Renewable and Sustainable Energy Reviews 49 (2015) 169–177

Contents lists available at ScienceDirect

Renewable and Sustainable Energy Reviews

journal homepage: www.elsevier.com/locate/rser

Life cycle assessment of food waste-based biogas generation

Changqing Xu, Wenxiao Shi, Jinglan Hong n, Fangfang Zhang, Wei Chen
Shandong Provincial Key Laboratory of Water Pollution Control and Resource Reuse, School of Environmental Science and Engineering, Shandong University,
Shanda South Road 27, Jinan 250100, PR China

art ic l e i nf o a b s t r a c t

Article history: Life cycle assessment was performed by using the ReCiPe model to estimate the environmental effects of
Received 5 December 2014 three food waste (FW)-based biogas generation scenarios. Uncertainty analysis was also conducted to
Accepted 26 April 2015 confirm and add credibility to the study. Results showed that the potential impacts of human toxicity,
freshwater eutrophication, marine ecotoxicity, and fossil depletion had dominant contributions to the
Keywords: overall environmental impact. Electricity consumption during anaerobic digestion (AD) and the
Life-cycle assessment transportation of raw materials during landfill stage exhibited high potential impacts. The FW to landfill
Food waste scenario with and without energy recovery had the highest environmental impact. Moreover, unce-
Energy recovery rtainty analysis indicated that landfill was unsuitable for treating FW. Increasing biogas generation
Anaerobic digestion
capacity, improving electricity generation efficiency, optimizing the energy structure of China, and
Biogas
decreasing electricity consumption during the AD stage are effective ways for reducing the adverse
effects on the environment.
& 2015 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
2. Scope definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
2.1. Functional unit. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
2.2. System boundary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
2.3. Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
2.4. Data sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
2.5. Life-cycle inventory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
3. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
3.1. Midpoint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
3.2. Uncertainty analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
3.3. Energy type and recovery capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
3.3.1. Energy recovery capacity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
3.3.2. Energy type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
3.4. Sensitivity analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
3.4.1. Sensitivity to the impact assessment method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
3.4.2. Sensitivity to main contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
5. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177

1. Introduction

Significant amounts of food waste (FW) have dramatically

n
Corresponding author. Tel.: þ 86 0531 88362328; fax: þ86 0531 88364513. increased worldwide because of the vast increase in population
E-mail address: [email protected] (J. Hong). and urbanization. FW includes a large proportion of the total

http://dx.doi.org/10.1016/j.rser.2015.04.164
1364-0321/& 2015 Elsevier Ltd. All rights reserved.
170 C. Xu et al. / Renewable and Sustainable Energy Reviews 49 (2015) 169–177

Energy and raw materials production

Road Transportation (Assume 100 km)

S-1 3.56 m
192.31 kWh Disposal
Anaerobic Wastewater
-812.47 kWh in
digestion with Electricity Sewage
-3684.17 kJ sewage sludge sewage sludge
FeCl
(1t)
Food waste collection (1t)

treatment Treatment
4.48 m
129.67 kWh S-2 plant
Wastewater
5.49 m Water Anaerobic
Biogas
-755.77 kWh digestion
CO2 residue
(1t)
-3427.06 kJ
Combustion
0.76kWh (5%)
Biogas
1.2L Diesel
S-3 Electricity
1.3 m Sand generation (70%)
Food Storage
0.35 m Water
waste to Heat supply
2.4 × 10 m LDPE
Landfill (25%)
0.12 kg Pesticide
(1t)
576.87 kg CaO
-758.00 kWh Leachate
-3440.00 kJ Treatment

Impact assessment (ReCiPe method)

LCIA results

Fig. 1. System boundary.

municipal solid waste (MSW) of the country. For instance, the (88%) produced from landfill or AD stages are emitted to the air
quantity of FW in the United States, United Kingdom, France, without being utilized [8].
Germany, Holland, Switzerland, Japan, Korea, and Singapore Accordingly, research needs to address certain issues to present a
accounted for approximately 12%, 27%, 22%, 15%, 21%, 20%, 23%, systematic and reliable assessment. The following should be con-
23%, and 30% of the total solid waste, respectively [1]. In China, the ducted: (a) all impact categories must be considered; (b) uncertainty
amount of FW reached 9  107 t in 2010 (i.e., approximately 38% of analysis should be conducted to provide a credible assessment;
the total solid waste) with a 10% increase each year [2]. FW (c) energy recovery from biogas generated from landfill and AD
contains numerous nutrient elements and has great energy reco- stages should be taken into consideration; (d) biogas production
very potential. It also contains large amounts of pathogenic org- capacity on commonly used FW treatment technologies (i.e., landfill
anisms that are harmful to humans and the environment. If and AD) in China should be compared to different biomasses in the
handled improperly, aggravated environmental problems will world; (e) the key factors for reducing the potential environmental
threat ecosystem human health. Therefore, a comprehensive burden generated by FW treatment should be identified. A LCA was
method for evaluating environmental burdens is highly needed. conducted to evaluate the environmental impact of three FW treat-
Life-cycle assessment (LCA) is used to evaluate the environ- ment scenarios in China.
mental impacts associated with the entire life-cycle treatment of
a product, process, or activity [3]. LCA has been widely used for
eco-labeling programs, strategic planning, and marketing. LCA 2. Scope definition
applications also include product design, process improvement,
and consumer education. The environmental impact of FW 2.1. Functional unit
treatment has been studied extensively by using the LCA method
[4–6]. However, no studies on FW treatment in China have been The functional unit is the base for the treatment comparison in
published in English-language peer-reviewed journals. In addi- the life-cycle inventory (LCI). In this study, the management of 1t
tion, most previous studies only focused on several impact volatile solid (VS) is selected. All emissions, transport, materials,
categories (e.g., climate change and acidification). Moreover, wastewater treatment, energy consumption and recovery levels
although the quantification of uncertainties in the LCA related are based on this functional unit.
to input and output results are important for correct interpreta-
tion and use, researchers have been conducting LCA studies of 2.2. System boundary
FW treatment without taking uncertainty into consideration.
Furthermore, China is well known as one of the largest energy Three scenarios for FW treatment are considered in this study.
consumers and greenhouse gas emitters in the world [7]. With These scenarios include (a) the AD of FW and sludge (S-1), (b) AD
increasing environmental and energy pressure, China has focused of FW (S-2), and (c) FW to landfill (S-3). Fig. 1 shows the system
on using renewable energy to reduce environmental effects. boundary and flow of main materials of each scenario. For all
Biogas generated from different types of biomasses (e.g., straw, scenarios, the common processes are biogas utilization, direct air
sugar beet, maize, grass silage, sewage sludge, and FW) has great emissions (i.e., carbon dioxide, nitrogen, hydrogen, sulfur dioxide,
potential to regenerate electricity. However, most of the biogas nitrogen oxides, and hydrogen sulfide), raw materials production
 C. Xu et al. / Renewable and Sustainable Energy Reviews 49 (2015) 169–177 171

Table 1 collected from Refs. [18,19]. Data (i.e., electricity consumption, raw
Characteristics of the FW considered in the study. materials consumption, direct water emissions) on wastewater
treatment were obtained from field research at the Guangda
Characteristics Unit Value
sewage treatment plant in Jinan. Uncertainty analysis was con-
Water content % 77.17 ducted by using the Monte Carlo model [20,21] to avoid the effect
pH – 4.57 of data sources and add credibility to the study.
BOD g/kg 173.92
CODcr g/kg 233.77
T-N g/kg 7.54 2.5. Life-cycle inventory
T-P g/kg 0.98
VS/TS % 92.60 The main inventory results for each FW treatment scenario are
illustrated in Table 2. The LCI analysis sums up emissions and
extractions of the consumption of raw materials, water, energy,
and consumption, energy consumption and recovery, road trans- wastewater and transport.
portation of raw materials, and infrastructure. Wastewater treat-
ment is an additional process for S-1 and S-2, and leachate
treatment is an additional process for S-3. The road transport for 3. Results
FW collection is excluded because it is common to all scenarios.
3.1. Midpoint
2.3. Methodology
The life-cycle impact assessment (LCIA) results of each scenario
The life cycle impact assessment (LCIA) results are calculated at were compared by using the ReCiPe method in Table 3. S-3
midpoint level by using the ReCiPe [9,10] method. This method is exhibited a significant impact on most categories, except for the
the most recent indicator approach available in LCA analysis which climate change and human toxicity categories. These findings
is based on the results of Eco-indicator 99 [11] and Centrum voor could be attributed to direct air emissions, transport of raw mat-
Milieukunde Leiden on LCA [12]. The ReCiPe method uses impact erials, and diesel consumption during the landfill stage. For the
mechanisms that have a global scope and considers 18 midpoint climate change and human toxicity categories, a high potential
categories: climate change, ozone depletion, human toxicity, ph- impact was observed in S-1, this result could be attributed to the
otochemical oxidant formation, particulate matter formation, direct carbon dioxide emissions and electricity consumption dur-
ionizing radiation, terrestrial acidification, freshwater eutrophica- ing the AD stage. For S-1, electricity consumption was an impor-
tion, marine eutrophication, terrestrial ecotoxicity, freshwater tant factor in each category. Direct emissions and infrastructure
ecotoxicity, marine ecotoxicity, agricultural land occupation, urban also exerted another dominant impact on the climate change and
land occupation, natural land transformation, water depletion, urban land occupation categories, respectively. For S-2, electricity
metal depletion, and fossil depletion. To compare midpoint im- consumption was an important factor in most categories except
pacts and to analyze the respective share of each midpoint impact for agricultural land occupation, natural land transformation,
to the overall impact, normalization which is determined by the water depletion, and metal depletion categories, which were dom-
ratio of the impact per unit of emission divided by the per capita inated by infrastructure. Infrastructure also played important roles
world impact is also applied in this study. The detailed methodol- in ozone depletion, ionizing radiation, terrestrial ecotoxicity, and
ogy and complete characterization factors for ReCiPe are available urban land occupation categories. Direct emissions and coal
on the website of the Institute of Environmental Science of Leiden consumption imposed another dominant impact on the climate
University of the Netherlands [13]. Moreover, the IMPACT2002þ change and fossil depletion categories, respectively. For S-3,
[14] and the TRACI [15] methods were used as a comparison to transport was an important factor in each category. Furthermore,
supplement and verify the applicability of the obtained results quicklime consumption provided an additional contribution to
from ReCiPe method. most categories. Diesel consumption was another important factor
in freshwater eutrophication and marine eutrophication categories.
2.4. Data sources Fig. 2 shows the normalized midpoint results. Impacts attrib-
uted to climate change, human toxicity, photochemical oxidant
In this study, data from China were used when available. In case formation, particulate matter formation, terrestrial acidification,
where in data were unavailable, relevant background data (i.e., freshwater eutrophication, marine eutrophication, freshwater eco-
chemicals production and infrastructure of each scenario) from toxicity, marine ecotoxicity, and fossil depletion categories con-
Europe were used [16]. For S-1, the annual average monitoring tributed significantly to overall environmental impact. The impacts
data (i.e., direct gas emissions, quantity of wastewater, biogas observed from the remaining categories (i.e., ozone deplation, ionizing
treatment, and biogas production from sewage sludge) of the AD radiation, terrestrial ecotoxicity, agricultural land occupation, unban
stage in the Maidao wastewater treatment plant (Qingdao City, land occupation, natural land transformation, and metal depletion
Shandong Province, China) were used. Biogas production capacity categories) were negligible. Table 3 showed that infrastructure played
from FW was calculated on the basis of the annual average important roles in the abovementioned remaining categories in S-2. As
running data of sewage sludge and FW treatment site in Japan mentioned earlier, data on infrastructure and chemicals production
because of the lack information of China. The mixing rate of FW was obtained from Europe because of the lack of detailed information
and sludge in the Japanese site was 4:1. For S-2, data on AD stage in China. These results indicated that infrastructure had a low
(i.e., electricity consumption and recovery, raw materials con- contribution to overall potential impact.
sumption, and direct gas emissions) were obtained from an envi- Fig. 3 depicts the most significant substances that contribute to
ronmental report on the Qingdao FW treatment plant. The main key categories. Results indicated that for all scenarios, carbon
characteristics of FW considered in this study are shown in Table 1. dioxide, phosphate, and coal contributed most to climate change,
For S-3, data on the landfill stage (i.e., electricity consumption and freshwater eutrophication, and fossil depletion categories, respec-
recovery, raw materials consumption, leachate treatment, and tively. In photochemical oxidant formation, terrestrial acidifica-
direct gas emissions) were obtained from Ref. [17]. Furthermore, tion, marine eutrophication, and particulate matter formation
data on electricity generation and road transport in China were categories, nitrogen oxides were main contributors. Nitrate was
172 C. Xu et al. / Renewable and Sustainable Energy Reviews 49 (2015) 169–177

Table 2
Life cycle inventory of three FW treatment scenarios in China. Values are presented per functional unit.

Unit S-1a (AD of FW and sludge) S-2 b

(AD of FW) S-3c (FW to landfill )

Food waste treatment Raw materials consumption Electricity consumption kW h 1.92  102 1.30  102 7.6  10  1
Electricity recovery kW h  8.12  102  7.56  102  7.58  102
Energy recovery kJ  3.68  103  3.43  103  3.44  103
Diesel L – – 1.20
Coal kg – 32.20 –
Biogas residue kg – 7.51  102 –
Sand m3 – – 1.30
Pesticide kg – – 0.12
Water m3 – 5.49 0.35
Wastewater m3 3.56 4.48 –
LDPE m3 – – 2.40  10  4
CaO kg – – 5.77  102
Transport (Assume100 km) tkm – – 4.33  102
Direct air emissions CO2 kg 43.74 40.69 40.81
N2 kg 10.08 2.56 –
H2 kg 0.71 0.27 –
SO2 kg – – 4.63  10  2
NOx kg – – 0.26
H2S kg 0.49 0.46 4.15
Wastewater treatmentd Raw materials consumption Electricity consumption kW h 1.68  102 2.08  102 –
FeCl3 kg 16.98 22.50 –
Direct water emissions COD kg 0.75 0.92 0.12
BOD5 kg 0.17 0.22 –
SS kg 0.89 1.10 3.58  10  2
TN kg 0.15 0.18 4.63  10  2
NH3–N kg 0.12 0.15 2.90  10  2
TP kg 8.54  10  3 1.06  10  2 4.76  10  4
T–Cd kg 7.16  10  5 8.86  10  5 1.16  10  4
T–Hg kg 7.48  10  5 9.27  10  5 1.17  10  6
T–Zn kg 3.77  10  2 4.66  10  2 –
T–Cr kg 3.31  10  3 4.10  10  3 1.16  10  5
T–Ni kg 1.73  10  3 2.14  10  3 –
T–As kg 7.19  10  4 8.91  10  4 1.16  10  4
T–Pb kg 2.57  10  3 3.19  10  3 1.16  10  4
T–Cu kg 7.80  10  3 9.67  10  3 –

a
Field research at Maidao sewage treatment plant in Qingdao. Two years monitoring data at a FW treatment plant in Japan.
b
Eevironmental report of Qingdao FW treatment plant. Field research at Guangda sewage treatment plant in Jinan.
c
Hong et al. [17].
d
Field research at Guangda sewage treatment plant in Jinan.

another dominant contributor for S-3. In the ecotoxicity category, environmental impact. Fig.4 depicts the sensitivity analysis of
vanadium and nickel served an important function. Similarly, electricity recovery capacity during the landfill stage. The potential
mercury and arsenic contributed most to the human toxicity effects on the aforementioned four key categories decreased with
category. increasing electricity recovery capacity. The electricity recovery
capacity in the landfill stage ranged from approximately 156 kW
3.2. Uncertainty analysis h/t-VS to 1051 kW h/t-VS [17,22]. A linear correlation between the
electricity recovery capacity and midpoint score of the four key
To date, extremely few LCA studies have analyzed the uncer- categories was observed. This linear relation showed that a
tainty propagation of FW treatment in detail to determine the 200 kW h/t-VS increase in electricity recovery decreased human
level of confidence. In this study, uncertainty analysis was con- toxicity, freshwater eutrophication, marine ecotoxicity, and fossil
ducted by using the Monte Carlo model [20,21] to present a more depletion potential score by approximately 6.32 kg 1,4-DB eq,
reliable assessment. Table 4 presented the probability by using the 4.07  10  3 kg P eq, 0.12 kg 1,4-DB eq, and 13.2 kg oil eq in the
Monte Carlo model without energy recovery. The probability of landfill stage, respectively. A similar tendency was also observed in
S-1ZS-2 and S-2ZS-3 was 26.2% and 2.5% in the fossil depletion S-1 and S-2: a 200 kW h/t-VS increase in electricity recovery
category, respectively. Thus, the environmental effect generated decreased the potential score of human toxicity, freshwater eutro-
from S-1 was similar to S-2, but much lower than S-3. A similar phication, marine ecotoxicity, and fossil depletion by approximately
tendency was also observed for the freshwater eutrofication and 6.32 kg 1,4-DB eq, 4.88  10  3 kg P eq, 0.13 kg 1,4-DB eq, and
marine eutrofication categories. For the remaining key categories, 33.03 kg oil eq in the AD process, respectively.
the environmental effects generated from S-1, S-2, and S-3 were
similar. The uncertainty analysis results showed that S-3 had the
largest environmental impact because of the road transportation 3.3.2. Energy type
of raw materials and diesel consumption. In China, electricity generation is dominated by thermal power
and hydropower sources [23]. In 2012, the thermal-, hydro-, wind-
3.3. Energy type and recovery capacity , nuclear-, and other power (e.g., biomass, geothermy) accounted
for approximately 78.22%, 17.52%, 1.96%, 1.93%, and 0.37% of
3.3.1. Energy recovery capacity electricity generation in China, respectively. In this study, coal-
Fig. 2 shows that the human toxicity, freshwater eutrophi- based electricity generation was considered. Therefore, the env-
cation, marine ecotoxicity, and fossil depletion had the highest ironmental impacts generated from different electricity types
 C. Xu et al. / Renewable and Sustainable Energy Reviews 49 (2015) 169–177 173

Table 3
LCIA results of each scenario without energy recovery and the contribution of most significant process to the ReCiPe mid-point scores.

Catogries Unit S-1 (AD of FW and sludge) S-2 (AD of FW) S-3 (FW to landfill)

Value Process Value Process Value Process

Climate change kg CO2 2.13  102 Electricity (77%)þdirect 1.69  102 Electricity (66%) þdirect 1.81  102 CaO (41%) þ transport (34%)
eq emissions (21%) emissions (24%)
7 7 6
Ozone depletion kg CFC- 2.22  10 Electricity (94%) 3.98  10 Electricity (35%)þ 5.33  10 Transport (76%)
11 eq infrastructure (62%)
Human toxicity kg 1,4- 8.12 Electricity (99%) 7.3 Electricity (74%) þ 7.58 CaO (34%)þ transport (60%)
DB eq infrastructure (24%)
Photochemical oxidant kg 1.17 Electricity (99%) 0.81 Electricity (96%) 1.25 Transport (42%) þCaO (36%)
formation NMVOC
Particulate matter kg PM10 0.26 Electricity (99%) 0.19 Electricity (92%) 0.34 Transport (44%) þ CaO (35%)
formation eq
Ionising radiation kg U235 0.64 Electricity (95%) 1.04 Electricity (40%) þ 4.31 Transport (64%) þ CaO (28%)
eq infrastructure (58%)
Terrestrial acidification kg SO2 0.69 Electricity (99%) 0.48 Electricity (95%) 0.80 Transport (38%) þCaO (37%)
eq
Freshwater kg P eq 6.25  10  3 Electricity (99%) 4.64  10  3 Electricity (90%) 8.88  10  3 Diesel (53%) þtransport (27%)
eutrophication
Marine eutrophication kg N eq 0.14 Electricity (99%) 0.10 Electricity (96%) 0.24 Diesel (41%)þ transport
(23%)þ CaO (22%)
Terrestrial ecotoxicity kg 1,4- 9.33  10  4 Electricity (98%) 1.18  10  3 Electricity (52%) þ 4.51  10  3 Transport (65%) þCaO (25%)
DB eq infrastructure (46%)
Freshwater ecotoxicity kg 1,4- 0.17 Electricity (98%) 0.14 Electricity (83%) 0.17 Transport (63%) þCaO (31%)
DB eq
Marine ecotoxicity kg 1,4- 0.17 Electricity (98%) 0.14 Electricity (80%) 0.18 Transport (63%) þCaO (31%)
DB eq
2 2
Agricultural land m2a 2.34  10 Electricity (90%) 8.53  10 Infrastructure (82%) 0.50 Transport (80%)
occupation
Urban land occupation m2a 0.28 Electricity (79%) þ 0.59 Electricity (25%) þ 3.18 Transport (84%)
infrastructure (20%) infrastructure (74%)
Natural land m2 5.85  10  4 Electricity (93%) 6.46  10  3 Infrastructure (94%) 2.68  10  2 Transport (80%)
transformation
Water depletion m3 1.96  10  2 Electricity (88%) 7.44  10  2 Infrastructure (83%) 0.40 Transport (74%)
Metal depletion kg Fe eq 0.28 Electricity (85%) 1.81 Infrastructure (90%) 6.11 Transport (79%)
Fossil depletion kg oil eq 42.19 Electricity (99%) 44.50 Electricity (63%) þ coal (33%) 114.46 Transport (21%) þCaO (76%)

0.16
Normalized Value

AD of FW and sludge
0.12 AD of FW
FW to landfill
0.08

0.04

Fig. 2. Normalized midpoint scores for each scenario without energy recovery.

should be compared. Table 5 presents the environmental effects of 3.4. Sensitivity analysis
two electricity types with energy recovery for key categories.
Results showed that coal-based electricity exhibited a higher 3.4.1. Sensitivity to the impact assessment method
environmental benefit than mixed-type electricity in all dominant To confirm and add credibility to the current study, Impact
categories for each scenario. This result could be attributed to the 2002þ and TRACI methods were used for comparison. The com-
environmental impact generated from coal-based electricity, pared results are presented in Table 6 and S-1 was chosen as an
which was higher than mixed-type electricity (data not shown). example. The results of Impact 2002 þ and TRACI were similar to
Specifically, S-1 had the highest environmental benefit as a result those of ReCiPe in the climate change, ozone depletion, and urban
of the higher electricity generation capacity in S-1 than in S-2 and land occupation impact categories. For fossil depletion, the Impact
S-3 (Table 2). 2002þ results were consistent with ReCiPe (1.79  103 MJ) by
174 C. Xu et al. / Renewable and Sustainable Energy Reviews 49 (2015) 169–177

100%

80% Others Others

Others
60% Methane NMVOC
Sulfur
dioxide
40% Carbon Nitrogen
dioxide oxides
Nitrogen
20% oxides

0%

100%

80% Nitrogen
Others oxides
Coal
60% Nitrate
Phosphate Crude Oil
40% Others
Others
20%

0%

100% Mercury
Arsenic
Vanadium
Mercury
80% Nickel Nitrogen
Selenium oxides
Beryllium
60%
Vanadium Bromine Sulfur
dioxide
Molybdenum Selenium
40%
Arsenic Others
Lead
20% Cobalt
Others
Others
0%
S-1 S-2 S-3 S-1 S-2 S-3 S-1 S-2 S-3

Fig. 3. Contributions of the most significant substances to key categories (a) climate change; (b) photochemical oxidant formation; (c) terrestrial acidification; (d) freshwater
eutrophication; (e) fossil depletion; (f) marine eutrophication; (g) human toxicity; (h) ecotoxicity (freshwater and marine ecotoxicity); (i) particulate matter formation
(S-1: AD of FW and sludge; S-2: AD of FW; S-3: FW to landfill).

Table 4
Human toxicity (kg 1,4-DB eq)

-5

Freshwater eutrophication
Probability using the Monte Carlo model without energy recovery. y = -2.00E-05x + 1.00E-05 -5.0E-03
R² = 1
Category P(S-1 4 ¼ S-2) (%) P(S-2 4 ¼ S-3) (%) -10

(kg P eq)
Climate change 48.80 42.3 -15
Fossil depletion 26.20 2.5 -1.5E-02
Freshwater ecotoxicity 41 33 -20
Freshwater eutrophication 42.30 14
Human toxicity 32.80 42.90
-25 -2.5E-02
Marine ecotoxicity 36.50 28.70
Marine eutrophication 53.20 8.00 y = -3.16E-02x - 3.60E-03
Particulate matter formation 49.90 18.10 -30
R² = 1
Photochemical oxidant formation 53.10 25.10
Terrestrial acidification 52.30 21.90 -35 -3.5E-02

S-1: AD of FW and sludge. y = -6.07E-04x + 8.37E-04

Fossil depletion (kg oil eq)

-10 -0.2
S-2: AD of FW. R² = 1
Marine ecotoxicity

S-3: FW to landfill.
(kg 1,4-DB eq)

-0.4
-30

conversion (i.e., change rate at 42.62 MJ/kg oil eq). For particulate -0.6
matter formation, the LCIA results obtained from Impact 2002 þ
and TRACI were lower than those from ReCiPe. This difference was -50
caused by the different equivalent values of the dominant sub- -0.8
y = -6.59E-02x - 1.08E-02
stances of these methods. If PM10 was considered, the results of R² = 1
Impact 2002þ (0.28 kg PM2.5 eq) and TRACI (0.24 kg PM2.5 eq) -70 -1
were consistent with ReCiPe (0.26 kg PM10 eq). For terrestrial 100 300 500 700 900 1100
acidification, the results from Impact 2002 þ were higher than Electricity recovery (kWh/t-VS)
those from ReCiPe because ReCiPe did not consider nitrogen Fig. 4. Relationship between the electricity recovery capacity and key categories.
dioxide and sulfur oxide. For marine eutrophication, the results
from TRACI were lower than those from ReCiPe. This difference
was attributed to the higher emission factor of NOx in ReCiPe (i.e.,
1.28) than in TRACI (i.e., 0.04429). A similar tendency was also differ widely. These comparisons indicated that the ReCiPe method
observed in S-2 and S-3 (data not shown). Other LCIA results were was reliable except for the terrestrial acidification and marine
difficult to compare because other categories and label substances eutrophication categories as far as the current study is concerned.
 C. Xu et al. / Renewable and Sustainable Energy Reviews 49 (2015) 169–177 175

Table 5
ReCiPe mid-point results of two different electricity types with energy recovery for key categories.

Category Unit Coal-based electricity generation Mixed power-based electricity generation

S-1 S-2 S-3 S-1 S-2 S-3

Climate change kg CO2 eq  4.83  102  4.78  102  4.68  102  3.71  102  3.65  102  3.31  102
Human toxicity kg 1,4-DB eq  25.71  24.16  23.98  23.04  21.47  20.72
Photochemical oxidant formation kg NMVOC  3.73  3.74  3.32  2.86  2.86  2.25
Particulate matter formation kg PM10 eq  0.83  0.83  0.68  0.67  0.66  0.48
Terrestrial acidification kg SO2 eq  2.18  2.19  1.88  1.75  1.75  1.35
Freshwater eutrophication kg P eq  1.98  10  2  1.96  10  2  1.54  10  2  1.48  10  2  1.46  10  2  9.37  10  3
Marine eutrophication kg N eq  0.46  0.46  0.32  0.35  0.36  0.19
Freshwater ecotoxicity kg 1,4-DB eq  0.55  0.53  0.50  0.44  0.42  0.37
Marine ecotoxicity kg 1,4-DB eq  0.53  0.51  0.47  0.45  0.43  0.37
Fossil depletion kg oil eq  1.34  102  1.20  102  50.00  1.08  102  92.73  17.61

S-1: AD of FW and sludge.

S-2: AD of FW.
S-3: FW to landfill.

Table 6
Compared ReCiPe mid-point results for AD of FW and sludge scenario without energy recovery.

Categories Unit S-1 (AD of FW and sludge)

ReCiPe IMPACT 2002 þ TRACI

2 2
Climate change kg CO2 eq 2.13  10 1.96  10 2.11  102
Ozone depletion kg CFC-11 eq 2.22  10  7 2.22  10  7 2.22  10  7
Human toxicity kg 1,4-DB eq 8.12 – –
Photochemical oxidant formation kg NMVOC 1.17 – –
Particulate matter formation kg PM10 eq 0.26 0.16 kg PM2.5 eq 0.14 kg PM2.5 eq
Ionising radiation kg U235 eq 0.64 –
Terrestrial acidification kg SO2 eq 0.69 1.77 –
Freshwater eutrophication kg P eq 6.25  10  3 – –
Marine eutrophication kg N eq 0.14 – 6.98  10  2
Terrestrial ecotoxicity kg 1,4-DB eq 9.33  10  4 – –
Freshwater ecotoxicity kg 1,4-DB eq 0.17 – –
Marine ecotoxicity kg 1,4-DB eq 0.17 – –
Agricultural land occupation m2a 2.34  10  2 – –
Urban land occupation m2a 0.28 0.20 –
Natural land transformation m2 5.85  10  4 – –
Water depletion m3 1.96  10  2 – –
Metal depletion kg Fe eq 0.28 – –
Fossil depletion kg oil eq 42.19 1.78  103 MJ primary –

3.4.2. Sensitivity to main contributors plant size, management attention, and applied type of processes
To recognize the crucial influence on the LCIA results obtained and technologies. Furthermore, most studies focused on the
from the current study, the sensitivity analysis of dominant limited environmental impacts of FW treatment. This research
processes for key categories are presented in Table 7. For S-1 and presented the LCIA of three commonly considered FW treatment
S-2, the efficiency of electricity consumption had the highest scenarios in China.
environmental benefit in each dominant category, whereas infra- As shown in Table 3 and Fig. 2, S-3 and S-2 exhibited the
structure exhibited low variability in all categories. For S-3, the highest and lowest environmental impacts to the overall environ-
efficiency of transport had the highest environmental benefit in mental impact with and without energy recovery, respectively.
most dominant categories except for the climate change, fresh- This result was attributed to the road transportation of raw mat-
water eutrophication, marine eutrophication, and fossil depletion erials and the consumption of quicklime and diesel during the
categories. The efficiency of diesel had lower variability in most landfill stage in S-3. When energy recovery was considered,
categories except for the freshwater eutrophication and marine significant environmental benefits were observed (Table 5) in all
eutrophication categories. Similarly, the efficiency of quicklime scenarios because of biogas utilization. Biogas production capacity
consumption had low variability in most categories except for the often varied with the changes in system conditions (e.g., pH,
climate change and fossil depletion categories. Therefore, optimiz- temperature, substrate, and process). The biogas production capa-
ing the efficiency of electricity consumption and road transport are city through the AD process for different types of biomasses (i.e.,
effective ways to reduce the overall environmental impacts gen- hay, straw, bracken, grass, sugar beet, maize, grass silage, Chinese
erated from FW treatment. cabbage, sewage sludge, and FW) of different system conditions
has been widely reported worldwide [27–31] (Fig. 5). The biogas
production capacity from abovementioned biomasses approxi-
4. Discussion mately ranged from 107 m3/t-VS to 761 m3/t-VS. In this study,
the biogas production potential for S-1, S-2, and S-3 was 635, 591
LCIA of FW treatment has been extensively studied [24–26]. and 593 m3/t-VS, respectively, which was consistent with the
However, the environmental potential impact for FW treatment reported results. Fig. 5 shows that biogas produced from grass
exhibited notable differences because of numerous factors such as silage (480 to 761 m3/t-VS), Chinese cabbage (434 to 677 m3/t-VS),
176 C. Xu et al. / Renewable and Sustainable Energy Reviews 49 (2015) 169–177

Table 7
Sensitivity analysis of main contributors.

Category Variation Electricity Transport Infrastructure CaO Diesel

5% 5% 5% 5% 5%

Climate change (kg CO2 eq) S-1 8.24 – 5.70  10  3 – –

S-2 5.55 – 0.22 – –
S-3 3.26  10  2 3.12 – 3.70 4.99  10  2
Human toxicity (kg 1,4-DB eq) S-1 0.40 – 8.09  10  4 – –
S-2 0.27 – 8.92  10  2 – –
S-3 1.58  10  3 0.23 – 0.13 1.54  10  2
Photochemical oxidant formation (kg NMVOC) S-1 5.79  10  2 – 1.96  10  5 – –
S-2 3.91  10  2 – 7.94  10  4 – –
S-3 2.29  10  4 2.60  10  2 – 2.28  10  2 2.09  10  4
Particulate matter formation (kg PM10 eq) S-1 1.29  10  2 – 1.22  10  5 – –
S-2 8.70  10  3 – 5.48  10  4 – –
S-3 5.1  10  5 7.37  10  3 – 5.92  10  3 1.14  10  4
Terrestrial acidification (kg SO2 eq) S-1 3.40  10  2 – 1.55  10  5 – –
S-2 2.29  10  2 – 7.09  10  4 – –
S-3 1.34  10  4 1.50  10  2 – 1.46  10  2 3.67  10  4
Freshwater eutrophication (kg P eq) S-1 3.09  10  4 – 5.44  10  7 – –
S-2 2.08  10  4 – 1.96  10  5 – –
S-3 1.22  10  6 1.18  10  4 8.52  10  5 2.35  10  4
Marine eutrophication (kg N eq) S-1 7.16  10  3 – 2.01  10  6 – –
S-2 4.83  10  3 – 8.56  10  5 – –
S-3 2.83  10  5 2.81  10  3 2.70  10  3 4.97  10  3
Ecotoxicity (kg 1,4-DB eq) S-1 1.69  10  2 – 6.92  10  5 – –
S-2 1.14  10  2 – 2.30  10  3 – –
S-3 6.67  10  5 1.11  10  2 – 5.48  10  3 7.50  10  4
Fossil depletion (kg oil eq) S-1 2.09 – 1.13  10  3 – –
S-2 1.41 – 5.58  10  2 – –
S-3 8.24  10  3 1.22 – 4.36 0.10

S-1: ADof FW and sludge.

S-2: AD of FW.
S-3: FW to landfill.

800
S-3 (  468 kg CO2 eq) was similar. Kim and Kim [32] reported that
the climate change score for the AD, co-AD, and landfill of FW
Biogas production (m3/t-VS)

treatment in Korea were 211, 180 and 1,050 kg CO2 eq, respectively.
600
The effect seen from climate change of AD and co-AD in this study
was consistent with reported results. However, the climate change
400
score for landfill was much lower than reported result. This finding
could be attributed to the direct methane emissions used by Kim
and Kim [32] to calculate the climate change score. If the remain-
200 ing biogas was burned, the climate change score would be
134.3 kg CO2 eq, which was close to the results presented in the
current study.
0 For S-1 and S-2, electricity consumption played an important
Hay Straw Bracken Grass Sugar Maize Grass Chinese Sewage FW role in the overall environmental potential impact. In this study,
beet silage Cabbage sludge
coal-based electricity generation was considered. However, in
Fig. 5. Biogas production of different biomasses. China, thermal power (78.22%) and hydropower (17.52%) sources
are the primary dominators of electricity generation in 2012. If
sewage sludge (439 to 673 m3/t-VS), and FW (534 to 669 m3/t-VS) mixed-type electricity was applied, the environmental impact
was relatively high. Moreover, the biogas production from S-2 was generated from the human toxicity, freshwater eutrophication,
slightly higher than that from S-3. These results indicated that marine ecotoxicity, and fossil depletion categories would decrease
biogas production capacity from mixed biomasses (e.g., grass by 2.67 kg 1,4-DB eq/t-VS, 5  10  3 kg P eq/t-VS, 0.11 kg 1,4-DB eq/
silage and FW, Chinese cabbage and FW) might also increase. t-VS, and 26.00 kg oil eq/t-VS for S-1 and S-2, respectively
Therefore, studies about these aspects should be further tested in (Table 5). A similar tendency was also observed for S-3. These
the future. results indicated that changing the energy structure of China was
Hong et al. [17] reported that electricity recovery capacity important for reducing the overall environmental effects. The
in the landfill stage ranged from approximately 156 kW h/t-VS Chinese government currently focuses on increasing the ratio of
to 1051 kW h/t-VS. The electricity regeneration capacity was clean energy (e.g., hydropower, solar power). During the Chinese
758 kW h/t-VS, thus indicating a 27.9% improvement potential to “12th Five-Year Plan” period (2011–2015), the ratio of hydropower
regenerate electricity during landfill stage. The climate change increased to 22.36%. Therefore, continuing the implementation of
result for S-1 (213 kg CO2 eq) was higher than S-2 (169 kg CO2 eq) this policy is important.
and S-3 (181 kg CO2 eq) without energy recovery because of the In China, FW is generally mixed in MSW, which is commonly
high electricity consumption of S-1. However, when energy treated by landfill (72.55%) and incineration (24.74%) technologies
recovery was considered, LCIA results revealed that the climate [33]. However, only few landfill sites are considered safe. Most
change score for S-1 (  483 kg CO2 eq), S-2 (  478 kg CO2 eq), and MSW are simply filled [17]. Accordingly, the leachate and use of
 C. Xu et al. / Renewable and Sustainable Energy Reviews 49 (2015) 169–177 177

biogas generated from landfill was extremely low. Furthermore, category indicators at the midpoint and the endpoint level. First edition report
the uncertainty analysis results indicated that landfill was unsui- I: characterization, available from: 〈http://www.lciarecipe.net〉; 2009.
[10] De Schryver A, Brakkee K, Goedkoop M, Huijbregts M. Characterization factors
table for treating FW. Hong et al. [17] reported that the electricity for global warming in life cycle assessment based on damages to humans and
recovery capacity of incineration was low (287 kW h/t-MSW). ecosystems. Environ Sci Technol 2009;43(6):1689–95.
Moreover, Kim et al. [34] presented that the total cost of incinera- [11] Goedkoop M Spriensma R. The ecoindicator 99: a damage oriented method for
life cycle impact assessment. Amersfoort, The Netherlands: PRE consultants,
tion was higher than co-AD and landfill scenarios. Therefore, in the
〈http:www.pre.nl/eco-indicator99/〉; 2000.
current situation, the separate disposal of FW from MSW and the [12] Guinée J Gorrée M Heijungs R Huppes G Kleijin R de Koning A. Life cycle
centralized disposal by the AD process are highly recommended. assessment; an operational guide to the ISO standards, Part 3: Scientific
background. Den Hag and Leiden, the Netherlands: Ministry of Housing,
Spatial Planning and Environment (VROM) and Centre of Environmental
5. Conclusion Science (CML), 〈http://www.leidenuniv.nl/cml/ssp/projects/lca2/lca2;2001.
html〉; 2001.
[13] Available from: 〈http://www.cml.leiden.edu/research/industrialecology/
This study evaluated three commonly used FW treatment researchprojects/finished/recipe.html〉.
scenarios in China and identified the contributions of FW treat- [14] Jolliet O, Margni M, Charles R, Humbert S, Payet J, Rebitze G, Rosenbaum R.
IMPACT 2002þ : a new life cycle impact assessment methodology. Int J Life
ment scenarios and their potential for improvement. The results Cycle Assess 2003;8:324–30.
indicated that the most suitable environmental scenario for treat- [15] Bare J, Norris G, Pennington D, McKone T. TRACI—the tool for the reduction
ing FW was S-2 (i.e., AD of FW) through the entire life cycle. By and assessment of chemical and other environmental impacts. J Ind Ecol
contrast, S-3 (i.e., FW to landfill) had the largest environmental 2003;6:56–68.
[16] Ecoinvent centre, 2010. Swiss centre for life cycle inventories, available from:
impact. Uncertainty analysis results also proved these results. The 〈http://www.ecoinvent.org〉.
biogas production capacity reported in this study was consistent [17] Hong J, Li X, Cui Z. Life cycle assessment of four municipal solid waste
with previously reported results worldwide. FW had higher biogas management scenarios in China. Waste Manage 2010;30:2362–9.
[18] Cui X, Hong J, Gao M. Environmental impact assessment of three coal-based
generation capacity than other biogases. The useful information electricity generation scenarios in China. Energy 2012;45:952–9.
identified in this study (i.e., the LCI, potential impacts, energy [19] Chen W Hong J Xu C. Pollutants generated by cement production in China,
recovery capacity, and key factors) will significantly help enrich their impacts, and the potential for environmental improvement. J Cleaner
Prod, 2014. Article in press. Available from: http://dx.doi.org/10.1016/j.jclepro.
the LCI database of FW treatment and help policymakers make
2014.04.048.
decisions regarding FW treatment plant problems. However, other [20] Hertwich E, McKone T, Pease W. A systematic uncertainty analysis of an
FW upgrading technologies such as composting and by-product evaluative fate and exposure model. an official publication of the Society for
production also need to be discussed to avoid pollutant transfer Risk Analysis. Risk Anal 2000;20:439–54.
[21] Huijbregts M, Gilijamse W, Ragas A, Reijnders L. Evaluating uncertainty in
and secondary pollution. We will conduct further LCA compar- environmental life-cycle assessment a case study comparing two insulation
isons with other FW treatment scenarios in the future. options for a Dutch one-family dwelling. Environ Sci Technol 2003;37:2600–8.
[22] Xu C, Chen W, Hong J. Life-cycle environmental and economic assessment of
sewage sludge treatment in China. J Cleaner Prod 2014;67:79–87.
Acknowledgments [23] Department of Energy Statistics. China energy statistical yearbook. National
Bureau of Statistic of China; 2013. 〈http://free.xiaze.com/nianjian/zgnytjnj2013/〉
Available from.
We gratefully acknowledge financial support from the Institute [24] Lundie S, Peters G. Life cycle assessment of food waste management options. J
of plateau meteorology, CMA, Chengdu, China (LPM2014002), Cleaner Prod 2005;13:275–86.
[25] Khoo H, Lim T, Tan R. Food waste conversion options in Singapore: environ-
China Energy Conservation and Emission Reduction Co. Ltd (GJN- mental impacts based on an LCA perspective. Sci Total Environ
14-07), and The Fundamental Research Funds of Shandong Uni- 2010;408:1367–73.
versity (2015JC016). [26] Vandermeersch T, Alvarenga R, Ragaert P, Dewulf J. Environmental sustain-
ability assessment of food waste valorization options. Resour Conserv Recycl
2014;87:57–64.
References [27] Pakarinen O, Lehtomaki A, Rissanen R, Rintala J. Storing energy crops for
methane production: effects of solids content and biological additive. Bior-
esour Technol 2008;99:7074–82.
[1] Chen E, Gu X. Disposal and recycling technology of food waste. Environ Study
[28] Caporgno M, Trobajo R, Caiola N, Ibanez C, Fabregat A, Bengoa C. Biogas
Monit 2012;3:57–61.
production from sewage sludge and microalgae co-digestion under mesophi-
[2] Xu D, Shen D, Feng H. Discussion on characteristics and resource recycling
technology of food residue. Bull Sci Technol 2011;27:130–5. lic and thermophilic conditions. Renewable Energy 2015;75:374–80.
[3] ISO (International Organization for Standardization). Environmental [29] Demirel B, Scherer P, Yenigun O, Onay T. Production of methane and hydrogen
management-life cycle assessment-general principles and framework. Geneva, from biomass through conventional and high-rate anaerobic digestion pro-
Switzerland: ISO; 2006. p. 14040. cesses. Crit Rev Environ Sci Technol 2010;40(2):116–46.
[4] Franchetti M. Economic and environmental analysis of four different config- [30] Kafle G, Bhattarai S, Kim S, Chen L. Effect of feed to microbe ratios on
urations of anaerobic digestion for food waste to energy conversion using LCA anaerobic digestion of Chinese cabbage waste under mesophilic and thermo-
for: a food service provider case study. J Environ Manage 2013;123:42–8. philic conditions: biogas potential and kinetic study. J Environ Manage
[5] Ebner J, Babbitt C, Winer M, Hilton B, Williamson A. Life cycle greenhouse gas 2014;133:293–301.
(GHG) impacts of a novel process for converting food waste to ethanol and co- [31] Lin J, Zuo J, Li P, Liu F, Wang K, Chen L, Gan H. Effects of mixture on anaerobic—
products. Appl Energy 2014;130:86–93. digestion with fruit and vegetable waste and food waste of China. J Environ Sci
[6] Yi S, Yoo K. Greenhouse gas emissions and cost analyses for the treatment 2011;23(8):1403–8.
options of food waste and human excrement. J Environ Prot 2014;5:597–607. [32] Kim M, Kim J. Comparison through a LCA evaluation analysis of food waste
[7] IEA (International Energy Agency). World energy outlook, 〈http://www.iea. disposal options from the perspective of global warming and resource
org/techno/etp/etp10/Chinese_Executive_Summary.pdf〉; 2010. recovery. Sci Total Environ 2010;408:3998–4006.
[8] Lee S, Choi K, Osako M, Dong J. Evaluation of environmental burdens caused by [33] China Statistical Yearbook. National Bureau of Statistic of China, available
changes of food waste management systems in Seoul, Korea. Sci Total Environ from: 〈http://www.stats.gov.cn/tjsj/ndsj/2013/indexch.htm〉; 2013.
2007;387:42–53. [34] Kim M, Song Y, Song H, Kim J, Hwang S. Evaluation of food waste disposal
[9] Goedkoop M Heijungs R Huijbregts M. De Schryver A Struijs J. Van Zelm R options by LCC analysis from the perspective of global warming: Jungnang
ReCiPe. A life cycle impact assessment method which comprises harmonised case, South Korea. Waste Manage 2011;31:2112–20.