Journal of Cleaner Production 131 (2016) 52e59

Contents lists available at ScienceDirect

Journal of Cleaner Production

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

Environmental impacts of reclamation and recycling processes of

refrigerators using life cycle assessment (LCA) methods
Rufeng Xiao, You Zhang, Zengwei Yuan*
State Key Laboratory of Pollution Control and Resources Reuse, School of the Environment, Nanjing University, Nanjing 210023, China

a r t i c l e i n f o a b s t r a c t

Article history: China has enacted a series of policies and regulations that require manufacturers to reclaim and recycle
Received 30 May 2015 obsolete refrigerators to mitigate the life cycle environmental impacts and to improve resource effi-
Received in revised form ciency. However, it is not clear whether the environmental benefits of recycling refrigerators can be
9 April 2016
balanced with the emissions from the reclamation and recycling processes. To address this issue, envi-
Accepted 15 May 2016
ronmental impacts of recycling refrigerators under different scenarios were quantified through a
Available online 24 May 2016
comparative life cycle assessment. The data were mainly acquired from a large professional refrigerator
recycling company. The CML 2001 method built in the GaBi software (Version 6.0) was used to quantify
Keywords:
Life cycle assessment
the environmental impacts. The result shows that the environmental benefits brought about by recycling
Refrigerator mainly come from the resource and energy savings in the upstream production, such as in the high
Recycling impact polystyrene recycling (22.17%), the steel recycling (23.94%), and the copper recycling (8.10%)
Environmental impacts phases. Compared to railway, motor is a more environmentally friendly transportation for refrigerators.
Environmental management Recycling refrigerators increase the ozone layer depletion potential because the crushing process inev-
itably releases the CFC-11. The environmental impacts of transportation may exceed the environmental
benefit of recycling refrigerators. This study will be useful for manufacturers to design new recycling
networks from the life cycle and environmental impact perspectives.
© 2016 Elsevier Ltd. All rights reserved.

1. Introduction Most published articles have emphasized the development of

In the past decade, China increased its refrigerator production
capacity with an annual growth rate of 5% and produced 93.37
million refrigerators in 2014 alone (National Bureau of Statistics of
China, 2014). Consequently, the waste refrigerators also increased
greatly and the highly valuable materials in waste refrigerators
have stimulated the unregulated recycling activities, which caused
severe pollution on the local soil and groundwater. To regulate the
reclamation and recycling activities and mitigate the pollution,
China has enacted a series of policies and regulations. These leg-
islations led to more environmentally friendly activities within the
specific reclamation or recycling stages regarding some environ-
mental impact categories (Zhou and Gao, 2007); however, the
question remains as to whether the environmental benefits of
recycling refrigerators can be balanced with the emissions from the
reclamation and recycling processes.
* Corresponding author. Tel.: þ86 025 89680532.
E-mail address: yuanzw@nju.edu.cn (Z. Yuan).
disassembly technologies for recovering the materials from waste
refrigerators rather than discussing the environmental impacts. For
example, Altekin et al. (2008) reported the development of a
refrigerator disassembly chain. An environmentally friendly pro-
duction line was reported in China for recovering the cabinet of
waste refrigerators with the recovery rate of 97.6% (Ruan and Xu,
2011). To securely recover valuable materials, Mitsubishi Electric
Corporation adopted a disassembling and classification process
before throwing waste refrigerators into crushing machines to
improve the purity of the recovered materials and to reduce the
harm to landfill sites (Kotera et al., 1999). Laner and Rechberger
(2007) summarized the European state-of-the-art recycling tech-
nologies of refrigerators. These studies considered the efficiency of
recycling resources but concerned little about the environmental
impacts of recycling systems. Some studies further discussed the
economic costs and the environmental benefits to improve the
effectiveness of policy-making for recycling waste refrigerators
(Lambert and Stoop, 2001; Nicol and Thompson, 2007).
As an effective way to analyze the life cycle environmental
performance of products, the life cycle assessment (LCA) was used

http://dx.doi.org/10.1016/j.jclepro.2016.05.085
0959-6526/© 2016 Elsevier Ltd. All rights reserved.
 R. Xiao et al. / Journal of Cleaner Production 131 (2016) 52e59 53

to quantify the environmental impacts of refrigerators (Kim et al.,
2006; Gehin et al., 2009). However, these studies either consid-
ered only specific processes or stages, or compared the environ-
mental impacts of refrigerators manufactured with different
technologies and materials. For example, the phosphating and
foaming processes result in environmental pollution. The former
discharges wastewater with phosphorus and can eutrophicate local
surface waters and the latter emits cyclopentane gas at a rate of
0.03 kg/h and is harmful to human health (Ruan and Xu, 2011;
Zhang, 2010; Chen, 2011; Liu, 2011). The energy consumption and
greenhouse gas emissions of refrigerators were also compared
between two different foamers, HFC-245fa (pentafluoropropane)
and pentane, as well as the life-cycle climate performance (LCCP) of
different refrigerating fluids based on measured data (McCulloch
and Lindley, 2003; Johnson, 2004; Yunho et al., 2007). Consid-
ering that the components contain heavy metals, they can pollute
air, soil, and water if the waste refrigerator is arbitrarily discarded
or sent to landfill directly at the end of a refrigerator's life (Liu and
Chi, 2013). The environmental consequences of reclamation and
recycling waste refrigerators across different regions still remain
unknown. This study aims to quantify the environmental impacts
of refrigerator reclamation and recycling systems and to examine
the rational reclamation distance in terms of environmental
benefits.
2. Methodology
This study does not cover the whole life cycle of the re-
frigerators, but the same methods used in LCA are used in it. The
report was only examined by coauthors and not tested by experts
outside. The study is completed with four steps including the goal
and scope definition, the life cycle inventory (LCI) analysis, the life
system boundary is shown in Fig. 1. The refrigerator is transported
from the collection site in Nanjing, Jiangsu province, to the disas-
sembly and recycling plant in Qingdao, Shandong province. In the
plant, refrigerants, compressors, and accessories are manually
separated firstly, and the compressors are used for direct, second-
hand sale. The remaining parts such as the metals and plastics
are further separated by follow-on processes. The plastics are
granulated and regenerated in the factory, then sold to refrigerator
manufacturing enterprises. The recovered metals (copper, iron and
aluminum) are sold to metallurgical enterprises. Polyurethane
(PUR), a type of hazardous solid waste, is incinerated then trans-
ported to a landfill. The recovered energy from gases generated by
incineration of PUR is used for upstream production of steel plate.
Previous research shows that the recovery and recycling of waste
materials and energy brings environmental benefits (Zhang, 2010;
Li, 2007).
2.3. Data sources and data quality requirements
2.3.1. Data sources
Types and sources of data may include a mixture of measured,
calculated or estimated data. In the first stage of life cycle, the
transportation distance of a refrigerator from Nanjing city to
Qingdao city is calculated with the assistance of Google Maps. We
enter the location names of the departure and the destination, and
the distance will be calculated automatically, which is 573 km by
road or 884.1 km by rail respectively in the Google Maps. In the
second stage of life cycle, we investigated Qingdao Chunqi
renewable resources Co. Ltd, a large professional refrigerator
System boundary

cycle impact assessment (LCIA), and the life cycle improvement

analysis and interpretation. We also took some methods (Guinee,
2002) to ensure that environmental impacts are assessed within
Waste
a consistent framework to avoid “problem shifting”. Refrigerator

2.1. Functional unit

Transportation
The functional unit is defined as a “1 kg waste refrigerator made
in China.” The direct-cooling double-door waste refrigerator was
selected because this type of refrigerator is the most popular, ac- Compressors
counting for nearly half of the market share, and because direct- Manual and
dismantling
cooling refrigerators hold 90.3% of the market share of double- accessories

door refrigerators (Internet Consumer Research Center, 2014).

2.2. System boundary Crushing

Material
Emissions
Energy
The recovery and disposal technologies for discarded re-
frigerators are still being developed in China (Li, 2007). To recover Magnetic
Steel
separation
waste refrigerators, private vendors go into communities to collect
obsolete refrigerators from the residents with little compensation;
the refrigerators are then sold to professional disposal firms which PUR
refurbish or dismantle them. We investigated a large professional Incineration and
Winnowing
landfill
refrigerator recycling company in Qingdao Shandong province and
acquired onsite information about the recycling process. The
company was selected primarily based on its scale and technical
Eddy current Aluminum&
level which was encouraged and supported by the Ministry of separation Copper
Environmental Protection to be the representative for recycling
white appliances. On the other hand, Jiangsu province has a large
amount of inventory and scrappage of refrigerator and ranked
HIPS
fourth nationwide below Shandong province, Sichuan province,
and Henan province (National Bureau of Statistics of China, 2014).
The choice of the system boundaries is important and affects the
LCA results (Guinee, 2002). Based on the research objectives, the Fig. 1. System boundary for the LCA study.
54 R. Xiao et al. / Journal of Cleaner Production 131 (2016) 52e59

recycling enterprise, to acquire the onsite information including bill Table 2

of materials and energy consumption on the recycling process. Recovery rate of the refrigerator parts.

Published data were used in the inventories as the secondary data Category Recovery rate Data source
to estimate the ratio of steel and copper in compressor (Zhang, Steela 90% Measured data
2010; Li, 2007), the ratio of copper and aluminum in the mixed High impact polystyrene (HIPS) 95% Measured data
metal and pollutant emission (Zhang, 2010; Li, 2007). Table 1 shows Polyurethane (PUR) 0% Measured data
the material balance of the waste refrigerator which is an ideal Rubber 100% Measured data
Metala (copper, aluminum) 90% Estimated data
situation with recovery rate of 100%. But in fact, due to the technical
Othersa Unknown
and worker's operating mistakes, there is a material loss in the
a
Metal does not contain the compressor that is sold directly on second-hand
recovery process. This study did not consider the electronic control
markets. Others refer to the no data material in Table 1.
system which were included within the missing data. Due to the
data on the input and output of materials was inaccessible, the
pollution and environmental benefits of the electronic control c) technology coverage: data based on existing large-scale
system do not be considered. technology;
As an adjunct, CFC11 is produced and wrapped in the PUR d) precision: data deviation of the two years, see Table 1;
foam. It is released into the environment again when the PUR e) completeness: the ratio of measured data to estimated data
foam is broken. In crushing process, not only PUR but all materials is 46 to 12. The data in square brackets of Table 1 refers to
of refrigerator are crushed together. Mass properties were measured data, and the data in parentheses of Table 1 refers
considered to estimate the CFC11 emission in crushing process. to estimated data. Considering that both mass of the mate-
The emission load of CFC11 was calculated with the emission rials and the fact that some materials are associated with
factor of CFC11 per refrigerator (Yang, 2006). Then it was allo- much higher impacts in some categories, some materials
cated according to the ratio of the mass of PUR in the total input. seem to be important for the benefit of recycling. For
The detailed results and the description of the data collection can example, although the CFC11 is estimated with the mass of
be found in Tables 1 and 2. 2 g, it is an important material for ODP category as it clearly
offsets the benefit to ODP from the recycling of steel and
other materials;
2.3.2. Data quality
f) reproducibility: based on the CML, bill of materials, energy
Descriptions of data quality are important to understand the
consumption and pollutant emission would allow an inde-
reliability of the study results and properly interpret the outcome
pendent practitioner to reproduce the results reported in the
of the study. The data quality requirements include the
study;
following:
g) sources of the data: see Section 2.3.1;
h) uncertainty of the information: see Section 4.2 limitation.
a) time-related coverage: The majority of data were derived
from the period 2013 to 2014 and is believed to be repre-
sentative of refrigerator recycled in Shandong Province dur- 2.4. Allocation procedures
ing this time frame;
b) geographical coverage: Shandong Province, China. Qingdao Allocation has been discussed by Cederberg and Stadig
Chunqi renewable resources Co. Ltd, a large professional (2003). Different allocation methods are used in LCA studies
home appliance recycling enterprise, represents over 40% of when more than one valuable output is yielded. Examples are:
recycled refrigerator in Shandong Province; mass based allocation (the heavier product is assigned with more

Table 1
Material balance of the waste refrigerator.

Processes Input (kg) Output (kg) Electricity Waste gas (kg)

(kW h)

Manual Total Compressor Compressor Refrigerant Door seal Magnetic The rest 0.082 [0.7%] Refrigerant
disassembly (steel) (copper) (isobutane) (rubber) powder (Isobutane)
65.047 4.521 1.258 (8.3%) 0.005 (0%) 0.287 [3.5%] 0.479 [2.1%] 58.452 0.045 (0%)
[3.1%]a (8.3%) [3.5%]

Crushing Total Cabinet Door Shelves & Evaporation Condensation Missing The rest 6.876 [0.1%] CFC11
drawers module module data
58.452 39.401 9.514 [1.1%] 2.182 [4.7%] 2.611 [3.9%] 2.221 [4.6%] 2.521 58.450 0.002 (0%)
[3.5%] [2.6%] [4.0%] [3.5%]

Magnetic Total Steel The rest 0.264 [0.8%]

separation 58.450 24.439 34.011
[3.5%] [4.2%] [6.1%]

Winnowing Total Polyurethane The rest 0.467 [0.4%]

(PUR)
34.011 10.684 23.327
[6.1%] [9.8%] [9.0%]

Eddy current Total Aluminum High impact Copper The rest 0.312 [0.6%]
separation polystyrene (HIPS)
23.327 6.096 (15.6%) 10.162 [1.1%] 7.069 (15.6%) 0
[9.0%]
a
The value refers to data deviation of two years. The square brackets refer to measured data. The parentheses refer to estimated data.
 R. Xiao et al. / Journal of Cleaner Production 131 (2016) 52e59
burdens), economic based allocation (the more expensive prod-
uct is assigned with more burdens), and energy based allocation
(the product with more energy content is assigned with more
burdens). In addition, allocation problem does not exclusively
arise in the case of a process with more than one product.
Considering the data acquisition, the mass based allocation is the
preferred method as it provides the most consistent solution for
this study.
(1) Here system-expansion is used for avoiding allocation
problem. First, the recovered energy from gases generated by
incineration of PUR is used for upstream production of steel
plate, then, the steel plate is used for producing refrigerators
again. So all the environmental burdens of incineration of
PUR are allocated to recycling phase of the refrigerator, and
the steel production takes the credit of the recovered energy
from incineration of PUR.
(2) In the subsequent recycling process, it is needed to allocate
environmental benefits based on the mass of the copper and
aluminum to know how much the copper and aluminum
each contribute to the recycled resources.
2.5. Assumptions
Two conveyances were used to reflect the different refrigerator
journeys. The first one is a train with a carrying capacity of 2000-
tonnes, and the second transportation method is a truck on a car-
rying capacity of 24-tonnes. Additionally, six different scenarios
were examined, as shown in Table 3. The LCI data of the refrigerator
transportation are acquired from the Ecoinvent database v 2.2 built
in the GaBi software (Weidema and Hischier, 2006; PE
International, 2012). The environmental impacts are quantified by
the multiplication of the average product mass (in tonnes) of a
truck or a train and the distance (km) which must be traveled to
transport these refrigerators.
Scenario 0 refers to the whole refrigerator is landfilled in
Nanjing city without any recycling process. Emission factors come
from Ecoinvent database.
Scenario 1 refers to the dataset of China: “operation, coal freight
train, and diesel”. The value of the diesel use per tonne-km is ob-
tained from the 2013 China Statistical Yearbook (National Bureau of
Statistics of China, 2014). Emission factors come from Ecoinvent
database.
Scenario 2 refers to the dataset of Europe: “operation, coal
freight train, hybrid drive by electricity and diesel”. Electricity
use per tonne-km is also obtained from the 2013 China Statis-
tical Yearbook. The diesel fuel used for shunting per tonne-km
is taken from the European dataset because no such informa-
tion exists in China. Emission factors come from Ecoinvent
database.
Inventories of Scenarios 3e5 refer to the following respective
datasets: “transport, lorry 16e32 t, EURO 3 (the Europe 3 emission
standard)”, “transport, lorry 16e32 t, EURO 4 (the Europe 4 emis-
sion standard)” and “transport, lorry 16e32 t, EURO 5 (the Europe 5
Table 3
Scenarios for the transportation of a refrigerator.
Scenario Conveyance Energy Source of emission factor
S0 None Hard coal Ecoinvent database
S1 Train Diesel Ecoinvent database
S2 Train Electricity Ecoinvent database
S3 Truck Diesel EURO 3
S4 Truck Diesel EURO 4
S5 Truck Diesel EURO 5
55
emission standard).” Here, it is assumed that the trucks use diesel
fuel on any condition of emission standards. The consumption of
diesel per tonne-km is obtained from the 2013 China Statistical
Yearbook.
2.6. Impact categories and impact assessment methodology
Various LCIA methods have been developed and are currently
available in the market database of LCA-dedicated software. The
methods include the environmental design of industrial products
(EDIP), a tool for the reduction and assessment of chemical and
other environmental impacts (TRACI), ecological scarcity, etc.
(Cavalett et al., 2013). The CML is a problem-oriented LCA method
developed by the Institute of Environmental Sciences of the Uni-
versity of Leiden, which aims to offer best practices for midpoint
indicators and operationalizes the ISO 14040 series of standards.
Three main impact categories are examined in this study. The
first category is the total raw material or resources consumption,
measured by abiotic depletion (ADP elements and ADP fossil). The
second category consists of the atmospheric and water impacts.
The atmospheric impacts are further divided into four sub-
categories, which include the acidification potential (AP), the global
warming potential (GWP), the photochemical ozone creation po-
tential (POCP), and the ozone layer depletion potential (ODP). The
water impacts are measured by the eutrophication potential (EP).
The last category is toxicity, which includes the freshwater aquatic
ecotoxicity (FAETP), the human toxicity potential (HTP), the marine
aquatic ecotoxicity (MAETP), and the terrestrial ecotoxicity poten-
tial (TETP).
3. Results
3.1. LCIA results of different transportation scenarios
Table 4 shows the characteristic values of the reclamation and
recycling processes of a refrigerator in different transportation
scenarios. As recycling of materials such as iron, copper etc. can
bring environmental benefits, the values of some environmental
categories are negative. Environmental benefits of 7 impact cate-
gories (AP, EP, ADP elements, FAETP, HTP, POCP and TETP) for sce-
nario 1 are 1e1.5 times higher than scenario 2. Furthermore, for the
GWP and MAETP, scenario 1 brings the environmental benefits
while scenario 2 causes the environmental pollution. The back-
ground process of electricity is from Ecoinvent database (CN:
electricity, production mix CN [production mix]) which is based on
Chinese situation. So the electricity we considered in this study is
the mix electricity grid. According to the data from China Electricity
Council, the electricity grid is dominated by coal-fired power (79%)
and hydropower generation only occupies 17%. On this basis,
burning diesel is cleaner than coal-fired power, so the carbon di-
oxide produced by an electric train (Scenario 2) is much more than
what is produced by an internal combustion engine train (Scenario
1). Especially, a high-speed rail with more electricity consumption
leads to more greenhouse gas emission than an internal combus-
tion engine train.
For scenario 3, 4 and 5, most categories are similar. With the
improvement of emission standards, the benefits to the environ-
ment are more obvious: for scenario 5, the GWP and ADP fossil
decreased by 77% and 62% compared to scenario 3. Also, it can be
found that motor transport is more environmentally friendly than
rail transport in general. For example, in consideration of GWP,
transporting a refrigerator 884.1 km by rail results in average car-
bon dioxide emissions of 38.3 kg equivalents. Similarly, trans-
porting a refrigerator 573 km by truck results in average carbon
dioxide emissions of 13.1 kg equivalent. Thus, for 1 km, the truck
56 R. Xiao et al. / Journal of Cleaner Production 131 (2016) 52e59

Table 4
LCIA for a 1 kg waste refrigerator based on the CML 2001 method.

Category S0 S1 S2 S3 S4 S5 Unit

GWP 3.30E02 2.11E01 2.94E01 2.49E01 3.46E01 4.41E01 kg CO2-Equiv.

AP 1.28E04 1.36E02 8.50E03 1.56E02 1.59E02 1.62E02 kg SO2-Equiv.
EP 9.41E05 1.53E02 1.48E02 1.58E02 1.59E02 1.60E02 kg P-Equiv.
ODP 1.94E11 3.29E08 3.98E08 3.81E08 2.28E08 7.80E09 kg R11-Equiv.
ADP elements 7.92E09 8.00E05 7.99E05 7.93E05 7.95E05 7.98E05 kg Sb-Equiv.
ADP fossil 4.90E01 4.74Eþ00 1.91E01 4.80Eþ00 6.29Eþ00 7.77Eþ00 MJ
FAETP 1.21E04 5.80Eþ00 5.72Eþ00 5.78Eþ00 5.79Eþ00 5.80Eþ00 kg DCB-Equiv.
HTP 1.26E03 6.09E01 4.73E01 6.09E01 6.23E01 6.37E01 kg DCB-Equiv.
MAETP 4.11Eþ00 4.60Eþ02 7.80Eþ01 4.27Eþ02 4.47Eþ02 4.68Eþ02 kg DCB-Equiv.
POCP 1.69E05 7.87E04 5.15E04 1.03E03 1.09E03 1.14E03 kg Ethene-Equiv.
TETP 8.03E04 1.47E02 1.33E02 1.42E02 1.44E02 1.47E02 kg DCB-Equiv.

Positive values refer to environmental pollution, and negative values refer to environmental benefits.

results in 0.023 kg CO2-Equiv versus 0.043 kg CO2-Equiv by train.
The reason is that the truck uses diesel, while the train consumes
electricity from thermal power plants. In order to produce the same
amount of power, however, the train results in more emissions of
greenhouse gases than the truck.
It is worth noting that under the existing recycling conditions,
recycling refrigerators causes ozone layer depletion potential
because the crushing process inevitably releases the CFC-11. The
polyurethane (PUR) foam is chemical combination formed by the
foamer, while trichlorofluoromethane (CFC-11) as an adjunct is
produced and wrapped in the PUR foam. And CFC-11 is released
into environment when the PUR foam is broken (Yang, 2006). And
due to S0 is based on Ecoinvent database which do not consider the
rupture of PUR foam, by comparing S0 with other scenarios, it can
be found that the recycling of refrigerator is more environmentally
friendly than landfill except ODP category. Considering the poly-
urethane foam also eventually break in the landfill, S0 results in
more ozone depletion than other scenarios.
3.2. LCIA contribution analysis
To further interpret the results and to examine the contribution
from different unit processes, an impact contribution analysis of
scenario 2 is conducted. The treatment system consists of
transportation (P1), steel recycling (P2), high impact polystyrene
recycling (HIPS) (P3), aluminum recycling (P4), copper recycling
(P5), PUR treatment and disposal (P6), electricity consumption in
the disposal stage (P7), and compressor recycling (P8). The relative
contributions to each impact category attributable to the recla-
mation and recycling processes of refrigerators are shown in Fig. 2.
The indicators are grouped and analyzed below.
The largest portion of the environmental benefit on the ADP
(elements) and FAETP lies in copper recycling (P5) (97.94% and
93.93%, respectively), followed by steel recycling (P2) (1.51% and
3.77%, respectively). The savings of coal, steam, water, and other
materials in the upstream production processes of copper and steel
contribute to the environmental benefit for ADP (elements).
The dominant contributors to the environmental benefit on the
GWP category are the steel recycling (P2) (17.79%), the HIPS recy-
cling (P3) (16.47%) and the copper recycling (P5) (6.02%) phases,
which together account for over 40% of the total GWP impact. The
largest portion of the GWP burden is primarily contributed by the
transportation phase (P1) (41.94%) caused by either direct CO2
emissions when hard coal is burned on-site to produce steam or by
indirect CO2 emissions from the on-site electricity consumption.
The dominant contributor to the environmental benefit on the
AP category is copper recycling (P5), which accounts for over 50% of
the total benefit. Copper recycling (P5) can reduce the contributors

80%

60%

40%

20%

0%

-20%

-40%

-60%

-80%

-100%
GWP AP EP ODP ADP ADP fossil FAETP HTP MAETP POCP TETP
elements

P1 P2 P3 P4 P5 P6 P7 P8

Fig. 2. Relative contributions to each impact category attributable to the recycling of a refrigerator. (Positive values refer to environmental pollution, and negative values refer to
environmental benefits. Here P1 refers to the treatment system consists of transportation, P2 refers to steel recycling, P3 refers to high impact polystyrene recycling (HIPS), P4 refers
to aluminum recycling, P5 refers to copper recycling, P6 refers to PUR treatment and disposal, P7 refers to electricity consumption in the disposal stage, and P8 refers to compressor
recycling.)
 R. Xiao et al. / Journal of Cleaner Production 131 (2016) 52e59
to AP burdens mainly caused by SO2 and NOx emissions, which are
either directly emitted from coal burning in the smelting stage or
indirectly emitted by electricity consumption in the upstream
production stage. Concurrently, the transportation (P1), PUR
treatment and disposal (P6) and electricity consumption in disposal
(P7) stages cause environmental pollution on the AP category, ac-
counting for 25.68%, 7.03%, and 4.08% of the total, respectively.
Copper recycling (P5) accounts for 71.68% of the environmental
benefit on the total EP of the refrigerator, followed by steel recycling
(P2), which contributes 2.67% to the total. The recycling of metals
comes from reducing in a high load of phosphorus, nitrates, and the
chemical oxygen demand (COD) in wastewaters, thus causing
environmental impacts. The dominant contributor to environ-
mental pollution on the EP category is PUR treatment and disposal
(P6), which accounts for 20.18% of the total category impact.
The contribution to ODP is predominantly the result of PUR
treatment and disposal (P6), which accounts for 26.89% of the total
ODP. The crushing and incineration of PUR releases chlorofluoro-
carbons (CFCs) wrapped in a plastic bubble, which can be potential
ozone depleting substances. Additionally, non-methane volatile
organic compounds (NMVOCs) from reclaimed and recycled
refrigerator transportation (P1) accounts for 38.12% of the total and
can lead to ozone layer depletion. Steel recycling (P2) accounts for
over 21.49% and brings about environmental benefit.
Similar to the ODP category, the environmental benefits of ADP
fossils come from steel recycling (P2), which accounts for over
17.07%; the environmental pollutions of ADP fossils come from
transportation (P1) and PUR treatment and disposal (P6), which
accounts for over 29.47% and 16.31%, respectively. HIPS recycling
(P3) accounts for over 24.49% of the total and provides environ-
mental benefits caused by the fossil resources saving in the up-
stream production of HIPS.
The largest portion of the three toxicity indicators (HTP, MAETP,
and TETP) is dominated by the steel recycling (P2), which accounts
for 40.53%, 30.43%, and 72.32%, respectively. Steel recycling (P2)
reduces waste heavy metals, particularly copper II, chromium, co-
balt and nickel II, and avoids selenium and vanadium emissions
into freshwater from the upstream coal and electricity production
processes.
POCP is primarily attributable to direct NOx and SO2 emissions
from heating during powder coating operations and to indirect NOx
and SO2 emissions from the upstream stage of copper production
due to electricity consumption. In this case, copper recycling (P5)
brings the largest environmental benefit to the POCP category,
43.42%.
3.3. Maximum reclamation distance
In China, railway transportation is still the prevalent logistics
approach, especially with the rapid development of high-speed
railways (Chen, 2007). The power for high-speed railways is elec-
tricity and so it is valuable to find the maximum reclamation dis-
tance of Scenario 2. As shown in Table 4, considering transportation
distance is one of important variables, recycling is not clearly ad-
vantageous from an environmental perspective for some impact
categories (GWP, ODP, ADP fossils and MAETP) in Scenario 2.
Recycling advantages depend on the convenient distance to collect
and transport waste refrigerators from residents to recycling plants
without jeopardizing international environmental conservation.
The final results whether it is worth to recycle refrigerator are
presented in Fig. 3, which shows that the specific values of the
environmental benefits turning point depend on the final disposal
and distance traveled. The x-axis refers to the transport distances of
a refrigerator (from 0 to 1000 km) and the y-axis shows the
dimensionless value of f(x)/f(0). Here x refers to distance of
57
2
Ep(MAETP)=769.23km
1
0
Impact categories
0 200 400 600 800 1000 1200
-1
Ep(GWP)=588.24km
-2 Ep(ODP)=163.93km
Ep(ADP)=909.09km
-3
-4
-5
-6
Distance (km)
GWP ODP ADP fossil MAETP
Fig. 3. Difference in the impact categories in scenario 2. (Positive values refer to
environmental benefits, and negative values refer to environmental pollution.)
recycling (kilometer), f(x) refers to the environmental benefits
under the recycling distance of x kilometer, f(0) refers to the envi-
ronmental benefits recycled in the local, and the environmental
benefits turning point refers to x when f(x) ¼ 0. The graphical
representation helps to immediately identify the different changes
in the gradient of each impact category. The environmental benefits
turning points of ADP and MAETP are situated at large distances
(Ep(ADP) ¼ 909.09 km, Ep(MAETP) ¼ 769.23 km). The ODP balance
becomes negative after a short distance (Ep(ODP) ¼ 163.93 km),
whereas the GWP begins with a large difference. However, due to a
very strong gradient, the breakeven point (Ep(GWP)) is situated at
588.24 km. For example, for the GWP, if the distance from the usage
location to the recycling location is less than 588.24 km, recycling a
refrigerator brings environmental benefit to control the green-
house effect and vice versa. The maximum reclamation distance of
Scenario 2 is 163.93 km below which the environmental benefits of
recycling are guaranteed to exceed its impacts. It is worth
mentioning that the LCIA results do not predict actual impacts on
category endpoints (Ruan and Xu, 2011), so the maximum recla-
mation distance in terms of environmental benefits is based on the
potential environmental effects.
4. Discussions
4.1. Sensitivity analysis
A sensitivity analysis was conducted using the operating pa-
rameters: the distance of transportation (Q1), the steel input in the
magnetic separation process (Q2), the HIPS input in the eddy cur-
rent separation process (Q3), the aluminum input in the eddy
current separation process (Q4), the copper input in the eddy cur-
rent separation process (Q5), the PUR input in the winnowing
process (Q6), and electricity consumption in the crushing stage
(Q7) and their contributions to the inventory. When compared with
the baseline operating conditions, each of these operating param-
eters increases by 10% while all other elements are kept constant.
The findings of the sensitivity analysis are shown in Table 5.
The sensitivity analysis results show that the HIPS input in the
eddy current separation process (Q3) has the most significant
benefit on the categories. With a variation of 10%, the values
calculated for HIPS in each impact category vary over 25% on the
baseline. The results also indicate that increase of electricity con-
sumption could increase the environmental pollution of refrigerator
reclamation and recycling. A variation of 10% in the amount of
electricity consumption in the crushing stage (Q7) results in a
change of 21.40% and 22.59% on the ADP fossils and POCP,
58 R. Xiao et al. / Journal of Cleaner Production 131 (2016) 52e59

Table 5
Sensitivity analysis of the important parameters.

Process Transporting Crushing Magnetic separation Winnowing Eddy current separation

Parameter Distance Electricity Steel PUR Aluminum Copper HIPS

Variation 10.00% 10.00% 10.00% 10.00% 10.00% 10.00% 10.00%

GWP 6.07% 15.61% 10.06% 10.26% 10.84% 14.65% 25.07%
AP 6.21% 15.65% 10.18% 10.20% 11.59% 7.93% 25.03%
EP 3.64% 15.48% 9.88% 6.67% 10.03% 9.75% 25.02%
ODP 1.42% 10.02% 10.00% 10.00% 10.00% 10.00% 25.00%
ADP elements 6.82% 10.00% 10.00% 10.00% 10.00% 10.00% 25.00%
ADP fossil 5.00% 21.40% 10.00% 10.24% 9.94% 9.96% 25.02%
FAETP 9.14% 15.73% 9.85% 10.11% 10.00% 9.95% 25.00%
HTP 7.60% 15.46% 10.00% 10.25% 9.84% 7.14% 25.11%
MAETP 9.49% 19.45% 9.94% 6.99% 10.04% 7.10% 24.90%
POCP 3.88% 22.59% 10.00% 10.00% 10.00% 10.00% 25.00%
TETP 8.17% 15.73% 10.00% 10.05% 10.00% 7.60% 25.00%

Positive values refer to environmental pollution, and negative values refer to environmental benefits.

respectively; the second significant parameter on ADP fossils and
POCP is the PUR input in the winnowing process (Q6). Except the
HIPS input in the eddy current separation process (Q3), the second
most sensitive factor for the GWP is electricity consumption in the
crushing stage (Q7) and the copper input in the eddy current sep-
aration process (Q5) which varied approximately 15% from the
baseline. Finally, the distance of transportation (Q1) has the weakest
influence on the impact categories, with changes of less than 10%.
4.2. Limitations
Due to limitations of time and funds, the LCA was performed
under the following conditions: 1) only environmental benefits of
separating materials of compressors at source (i.e., copper and
steel) were considered, while environmental benefits of saved en-
ergy and other materials in the production of compressors were
ignored due to the lack of inventory. 2) in order to find the share of
each impact category from the avoided potential damage to the
environment, the result of each impact category needs weighting
and normalization. But the CML 2001 method does not support
weighting and aggregation into one score (Guinee, 2002).
Some components of the refrigerator, such as the electronic
components, are outsourced, and data on their manufacturing was
unavailable, thus leading to an underestimation of the environ-
mental burdens. Meantime, a mass-based allocation for materials
may bring uncertainty on the inventory. For example, in eddy
current separation process environmental benefits are calculated
based on the mass ratio of steel and aluminum we choose. In fact,
due to the fact that there are many different kinds of refrigerators,
the mass ratio of the steel and aluminum changes in a certain range.
Therefore, we chose an average which may bring uncertainty to the
inventory. Furthermore, limited by data unavailability, we had to
make extrapolations based on calculations available in the litera-
ture, which may bring uncertainty to the results. In addition, the
average recycling rates of materials in China may be below the
recycling rates of materials acquired from representative enterprise
because of its large scale and high technology level. At last, the
collection of waste refrigerators is excluded in this study due to a
lack of data, so the environmental benefits from recycling may be
overestimated.
4.3. Policy implications
recycling system so as to take back and recycle the waste re-
frigerators in China. In consideration of the recycling technology of
refrigerators, the number of specialized appliance recycling com-
panies is relatively small in China which cannot meet the demand
for recycling a large number of obsolete household appliances (He
and Xu, 2014). Therefore, it is necessary for the government to
make preferential industrial policies to expand specialized appli-
ance recycling companies. For example, financial subsidies can be
provided to encourage the application of cleaner techniques and
equipment. Preferential tax policies can be used to motivate more
obsolete household appliances to be recycled with environment-
friendly treatment and disposal methods. Furthermore, the inte-
grated system needs to be encouraged which combines family-
workshop collection, manual separation and professional refining
to reduce the competition for recyclable materials and simulta-
neously to safeguard job opportunities for existent individual col-
lectors and dismantlers (Chi et al., 2014). As for the optimization of
transport distance, the status quo of Chinese obsolete refrigerators
should be considered to build a reasonable transport network for
recycling so as to balance the recycling costs and environmental
impacts.
5. Conclusions
In this study, recycling and disposal phase of a refrigerator was
studied, mainly based on first-hand survey data. Under the existing
recycling conditions, regardless of the mode of transportation,
recycling refrigerators cause the ozone layer depletion potential
because the crushing process inevitably releases the CFC-11. The
environmental benefits brought about by recycling mainly come
from the resource and energy savings in the upstream production,
such as in the HIPS recycling (P3) (22.17%), the steel recycling (P2)
(23.94%), and the copper recycling of (P5) (8.10%) phases, which
together account for over 50% of the total GWP benefits. Each
impact category has its own optimal reclamation distance. The
maximum distances (roughly no farther than 163.93 km for the
return trip) should be considered as a starting point when
designing new reverse logistics networks, taking into account the
reclamation and recycling of refrigerators. Given the importance of
reducing the transportation impact, the optimization of these
networks is crucial. Such optimization must include the location of
the facility and planned transportation routes.

The results of this study indicate that the refrigerator recycling Acknowledgments
technology and transportation distance are the key factors deciding
whether the obsolete refrigerators can be recycled effectively. This research was financially supported by the Sustainability
Therefore, it is important to build a reasonable waste refrigerator Consortium and the Social Science Foundation of China
 R. Xiao et al. / Journal of Cleaner Production 131 (2016) 52e59
(15ZDA053). We would like to thank the Sustainability Consortium
for improving the paper.
References
Altekin, F.T., Kandiller, L., Ozdemirel, N.E., 2008. Profit-oriented disassembly-line
balancing. Int. J. Prod. Res. 46 (10), 2675e2693.
Cavalett, O., et al., 2013. Comparative LCA of ethanol versus gasoline in Brazil using
different LCIA methods. Int. J. Life Cycle Assess. 18 (3), 647e658.
Cederberg, C., Stadig, M., 2003. System expansion and allocation in life cycle
assessment of milk and beef production. Int. J. Life Cycle Assess. 8 (6), 350e356.
Chen, Y., 2007. Research on System of Reverse Logistics Transportation for Waste
Household Appliances. Hefei University of Technology.
Chen, W., 2011. Analysis on reducing energy loss by start/stop control of three
systems refrigerator. Refrigeration Technol. 1, 35e37 (in Chinese).
Chi, X., Wang, M.Y.L., Reuter, M.A., 2014. E-waste collection channels and household
recycling behaviors in Taizhou of China. J. Clean. Prod. 80 (7), 87e95.
Gehin, A., Zwolinski, P., Brissaud, D., 2009. Integrated design of product life-
cyclesdthe fridge case study. CIRP J. Manuf. Sci. Technol. 1 (4), 214e220.
Guinee, J.B., 2002. Handbook on life cycle assessment operational guide to the iso
standards. Int. J. Life Cycle Assess. 7 (5), 311e313.
He, Y., Xu, Z., 2014. The status and development of treatment techniques of typical
waste electrical and electronic equipments in China: a review. Waste Manag.
Res. 32 (4), 254e269.
Internet Consumer Research Center, 2014. The Annual Report of Market Research of
China Double Door Refrigerator in 2013e2014. Available from: http://zdc.zol.
com.cn/430/4308525_all.html.
Johnson, R.W., 2004. The effect of blowing agent choice on energy use and global
warming impact of a refrigerator. Int. J. Refrigeration 27 (7), 794e799.
Kim, H.C., Keoleian, G.A., Horie, Y.A., 2006. Optimal household refrigerator
replacement policy for life cycle energy, greenhouse gas emissions, and cost.
Energy Policy 34 (15), 2310e2323.
Kotera, Y., Hirasawa, E., Sakitani, H., 1999. Integrated recycle plant for electric
home appliances-automated primary disassembling for refrigerators. In:
59
Environmentally Conscious Design and Inverse Manufacturing, 1999. Pro-
ceedings. EcoDesign'99: First International Symposium on. IEEE.
Lambert, A., Stoop, M., 2001. Processing of discarded household refrigerators: les-
sons from the Dutch example. J. Clean. Prod. 9 (3), 243e252.
Laner, D., Rechberger, H., 2007. Treatment of cooling appliances: interrelations
between environmental protection, resource conservation, and recovery rates.
Resour. Conserv. Recycl. 52 (1), 136e155.
Li, H., 2007. Research on the System of Recovery and Disposal and Economy of
Recovery about Refrigerator. Hefei University of Technology.
Liu, Y., 2011. The Study on Green Product of Refrigerator Based on Life Cycle
Assessment Method. Taiyuan University of Science & Technology.
Liu, Y., Chi, X., 2013. Harm and treatment technologies of Chinese e-waste. Environ.
Sci. Manag. 38 (5), 57e60 (in Chinese).
McCulloch, A., Lindley, A., 2003. From mine to refrigeration: a life cycle inventory
analysis of the production of HFC-134a. Int. J. Refrigeration 26 (8), 865e872.
National Bureau of Statistics of China, 2014. China Statistical Yearbook 2013. China
Statistics Press, Beijing.
Nicol, S., Thompson, S., 2007. Policy options to reduce consumer waste to zero:
comparing product stewardship and extended producer responsibility for
refrigerator waste. Waste Manag. Res. 25 (3), 227e233.
PE International, 2012. GaBi 6 Software and Databases. Leinfelden-Echterdingen,
Germany.
Ruan, J., Xu, Z., 2011. Environmental friendly automated line for recovering the
cabinet of waste refrigerator. Waste Manag. 31 (11), 2319e2326.
Weidema, B., Hischier, R., 2006. Ecoinvent Data v2. 2. St. Gallen, Switzerland.
Yang, Y., 2006. Study on Waste Refrigerators Dismantling and Release and
Removing of CFC-11 in Rigid Polyurethane Foams. Southwest Jiao Tong
University.
Yunho, H., Jin, D.-H., Radermacher, R., 2007. Comparison of R-290 and two HFC
blends for walk-in refrigeration systems. Int. J. Refrigeration 30 (4), 633e641.
Zhang, J., 2010. Research on Refrigerator Life Cycle Environmental Impact. Shanghai
Jiao Tong University.
Zhou, H., Gao, Y., 2007. Comprehensive interpretation of Chinese RoHS e “electronic
information products pollution control regulations”. China Insp. Quar. 5, 14e15.