2006). was sent to the sewage treatment plant for treatment. We
b
Average concentration of raw leachate generated at the closed site assumed wastewater was treated with a removal
of landfill S for 1993–2005 (SLC, 2006). efficiency of 90%, and the electricity consumption for
c
It regarded all combusted CH4 became CO2.
d wastewater treatment was taken from the leachate
Typical concentration of LFG (White et al., 1995).
treatment of landfill. Wastewater generation and quality
at the selected facility are shown in Table 5. To consider
the displacement effect of the product, it was assumed
carbon content of input food waste (158 kg C/t of food that the final product could replace soybean as an
waste). Treated flue gas and wastewater characteristics ingredient of commercial feed, which could reduce the
used in this study are shown in Table 4. resources consumption and emissions caused by the
Composting: The considered composting facility is soybean cultivation after all. Other residues were sent to
an aerobic type located in east Seoul, and uses sawdust Landfill S, as in the case with composting.
with food waste to produce compost (food waste :
sawdust = 85 : 15). Sawdust is produced in the com-
posting facility for food waste treatment, so its effect was Table 4
considered as a part of the burden from food waste Flue gas and wastewater characteristics at incineration
composting in this study. About 17% of input food Flue gas Wastewater
wastes were converted to compost, but 23% of them 3
Generation (m /t) 2409 a
Generation (l/t) 432
were sent to Landfill S as residue. To estimate the emis- Particles 2 BOD (mg/l) 7
sions from the process, we used the characteristics of CO (mg/m3) 15 COD (mg/l) 17
compost given by McDougall et al. (2001): moisture CO2 (mg/m3) 240,349 SS (mg/l) 12
NOx (mg/m3) 66 Phenol (mg/l) 0.001
content of 40%, and N, P2O5, and K2O content of 1.18%,
SOx (mg/m3) 9 Cr (mg/l) 0.029
0.68%, and 0.9% dry weight, respectively, and assumed HCl (mg/m3) 3 Cu (mg/l) 0.026
that all produced compost was a competitive replace- As (mg/m3) 3.05E− 06 a Pb (mg/l) 0.02
ment for chemical fertilizers. On the other hand, only Cd (mg/m3) 5.73E− 04 a
CO2 and NH3 were considered as air emissions from Cr (mg/m3) 9.82E− 04 a
composting reactions. CO2 emission was calculated to Cu (mg/m3) 1.30E− 04 a
Pb (mg/m3) 4.82E− 03 a
be 210 kg/t of food waste, assuming that C accounted for Hg (mg/m3) 1.31E− 02 a
44% (theoretical composition of cellulose) of the de- Ni (mg/m3) 3.75E− 03 a
graded solid material obtained from the difference be- Zn (mg/m3) 7.46E− 03 a
tween the solid fraction of input material (food waste and a
Typical flue gas characteristics from organics incineration including
sawdust) and that of output (residue and compost). NH3 gas cleaning (McDougall et al., 2001).
46 S.-H. Lee et al. / Science of the Total Environment 387 (2007) 42–53
Table 5
Wastewater characteristics at feed manufacturing process
Generation (l/t) Raw wastewater
370
BOD (mg/l) 210,797
COD (mg/l) 242,213
SS (mg/l) 45,854
NH4 + (mg/l) 4113
PO3−
4 (mg/l) 1250
3.2. LCI modeling and impact assessment
The LCI of each system consists of not only the
burdens from the energy consumption of the system,
treatment process, and residue disposal, but also the
saving from product displacement. In detail, the LCI
includes the burdens from the production and use of
energy, and environmental emissions from disposal of
solid residues or by-products were considered as a part of
each system. If a product such as electricity, compost or
feed displaced a requisite for the external system, it was
also included as a benefit of that system. To obtain the
LCI for fuel and electricity, we used the IWM-2 model
(McDougall et al., 2001). This is the second version of
IWM, developed in 1995 (White et al., 1995), and
provides a useful tool to reflect the system electricity. In
considering the electricity consumed on site, the power
sources distribution (coal 37.3%, oil 8.2%, gas 12.1%,
nuclear 40.2%, hydro 2.1%) and efficiency (coal 37.4%,
oil 35.4%, gas 34.7%, nuclear 27.2%, hydro 76.5%) of
the Korean national grid were used. However, if the
electricity generated in the process exceeded the amount
required onsite, it was assumed that the surplus elec-
tricity replaced the conventional electricity (coal 54.8%,
oil 8.1%, gas 13.2%, nuclear 23.2%, hydro 0.8%), cal-
culated from the build margin (CDM PDD, 2004) and
operating margin of national grid. This means that the
exported electricity by waste source can reduce the use of
grid electricity, as well as delay new power plant con-
struction. Because chemical fertilizer and soybean de-
pend on greatly imports in Korea, their LCI for recycling
processes were taken from the food LCA database in
Simapro 7.0. In this study, the collection and transpor-
tation steps were excluded from the system boundary,
and the burdens from consumption of water, and che-
micals as well as processes of construction and demo-
lition of facilities were ignored due to their relatively
small effect compared with the treatment process
(Eriksson et al., 2005; Conssonni et al., 2005). The
schematic flows and parameters in each system are pre-
sented in Fig. 1. All predicted environmental burdens
were classified and characterized into five impact cate-
gories with CML method (Guinée, 2002): global
warming (GWP100 in kg CO2 eq./kg), human toxicity
(HTPinf in kg 1,4-DCB eq./kg), fresh water aquatic
ecotoxicity (FAETPinf in kg 1-4-DCB eq./kg., hereafter
ecotoxicity), acidification (AP in kg SO2 eq./kg), and
eutrophication (EP in kg PO43− eq./kg).
4. Results and discussion
4.1. Environmental impacts of each system
LCI results obtained from each system in Fig. 1 are
presented in Table 6,and the environmental impacts
characterizing the result of Table 6 are shown in Table 7.
In the global warming category, feed manufacturing
showed the least impact (about 36 kg CO2 eq. per
treatment of 1 t of food waste), because the selected
facility was based on the dry-type process which does
not involve any operation related with global warming.
However, incineration produced a large burden on
global warming (about 16 times higher than that of feed
manufacturing). This seems reasonable because com-
bustion itself is a sort of conversion of organic carbon to
CO2. Lundie and Peters (2005) compared several food
waste disposal options, and showed that codisposal at
landfill made a greater contribution than aerobic com-
posting or central composting. In addition, Güereca
et al. (2006) reported that a composting based system
had less impact on global warming than an incineration
based system in the biowaste treatment of Barcelona,
Spain. Also from the study on MSW disposal by
Eriksson et al. (2005) and Hong et al. (2006), landfill
had more impact on global warming than incineration.
From those researches, it could be predicted that landfill
would have more impact than incineration in waste
treatment, but landfill showed less impact than inciner-
ation assuming 90% flaring of LFG as well as large
consumption of fuel during flue gas cleaning in inci-
neration in this study. As other studies showed, it has
been evaluated that landfill has more impact on global
warming than composting due to CH4 in LFG, however,
LFG power generation in landfill (12% of collected
LFG) scarcely affected global warming, with a reduc-
tion of less than 1%.
In recent days, global warming has become a special
interest in view of the clean development mechanism
(CDM). Though some researchers consider that biogen-
ic CO2 from the landfilling process does not need to be
considered as a net contribution to global warming since
it is a consequence of the natural process of biodegrad-
able wastes in a strict sense (Finnveden et al., 2000;
 S.-H. Lee et al. / Science of the Total Environment 387 (2007) 42–53 47

Mendes et al., 2004; Bovea and Powell, 2006), to date it
has been widely applied also to other processes relating
to biodegradable waste in the field of CDM and national
greenhouse gas (GHG) inventory (IPCC, 2006). There-
fore, we divided CO2 into biogenic CO2 and fossil CO2,
based on the assumption that the entire carbon content in
food waste is biogenic carbon. Accordingly, CO2
converted from CH4 in LFG by flaring or combustion
in the landfill process is regarded as being of biogenic
origin, too. When biogenic CO2 was excluded from the
global warming category, the impacts were completely
changed, as shown in Fig. 2. Especially, the decrease in

Fig. 1. System flows and parameters considered in food waste treatment in Seoul.
48 S.-H. Lee et al. / Science of the Total Environment 387 (2007) 42–53

Fig. 1 (continued ).

incineration and composting was more dominant than in
other systems, of which biogenic CO2 accounted for
about over 99% and 92% of total GHG emissions,
respectively. Accordingly, incineration caused the least
impact. Though recycling products reduced GHG emis-
sions by 150 g and 218 g CO2 eq./t of food waste in
composting and feed manufacturing, respectively, this is
only about 1% of total GHG emissions. On the other
hand, the production of surplus electricity by inciner-
ation greatly reduced GHG, avoiding 1.1 kg CO2 eq./t of
food waste. Seen in the perspective of CDM, therefore,
the net contribution of incineration to global warming
no longer seems remarkable, and perhaps even better,
whereas the landfill system would contribute signifi-
cantly to global warming.
Compared to others, landfill systems caused large
burdens on human toxicity accounting for over 60 kg
1,4-DCB eq./t of food waste, mainly due to unburnt
 S.-H. Lee et al. / Science of the Total Environment 387 (2007) 42–53 49

Table 6
LCI results of each system (g/t of food waste)
Landfill a Landfill b Incineration Composting Feed manufacturing
Air Emissions Particulates 2.0E+00 − 2.0E+01 − 1.2E+01 3.3E+01 1.3E+01
CO c 1.0E+01 8.3E+00 5.2E+01 4.4E+00 2.3E+00
CO2_biogenic 2.8E+05 2.8E+05 5.8E+05 2.5E+05 2.0E+04
CO2_fossil 2.3E+03 − 1.4E+04 2.5E+04 2.5E+04 8.5E+04
CH4 6.3E+03 6.2E+03 5.8E+01 9.7E+02 7.2E+02
NOx 4.7E+01 7.0E+00 1.7E+02 6.9E+01 − 3.9E+00
N2O 3.6E− 02 − 9.7E− 02 3.1E− 01 2.1E−01 8.1E− 01
SOx 1.0E+01 − 5.8E+01 − 1.4E+01 1.1E+02 7.5E+01
HCl 1.5E+00 −2.2E+00 4.9E+00 5.5E+00 2.4E+00
HF 2.4E− 01 − 1.5E− 01 − 3.1E− 01 6.0E−01 2.6E− 01
H2S 3.3E+00 3.3E+00 0.0E+00 4.6E−01 2.4E− 01
Total HC 3.3E+01 3.3E+01 0.0E+00 4.6E+00 2.4E+00
Chlorinated HC 5.7E− 01 5.7E− 01 8.6E− 07 8.0E−02 4.2E− 02
Dioxins/Furans 0.0E+00 0.0E+00 0.0E+00 −1.9E− 15 − 3.4E− 16
Ammonia 3.7E− 02 − 3.7E− 02 − 6.0E− 02 6.0E+02 4.8E− 02
Arsenic 0.0E+00 0.0E+00 7.4E− 06 0.0E+00 0.0E+00
Cadmium 1.2E− 04 − 6.2E− 05 1.3E− 03 3.2E−04 1.1E− 04
Chromium 1.1E− 05 1.1E− 05 2.4E− 03 1.5E−06 7.9E− 07
Copper 0.0E+00 0.0E+00 3.1E− 04 0.0E+00 0.0E+00
Lead 3.8E− 04 − 2.8E− 03 9.2E− 03 5.3E−03 2.1E− 03
Manganese 1.2E− 04 − 1.5E− 03 1.3E− 03 2.6E−03 1.1E− 03
Mercury 4.5E− 05 − 4.8E− 04 3.2E− 02 7.4E−04 2.9E− 03
Nickel 1.7E− 03 − 1.2E− 02 − 1.0E− 03 2.6E−02 9.4E− 03
Zinc 1.9E− 03 − 3.1E− 03 1.4E− 02 8.2E−03 2.0E− 03
Water Emisions BOD 6.2E+01 6.2E+01 1.4E+01 1.4E+01 7.8E+03
COD 1.1E+02 1.1E+02 2.8E+01 2.6E+01 9.0E+03
Suspended Solids c 8.0E+00 4.6E+00 2.3E+01 8.0E+00 1.7E+03
TOC c 2.6E− 01 − 9.6E− 01 1.8E+01 1.3E+00 4.0E+01
AOX 7.2E− 05 − 4.6E− 05 − 6.6E− 05 2.0E−04 − 2.6E− 04
Chlorinated HC 1.7E− 05 − 1.3E− 05 2.7E− 04 4.8E−05 5.6E− 04
Phenols 2.6E− 03 − 2.1E− 03 8.7E− 04 9.5E−03 − 1.0E− 03
Aluminium c 8.4E− 01 − 1.1E+01 − 9.9E+00 1.8E+01 7.4E+00
Ammonium 5.8E+01 5.7E+01 1.0E+01 1.3E+01 1.6E+02
Arsenic 2.8E− 03 − 2.2E− 02 − 2.0E− 02 3.7E−02 1.5E− 02
Barium 1.1E− 01 − 9.5E− 01 − 8.3E− 01 1.6E+00 4.3E− 01
Cadmium 3.2E− 04 − 3.5E− 04 − 4.9E− 04 1.1E−03 3.7E− 04
Chloride c 1.5E+01 − 7.9E+01 − 7.8E+01 1.5E+02 1.7E+00
Chromium 1.1E− 02 − 1.1E− 01 − 8.5E− 02 1.8E−01 7.7E− 02
Copper 6.6E− 03 − 5.5E− 02 − 3.8E− 02 9.0E−02 3.7E− 02
Cyanide c 3.0E− 04 − 1.5E− 04 − 2.8E− 04 8.7E−04 2.5E− 05
Iron c 2.7E− 01 − 3.6E+00 − 2.3E+00 5.7E+00 3.9E+00
Lead 6.7E− 03 − 5.9E− 02 − 4.3E− 02 1.0E−01 4.2E− 02
Mercury 3.4E− 05 3.8E− 06 4.3E− 05 6.6E−05 1.6E− 04
Nickel 4.3E− 03 − 5.8E− 02 − 5.0E− 02 9.1E−02 3.7E− 02
Nitrate 3.2E− 02 − 2.8E− 01 − 2.4E− 01 4.7E−01 1.3E− 01
Phosphate 1.2E+00 4.8E− 01 − 3.8E− 01 1.3E+00 4.7E+01
Sulphate c 6.5E+00 − 6.4E+01 − 5.8E+01 1.3E+02 4.5E+01
Sulphide c 6.2E− 04 − 6.1E− 04 − 7.2E− 04 2.5E−03 − 1.7E− 03
Zinc 8.6E− 03 − 1.2E− 01 − 1.0E− 01 1.8E−01 7.3E− 02
a
No electricity generation.
b
Electricity generation with 12% of the collected LFG.
c
No characterization factor for considered impact categories.

LFG emissions of hydrocarbons, hydrofluoride, and method, and the characterization model used for toxicity
heavy metals etc. Hirai et al. (2000) reported that human categories was the USES-LCA model for infinite time
toxicity is affected significantly by the characterization horizon of Huijbregts et al. (2000), which models fate
50 S.-H. Lee et al. / Science of the Total Environment 387 (2007) 42–53
Table 7
Enviromental impacts of each system
Category Unit (eq./t) Landfill a
Global warming g CO2 409,433
Human toxicity g 1,4-DCB 62,847
Acidification g SO2 51
Eutrophication g PO3−
4 30
Ecotoxicity g 1,4-DCB 53
a
No electricity generation.
b
Electricity generation with 12% of the collected LFG.
and exposure routes more realistically and comprehen-
sively than earlier methods (Guinée, 2002). Based on
this model, LFG showed an intimate relation to human
toxicity. In every system except incineration, untreated
LFG contributed over 70% to the human toxicity im-
pact, whereas incineration, which is assumed not to
generate LFG in ash disposal, showed net savings by the
displacement of conventional electricity. It is unrealistic
to expect 100% collection and flaring of LFG in real
landfill sites, therefore landfill avoiding biodegradable
residues is an alternative to reduce human toxicity im-
pact in the recycling systems of Seoul.
In the acidification category, composting showed the
greatest potential impact of at least 12-fold within the
examined systems, mainly due to the NH3 emissions
during the composting process. In other systems, SOx
and NOx emissions caused by fuel consumption were
main contributors of acidification impact. Accordingly,
landfill with LFG recovery showed the least impact not
only less fuel consumption but also electricity generation
from LFG. Bovea and Powell (2006) and Özeler et al.
(2006) indicated fuel consumption during collection as
main contributors to acidification. However, the collec-
tion process was outside of the system boundary in this
study, consequently fuel consumption of system life
cycle showed intimate relations in this category.
Fig. 2. Global warming impacts with and without biogenic CO2 emission.
Landfill b Incineration Composting Feed manufacturing
408,110 580,317 269,647 35,748
60,394 − 909 12,706 6002
− 49 106 1293 75
24 26 226 468
−497 − 405 822 283
The eutrophication impact caused by recycling sys-
tems was about 10 to 20 times compared to other sys-
tems due to the contribution of the nitrogen source and
organic compounds, etc. The main contributor in the
composting process was NH3 as indicated by other
studies (Bovea and Powell, 2006), and wastewater of
feed manufacturing containing high concentrations of
BOD, COD etc. contributed significantly to eutrophi-
cation. A few researchers indicated that the main con-
tributor of eutrophication was leachate from traditional
landfills (Eriksson et al., 2005; Güereca et al., 2006;
Hong et al., 2006), however leachate from landfill was
treated with over 90% efficiency, and so its contribution
became relatively insignificant. On the other hand,
though we assumed 90% removal of emissions in the
recycling processes, in real facilities, odor problems
arising from the entire process including storing raw
food waste are hard to control fully. Moreover, the
disposal of high-concentration wastewater is another
issue to be resolved in food waste recycling.
Regarding ecotoxicity, systems including electricity
generation showed net savings: incineration and landfill
with LFG power generation, moreover, LFG power gen-
eration significantly reduced the impact in the landfill
system, avoiding 550 g 1,4-DCB eq. per ton of food waste.
This reflected the proportion of net electricity usage in
 S.-H. Lee et al. / Science of the Total Environment 387 (2007) 42–53
each system because the dominant contributors in this
category were metals such as nickel, barium etc in water
emission, which were provided mainly by conventional
electricity production processes. Although recycling
products reduces the impact slightly by reducing energy
consumption during substitute production by 5 g and 66 g
1,4-DCB eq./t of food waste in composting and feed
manufacturing, respectively, the benefit of surplus
electricity produced during incineration and landfill was
far more than that of the recycling product. Finally, the
landfill system with LFG power generation producing
more surplus electricity than incineration showed the least
impact, but composting showed the most impact. Like in
eutrophication category, recycling systems imposed more
impact in ecotoxicity due to the large energy consumption.
4.2. Environmental impacts in 1997 and 2005
Based on the results of environmental impacts as well
as statistical data for food waste management, we
Fig. 3. The change of environmental impacts between 1997 and 2005.
51
investigated the change of food waste management
system in Seoul. As mentioned already, it was con-
sidered that LFG from not only food waste but also
residue disposal in each system was used to generate
electricity in 2005. In addition, though some food waste
was treated by anaerobic digestion, co-treatment with
sewage, etc. in 2005, we included only the composting
and feed manufacturing amount, which accounted for
97% of the total recycling stream. The total amount of
food waste was 1,006,670 t and 1,256,650 t in 1997 and
2005 respectively, and the treatment share of each sys-
tem was 87:7:6:0 (landfill : incineration : composting :
feed manufacturing) and 9:2:36:53 in 1997 and 2005
respectively. Fig. 3 shows the changes of environmental
impacts in food waste management systems between
1997 and 2005. In spite of the increase of food waste
generation since 2003, it was evaluated that the impacts
have decreased greatly in the global warming and
human toxicity categories, but have increased in acidi-
fication, eutrophication, and ecotoxicity. This is because
52 S.-H. Lee et al. / Science of the Total Environment 387 (2007) 42–53
the main contributors of the considered systems have
completely changed. Compared with that in 1997, the
impact on global warming has decreased about 50%,
moreover, when biogenic CO2 is excluded, the impact
was reduced by 70%. Human toxicity also decreased
about 70% in 2005. This has obviously resulted from the
rapid decrease of landfilling that used to cause serious
environmental burdens in these categories. On the other
hand, the increase of food waste recycling increased the
impact by four-, nine-, and six-fold in acidification,
eutrophication, and ecotoxicity, respectively. Especially,
the appearance of feed manufacturing has contributed
greatly to the increased impact of eutrophication due to
wastewater discharge. This result shows that attention
needs to focus on emission control in the recycling
process, and indeed, odor problems have often angered
residents living near recycling facilities. In addition, the
LFG power generation at landfill in 2005 reduced the
burden in the ecotoxicity category by about 17%
compared to no power generation, however, the benefit
in the other impact categories was small, with a re-
duction of less than 3%. Although the 90% flaring of
LFG showed relatively small environmental benefits, in
view of energy conservation, power generation using
LFG may be highly attractive due to the high energy
potential of LFG.
5. Conclusions
The food waste management system in Seoul has
undergone rapid and dramatic changes. The results of this
study clearly show that their environmental burdens also
have changed dynamically in recent years as recycling has
increased while landfilling has decreased. Though there is
no doubt that the changes have reduced the global
warming and human toxicity impact from food waste
substantially, the outcome is not without problems due to
the increase of recycling share imposing more impacts in
some other categories. Moreover, far from the optimistic
assumption in this study, namely 100% displacement by
the produced compost or feed, these rarely guarantee
consumer choice in the marketplace due to their low
product quality and fear of adverse effects on livestock.
Besides on it, strong food waste policy can cause
excessive social burdens, as witnessed by the increase in
food waste generation in the past three years. Especially,
as mentioned at the beginning, sorting residues from
recycling facilities has tended to increase since direct
landfilling was banned in 2005, the presence of non-
recyclable materials in food waste occasionally causes
serious problems in the food waste recycling process.
Hence, to gain maximum merits and reduce environmen-
tal burdens in the recycling process, it is necessary to
separate food waste perfectly from the sources as well as
to control the emission and product quality by appropriate
regulations. The food waste policies of the Korean
government have given priority to reducing generation
rather than recycling. In terms of food waste generation
excepting nonrecyclables, it may have decreased since the
policy was introduced, however the effect of this action
was not outstanding compared to the change of treatment
share. As shown in the above results, every system im-
poses some environmental burden, consequently the most
effective way to reduce potential environmental burdens
in the food waste management system is to reduce the
generation of waste itself.
References
Bovea MD, Powell JC. Alternative scenarios to meet the demands of
sustainable waste management. J Environ Manage 2006;79:115–32.
CDM PDD. Project Design Document of the Gangwon Wind Park
Project Ver.2. Unison Co., Korea; 2004 (http://cdm.unfccc.int/).
Conssonni S, Giugliano M, Grosso M. Alternative strategies for
energy recovery from municipal solid waste Part B: emission and
cost estimates. Waste Manage 2005;25:137–48.
Durlak SK, Biswas P, Shi J. Equilibrium analysis of the effect of
temperature, moisture and sodium content on heavy metal
emissions from municipal solid waste incinerators. J Hazard
Mater 1997;56:1–20.
Eriksson O, Carlsson RM, Frostell B, Bjorklund A, Assefa G,
Sundqvist JO, et al. Municipal solid waste management from a
systems perspective. J Clean Prod 2005;13:241–52.
Finnveden G, Johansson J, Lind P, Moberg A. Life cycle assessment of
energy from solid waste. Fms report, vol. 137. Sweden: Stockholm
University; 2000.
Guinée JB. Handbook on life cycle assessment-operational guide to
the ISO standards. Kluwer Academic Publishers; 2002.
Güereca LP, Gassó S, Baldasano JM, Guerrero PJ. Life cycle
assessment of two biowaste management systems for Barcelona,
Spain. Resour Conserv Recy 2006;49(1):32–48.
Hirai Y, Murata M, Sakai S, Takatsuki H. Life cycle assessment for
food waste recycling and management. 4th international confer-
ence on ecobalance; 2000. p. 335–8.
Hong RJ, Wang GF, Guo RZ, Cheng X, Liu Q, Zhang PJ, Qian GR.
Life cycle assessment of BMT-based integrated municipal solid
waste management: case study in Pudong, China. Resour Conserv
Recy 2006;49(2):129–46.
Huijbregts MAJ, Thissen U, Guinée JB, Jager T, Meent D, Ragas
AMJ, et al. Priority assessment of toxic substances in life cycle
assessment, Part I: calculation of toxicity potentials for 181
substances with the nested multi-media fate, exposure and effects
model USES-LCA. Chemosphere 2000;41(4):541–73.
IPCC (Intergovernmental Panel on Climate Change). Good practice
guidance and uncertainty management in national greenhouse gas
inventories; 2000.
IPCC (Intergovernmental Panel on Climate Change). 2006 IPCC
guidelines for national greenhouse gas inventories; 2006.
Lim, ST. A LCA study on the Incineration and Landfill Process of
Municipal Solid Waste (Master's thesis). Korea: Konkuk Univer-
sity; 2003.
 S.-H. Lee et al. / Science of the Total Environment 387 (2007) 42–53
Lundie S, Peters MG. Life cycle assessment of food waste manage-
ment options. J Clean Prod 2005;13:275–86.
McDougall F, White P, Franke M, Hindle P. Integrated solid waste
management: a life cycle inventory. 2nd edition. Blackwell
Publishing; 2001.
Mendes MR, Aramaki T, Hanaki K. Comparison of the environmental
impact of incineration and landfilling in Sao Paulo City as
determined by LCA. Resour Conserv Recycl 2004;41:47–63.
MOE. Operation status of municipal solid waste incinerators in 2005.
Korea: Ministry of Environment; 2006.
Özeler D, Yetis Ü, Demirer GN. Life cycle assessment of municipal
solid waste management method: Ankara case study. Environ Int
2006;32:405–11.
53
SLC. The fourth number of Sudokwon landfill statistics yearbook.
Korea: Sudokwon Landfill Site Management Corp; 2006.
Wang KS, Chiang KY, Lin SM, Tsal CC, Sun CJ. Effects of chlorides
on emissions of toxic compounds in waste incineration: study on
partitioning characteristics of heavy metal. Chemosphere
1999;38:1833–49.
White PR, Franke M, Hindle P. Integrated solid waste management: a
life cycle inventory. Blackie Academic and Professional; 1995.