Journal of Cleaner Production 162 (2017) 1368e1375

Contents lists available at ScienceDirect

Journal of Cleaner Production

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

Life cycle assessment of clothing libraries: can collaborative

consumption reduce the environmental impact of fast fashion?
Bahareh Zamani a, Gustav Sandin b, *, Greg M. Peters a, c
a
Chalmers University of Technology, 412 96 Gothenburg, Sweden
b
RISE Research Institutes of Sweden, Eklandagatan 86, 412 61 Gothenburg, Sweden
c
School of Civil and Environmental Engineering, UNSW, Sydney, NSW 2052, Australia

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

Article history: Fast fashion is a clothing supply chain model that is intended to respond quickly to the latest fashion
Received 26 January 2016 trends by frequently updating the clothing products available in stores. The shift towards fast fashion
Received in revised form leads to shorter practical service lives for garments. Collaborative consumption is an alternative way of
8 May 2017
doing business to the conventional model of ownership-based consumption, and one that can potentially
Accepted 14 June 2017
Available online 18 June 2017
reduce the environmental impacts of fashion by prolonging the practical service life of clothes. In this
study, we used life cycle assessment to explore the environmental performance of clothing libraries, as
one of the possible ways in which collaborative consumption can be implemented, and compared the
Keywords:
Problem shifting
advantages and disadvantages in relation to conventional business models. Furthermore, the key factors
Apparel influencing the environmental impact of clothing libraries were investigated. We based our assessment
Climate change on three key popular garments that are stocked in clothing libraries: jeans, T-shirts and dresses. The
Ecotoxicity results showed the benefits of implementing clothing libraries associated with the garments’ prolonged
Eutrophication service lives. Therefore to achieve environmental gains, it is important to substantially increase garment
Water use service life. Moreover, the results quantitatively demonstrated the potential risk of problem shifting:
increased customer transportation can completely offset the benefits gained from reduced production.
This highlighted the need to account for the logistics when implementing collaborative consumption
business models.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction Collaborative consumption is an alternative way of doing busi-

“Fast fashion” is a clothing supply chain model that is intended
to respond quickly to the latest fashion trends by frequently
updating the clothing products available in stores (Byun and
Sternquist, 2011; Moore and Fernie, 2004). Major retailers, for
example H&M and Zara, have adopted fast fashion by introducing
new clothing designs to stores every three to five weeks (Hu et al.,
2014). With rapidly shifting fashion trends, the practical service life
of clothes (how long they are actually worn) is short, well below the
technical service life (how long they could be worn, even if they are
no longer in fashion). Furthermore, clothes may not be taken care of
properly and may be disposed of without much thought. This
makes fast fashion an important driver of environmental impacts
(Roos et al., 2015).
* Corresponding author.
E-mail address: gustav.sandin@ri.se (G. Sandin).
ness that can potentially reduce the environmental impacts of
fashion, by prolonging the practical service life of clothes. Collab-
orative consumption has been defined as “people coordinating the
acquisition and distribution of a resource for a fee or other
compensation” (Belk, 2014), which can include the renting, trading,
swapping and borrowing of goods (Piscicelli et al., 2014). Collabo-
rative consumption is part of the so called “sharing economy”,
which has gained increased attention in recent years (Belk, 2014).
Apart from collaborative consumption, a plethora of similar and
related terms have emerged, for example “commercial sharing
systems”, “product-service systems”, and “access-based consump-
tion” (Belk, 2014).
There is considerable hope that collaborative consumption and
other sharing economy business models will transform society and
reduce environmental impacts (Mont, 2002; Botsman and Rogers,
2011; Gansky, 2010; Armstrong and Lang, 2013; Heinrichs, 2013;
Agrawal et al., 2011; Pedersen and Netter, 2015). Several studies
have evaluated the environmental performance of various business

http://dx.doi.org/10.1016/j.jclepro.2017.06.128
0959-6526/© 2017 Elsevier Ltd. All rights reserved.
 B. Zamani et al. / Journal of Cleaner Production 162 (2017) 1368e1375
models of the sharing economy (Goedkoop et al., 1999; Tukker,
2004; Agrawal et al., 2011), including the reuse of clothes
(Woolridge et al., 2006; Farrant et al., 2010). These studies show
both opportunities and potential pitfalls of collaborative con-
sumption. A potential pitfall is the risk of rebound effects that offset
the environmental gains. For example, Leismann et al. (2013)
hypothesised that the impact of more frequent transactions of
goods (e.g. related to transportation and/or packaging) may be
higher than the impact reduction in production.
In the fashion industry, an example of a collaborative con-
sumption business model is the clothing library, in which a
monthly membership fee allows members to borrow a specific
number of clothing pieces in a set time, typically a few weeks.
Successful small scale enterprises in Sweden demonstrate the
business potential of clothing libraries, including Lånegarderoben
(www.lanegarderoben.se) and Kl€ adoteket (www.kladoteket.se).
From the perspective of the user, clothing libraries can maintain or
even increase the speed of fashion, as each user may update his/her
wardrobe at least as frequently as with conventional fashion con-
sumption. On the other hand, from the perspective of the clothing
item, clothing libraries can slow down the speed of fashion,
assuming that each item is used more times before disposal
compared to conventional fashion consumption. In this way,
clothing libraries can prolong the practical service life of clothing
items, and reduce the production of new clothing and the associ-
ated environmental impacts. However, while Leismann et al. (2013)
made a qualitative hypothesis regarding the general idea of clothes
swapping, no previous study has quantitatively assessed whether
the environmental gains in production induced by clothing libraries
could be offset by environmental losses elsewhere in the product
life cycle. This topic is explored in this article, for the geographical
context of Sweden, by means of the following research questions:
1. What are the environmental advantages and disadvantages of
clothing libraries in relation to conventional business models?
2. What are the key controlling factors influencing the environ-
mental impact of clothing libraries?
2. Method
In order to assess the environmental impact of clothing libraries,
life cycle assessment (LCA) was conducted according to the ISO
14044 standard (ISO, 2006). Three different garments were ana-
lysed: a 100% cotton T-shirt, a pair of 98% cotton/2% elastane jeans
and a 100% polyester dress. Statistics on Swedish fashion con-
sumption were used to select these garments so that they together
would provide a representative picture of the Swedish fashion
consumption in terms of material production, fabric construction,
finishing, customer behaviour and end-of-life handling. For details
on selection of garments, see Roos et al. (2015).
GaBi software was used for modelling the product systems
(www.gabi-software.com). Background processes were modelled
with data from databases, mainly the Ecoinvent database (Hischier,
2003).
2.1. Goal definition, functional unit and system boundaries
The goal of the study was to assess the environmental impacts
associated with different clothing library setups. Therefore, 12
scenarios representing different types of clothing library business
models were created. These scenarios were compared with the
baseline scenarios reflecting a regular fashion business model, i.e.
that the customer purchases and owns the clothing. The functional
unit was defined as “one average use” of each of the three
1369
garments. “One use” means using the garment at some point within
a 24 h time period. The choice of one use as the functional unit was
made to facilitate evaluation of the influence of interventions to
prolong the service life of the garment (i.e. more users during the
garment’s life cycle).
This is a cradle to grave LCA study, spanning from raw material
extraction to end of life treatment. The life cycle of each garment
includes fibre, textile and garment production, distribution,
wholesaling and retailing, use (including the customer’s trans-
portation and laundry) and end of life handling, as shown in Fig. 1.
2.2. Selection of impact categories and characterisation methods
The focus of the study lay on investigating environmental im-
pacts for climate change, fresh water consumption, fresh water
ecotoxicity and fresh water eutrophication since these environ-
mental impacts appeared to be among the indicators with highest
relevance for the clothing sector (Roos et al., 2015). Results for land
use, human toxicity and acidification were excluded from this pa-
per since result for the land use was correlated with freshwater
consumption, results for human toxicity were correlated to fresh
water ecotoxicity, and results for acidification were correlated with
freshwater eutrophication. Results for the excluded impact cate-
gories are provided in Roos et al. (2015). The characterisation
models used in this study were global warming potential with a
100 years perspective (GWP100) excluding biogenic CO2 emissions
(IPCC, 2013), consumptive freshwater use (Swiss Ecoscarcity
model) (Frischknecht and Büsser Kno €pfel, 2013), ecotoxicity po-
tential (USEtox model) (Rosenbaum et al., 2008) and freshwater
eutrophication potential (EUTREND model) (Struijs et al., 2009) as
implemented in GaBi v 6.0 (Thinkstep, 2015).
2.3. Life cycle inventory
2.3.1. Scenarios
In this study, we aimed to map the life cycle of garments in a
way which is statistically representative of Swedish fashion con-
sumption, both in terms of production and use. For the fibre, fabric
and garment production phase it was assumed that each garment
was produced in the three countries that dominate clothing im-
ports to Sweden e China (32% of clothing imports to Sweden),
Bangladesh (11%) and Turkey (6%) (Statistics Sweden, 2015) e and
that retailing, use and end of life handling occur in Sweden.
For construction of the clothing library scenarios different pa-
rameters were varied including: (i) the extension of the garment’s
service life (two or four times the garment’s service life assumed in
the baseline scenarios; to make this feasible one may have to
extend the technical service life of current garments, or only use
particularly durable garments), (ii) the type of customer trans-
portation (the same means of transportation as in the baseline
scenarios for medium impact, or low or high impact means of
transportation, respectively), and (iii) whether the setup is an off-
line (physical store) or online (internet, with a pick-up point for
deliverables) solution (this influences mode of transportation to
and from the store/pick-up point and distances). The scenarios are
shown in Table 1.
The use phase includes the customer’s transport to and from the
store and the residential laundry. The influence of the type of
customer transportation was explored in the clothing library sce-
narios by assuming scenarios with medium, low and high trans-
portation impacts. Medium impact transportation is 50% car/50%
bus for both online and offline scenarios (this is the same ratio
between car and bus as in the baseline scenario). High impact
transportation is 100% car for both online and offline scenarios. Low
impact transportation is using bicycle or walking to the pick-up
1370 B. Zamani et al. / Journal of Cleaner Production 162 (2017) 1368e1375

Baseline scenarios

Fibre Textile Garment Wholesale Use phase, Use phase,

Distribution End of life
production production production and retail transport laundry

Collaborative consumption scenarios

Fibre Textile Garment Wholesale Use phase, Use phase,

Distribution
production production production and retail transport laundry

Use phase,
transport

End of life

Fig. 1. Life cycle phases included in the baseline scenarios and the clothing library scenarios.

Table 1
Collaborative consumption scenarios.

Scenario Service life compared to baseline Type of customer transportation Type of setup

1 2 Medium impact customer transport: online

50% car/50% bus to/from store or pickup-point
2 4 Medium impact customer transport: online
50% car/50% bus to/from store or pickup-point
3 2 Medium impact customer transport: offline
50% car/50% bus to/from store or pickup-point
4 4 Medium impact customer transport: offline
50% car/50% bus to/from store or pickup-point
5 2 Low impact customer transport: online
Bike/walk to/from pickup-point
6 4 Low impact customer transport: online
Bike/walk to/from pickup-point
7 2 Low impact customer transport: offline
100% bus to/from store (e.g. city centre)
8 4 Low impact customer transport: offline
100% bus to/from store (e.g. city centre)
9 2 High impact customer transportation: online
100% car to/from store
10 4 High impact customer transportation: online
100% car to/from store
11 2 High impact customer transportation: offline
car to/from pickup-point
12 4 High impact customer transportation: offline
car to/from pickup-point

point for the online scenarios, and using 100% bus for the offline
scenarios (these assumptions are further discussed below). The
datasets used for transport modelling for the car and bus were the
Ecoinvent v2.0 datasets “RER: transport, passenger car” and ”CH:
transport, regular bus”, respectively (Spielmann et al., 2007). For
the baseline scenarios and the offline scenarios we assumed a
customer transport of 17 person-km/kg of purchased garment
based on Granello et al. (2015). For online scenarios, transportation
distance from the customer’s home to the pick-up point (where the
online orders are delivered) was assumed to be one third of the
distance in the offline scenarios. This change in distance between
the physical stores in the offline scenarios and the pick-up points in
the online scenarios is an estimation based on the fact that pick-up
points in Sweden typically are located in small stores closer to
residential areas than city centres or shopping malls (the normal
locations of physical clothing stores). Furthermore, the possibility
of home delivery was not explored as clothing most often renders a
size of packages which, in Sweden, seldom is home delivered.
Furthermore, as mentioned above, the differences in distances
between online and offline scenarios influenced our assumptions of
modes of transport: as pick-up points in the online scenarios can be
assumed to be within walking distances to the home of many
customers, we assumed that the low impact transport mode for the
online scenarios to be bike or walking. In contrast, as the distance to
physical stores are longer, we assumed the low impact mode of
transportation in the offline scenario to be public bus. In building
the scenarios, our aim was to reflect on the main uncertainties of
relevance for clothing libraries in Sweden. In studies of clothing
libraries in other geographical contexts, other assumptions and
scenarios might be relevant to set up, e.g. because of differences in
city planning, urban density, postal services, and other factors
potentially influencing the logistics of collaborative consumption.
In the clothing library scenarios, the store could take re-
sponsibility for some of the laundry and thus do this more
 B. Zamani et al. / Journal of Cleaner Production 162 (2017) 1368e1375
efficiently, but there could be more frequent laundry compared to
the baseline scenarios (e.g. the customer washes the garment
before handing it back to the store, the store washes the garment,
and then the customer washes it before use). Therefore we
assumed the garment is washed, dried and ironed in the same
manner as in the baseline scenarios when the customer does the
laundry at home. For the laundry phase, Swedish conditions were
assumed, therefore the Ecoinvent data set “Electricity, low voltage,
at grid/SE S” (Frischknecht et al., 2007) was used to model the
electricity use.
Table 2 summarises the specification of each garment and the
use phase parameters for the baseline scenarios and the clothing
library scenarios. The average number of uses per garment in the
baseline scenarios e which determines the actual service life of the
garment e is estimated based on the number of garments an
average Swede purchases per year according to Swedish statistics
on net annual imports of garments in 2008 (Statistics Sweden,
2015) and the SMED study of the textiles flow in Sweden
(Carlsson et al., 2011). The number of days each garment is used per
year and number of washes per garments life cycle are based on a
study of customer behaviour carried out by Gwozdz et al. (2013)
and on a survey among 225 Swedish fashion customers carried
out by Granello et al. (2015); also, some complementary assump-
tions were necessary. For further details on the data inventory, see
Roos et al. (2015). In the clothing library scenarios, more customers
are assigned to each garment compared with the baseline sce-
narios. Since this number is an estimate, the effect of the choice of
number of customers on the environmental burden was tested by a
sensitivity analysis, in which we halved and doubled the number of
customers per T-shirt service life in scenario 3 (22 customers), i.e.
we varied the number of customers who rent the garment from the
clothing library (assuming the number of uses per service life re-
mains the same).
3. Results and discussion
To acquire an overall perspective regarding the environmental
impact associated with each collaborative consumption scenario,
the results for climate change, freshwater eutrophication, fresh-
water consumption and freshwater ecotoxicity are shown for the
medium transport impact scenarios. The results are normalised to
the baseline scenario with medium impact transportation results,
which are shown with black bars in Fig. 2. The results for low and
high transport impact are shown in error bars for each scenario. For
details on the absolute value of the results, see Roos et al. (2015).
The results for global warming show the online scenarios lead to
Table 2
Garments specifications and user behaviour for the baseline and clothing library scenarios.
Garment T-shirt
Mass 110 g
Textile material 100% cotton
Baseline scenarios
Medium impact transportation 50% car
50% bus
Low impact transportation 100% bus
High impact transportation 100% car
Customer transport distance 17 km distance back and forth
Number of uses 22
Number of uses per laundry cycle 2 10
Number of customers per service life 1 1
Collaborative consumption scenarios e twice (“life x2”) and four times (“life x4”) the garment service life
Number of uses before disposal (life 2) 44
Number of uses before disposal (life 4) 88
Number of customers before disposal (life 2) 11
Number of customers before disposal (life 4) 22
1371
more savings comparing with the offline scenarios. The offline
scenario which doubles the garment’s service life and which has
the highest transportation impact potentially leads to 26% higher
global warming potential compared to the baseline scenario with
medium impact transportation. It is necessary to lengthen the
service life by a factor of four for the improvement in global
warming potential to be independent of the transportation used.
In terms of freshwater eutrophication, the online scenario with
four times longer service life and low transportation impact
potentially leads to 60% lower impact. Among the medium trans-
portation impact scenarios, the online scenarios lead to lower
impact compared with the offline scenarios.
The results for freshwater consumption show savings up to 75%
in the clothing library scenarios in comparison with the baseline
scenario with medium impact transportation. For freshwater con-
sumption the type of transportation does not play a major role
since most of the water consumption is related to the production
phase. Since there were almost no variations in the results due to
the type of transportations in the scenarios, no error bars are
shown for this indicator.
The results for freshwater ecotoxicity show that assumptions
regarding the type of customer transportation have considerable
influence on the environmental burden. Online services with walk/
bike transport leads to 35% savings in ecotoxicity potential
comparing to the baseline scenario with medium impact
transportation.
Figs. 3 and 4 show the normalized results of the assessment of
the environmental burden of a pair of jeans and a dress respec-
tively. The results for global warming and freshwater consumption
for both a pair of jeans and a dress are similar to those of the t-shirt.
Regarding freshwater eutrophication, all clothing library sce-
narios show more saving than the baseline scenario for a pair of
jeans. The results for the dress show the robust difference of results
with respect to a change of type of transportation with the
exception of the offline scenario with a doubling of the service life.
Considering freshwater consumption, all clothing library sce-
narios are more favourable than the baseline scenarios and leads to
up to 75% savings. For the dress, the clothing libraries scenarios are
more favourable in terms of freshwater consumptions compared to
the baseline scenario with medium impact transportation.
In terms of freshwater ecotoxicity, all medium and low impact
customer transport scenarios are more favourable to the baseline
scenarios for the pair of jeans. On the other hand, the results of
ecotoxicity for high impact customer transport scenarios are up to
38% higher than the baseline scenario with medium impact
transportation.
Jeans Dress
477 g 478 g
98% cotton 100% polyester
50% car 50% car
50% bus 50% bus
100% bus 100% bus
100% car 100% car
17 km distance back and forth 17 km distance back and forth
200 10
3
1
400 20
800 40
10 15
20 30
1372 B. Zamani et al. / Journal of Cleaner Production 162 (2017) 1368e1375

2.5

Baseline
2

Scenarios 3,7,11
1.5 (life x2, offline)

Scenarios 1,5,9
1 (life x2, online)

Scenarios 4,8,12
0.5 (life x4, offline)

Scenarios 2,6,10
0 (life x4, online)
Global Freshwater Freshwater Freshwater
warming ecotoxicity

Fig. 2. Normalised results of environmental impact potential indicators of one use of a T-shirt for the medium transport impact baseline and clothing library scenarios, the results
related to low and high impact transport for each scenario are shown as error bars. (In the legend: life 2 and life 4 mean: two or four times longer garment service life compared
to the baseline scenarios). This implies 44 or 88 uses before disposal, and 11 or 22 customers before disposal, respectively.

1.4

Baseline
1.2

1 Scenarios 3,7,11
(life x2, offline)
0.8
Scenarios 1,5,9
0.6 (life x2, online)

0.4 Scenarios 4,8,12

(life x4, offline)
0.2
Scenarios 2,6,10
(life x4, online)
0
Global Freshwater Freshwater Freshwater
warming ecotoxicity

Fig. 3. Normalised results of environmental impact potential indicators of one use of a pair of jeans for the medium transport impact baseline and clothing library scenarios, the
results related to low and high impact transport for each scenario are shown as error bars. (In the legend: life 2 and life 4 mean: two or four times longer garment service life
compared to the baseline scenarios. This implies 400 or 800 uses before disposal, and 10 or 20 customers before disposal, respectively.)

For a dress, the results in relation to freshwater ecotoxicity for
medium impact customer transport for both online and offline
scenarios are up to 3.4 times higher than the baseline scenarios.
Since major ecotoxicity impacts are related to the customer trans-
portation activities, only the low impact customer transportation
online collaborative consumption scenarios showed less environ-
mental burden comparing to baseline scenarios.
The overall results for online scenarios show more environ-
mental benefits compared to offline scenarios due to the closer
package pickup-point to customers (one third of the distance). This
underlines the importance that the locations of stores and/or
pickup points are close to customers or accessible by public
transportation.
One surprising result from the assessment of the collaborative
consumption business models was the increase of freshwater
ecotoxicity potential especially in scenarios using high impact
transportation systems. It is possible that this represents an
imbalance between the level of detail in LCI data for textile pro-
duction compared with transportation e toxic emissions from the
production phase are seldom inventoried satisfactorily in the
database data (Roos, 2016). USEtox was the characterisation
method used in this study and while this consensus method covers
most chemicals, it lacks characterisation factors for many textile
chemicals (Roos and Peters, 2015). On the other hand, the toxic
emissions from energy production and transportation systems are
relatively well inventoried and characterisation factors are typically
available for the main toxicants, so that these processes tend to
dominate toxicity calculations.
The previous discussion has enabled us to identify the scenarios
exhibiting the greatest absolute deviation from the baseline sce-
narios. In order to illustrate the influence of key variables on
different elements of the life cycle, Fig. 5 shows each life cycle phase
 B. Zamani et al. / Journal of Cleaner Production 162 (2017) 1368e1375 1373

7
Baseline
6

5 Scenarios 3,7,11
(life x2, offline)
4 Scenarios 1,5,9
(life x2, online)
3
Scenarios 4,8,12
2 (life x4, offline)

Scenarios 2,6,10
1
(life x4, online)

0
Global Freshwater Freshwater Freshwater
warming ecotoxicity
Fig. 4. Normalised results of environmental impact potential indicators of one use of dress for the medium transport impact baseline and clothing library scenarios, the results
related to low and high impact transport for each scenario are shown as error bars. (In the legend: life 2 and life 4 mean: two or four times longer garment service life compared
to the baseline scenarios. This implies 20 or 40 uses before disposal, and 15 or 30 customers before disposal, respectively.)

0.14

0.12

0.1 Wholesale and

kg CO2 eq/use

retail
0.08

0.06 Use phase,

transport
0.04
Use phase,
0.02 laundry

0 End of life
-0.02
Baseline Scenario 11 Scenario 6
Fig. 5. Global warming potential of the baseline with medium impact transport vs. the clothing library scenarios: scenario 11 (high impact transport, life 2, offline) as worst case
scenario and scenario 6 (low impact transport, life 4, online) as best case scenario for one (1) use of the T-shirt.

contribution to the global warming potential for the baseline sce-
nario with medium impact transport and two collaborative con-
sumption scenarios: the “high transport impact, life 2, offline”
scenario is the worst case scenario and the “low transport impact,
life 4, online” scenario is the best case scenario for the T-shirt.
Fig. 5 shows which processes that contribute the most to the
carbon footprint for the different life cycle phases. The results in
Fig. 5 quantitatively indicate the significant contribution of the use
phase transport in the collaborative consumption scenarios. One
can observe the burden of production in Asia and the Middle East
being exchanged for transport impacts in Sweden e an example of
geographical problem shifting.
In the baseline scenario, statistics on Swedish apparel imports
and a survey of Swedish user behaviour were used to determine the
actual service lives of average garments purchased in Sweden, as
well as laundry practices and the customer’s transportation to and
from the retailer. The results show that the customer transportation
has a bigger impact on global warming potential comparing with
laundry, which contrasts with many previous studies. This is both
because previous studies often exclude the customer trans-
portation, and because they often overestimate the importance of
the laundry, because the number of uses per garment life cycle is
based on technical service life rather than actual user behaviour
(Roos et al., 2016). The comparable low laundry contribution of our
study is also due to the geographical context. In Sweden, the energy
used for laundry has a relatively low carbon intensity (nuclear
power, hydroelectricity and wind provided 81% of local power in
2014 (Statistics Sweden, 2015)), and tumble drying is not used as
frequently as elsewhere (tumble drying usually is the key electricity
consuming process in LCA studies (on garment level) where
laundry dominates the global warming potential results).
The results for the two other garments follow the same trend, as
shown in Roos et al. (2015). The results highlight the need to ac-
count for the logistics when implementing a collaborative con-
sumption business model, for example by locating physical rental
services or clothing libraries in locations close to customers and/or
public transportation.
Fig. 6 shows the results of the sensitivity analysis on global
warming potential in which we varied the number of customers per
T-shirt service life in scenario 3 (life 2, medium impact trans-
portation, and offline). In other words, the number of customers
who rent the garment from clothing library is changed, but the
1374 B. Zamani et al. / Journal of Cleaner Production 162 (2017) 1368e1375
0.2
0.18
0.16
0.14
kg CO2 eq/use
0.12
0.1
0.08
0.06
0.04
0.02
0
-0.02
Baseline Scenario 3, 22
customers
Fig. 6. Sensitivity analysis on the number of customers per T-shirt service life in one of the clothing library scenarios, for global warming potential. The number of customers is
assumed to be 22 in the regular scenario. This value is halved and doubled for the sensitivity analysis.
number of uses per service life remains the same.
Fig. 6 shows that the number of customers per garment life cycle
can considerably influence the environmental impact of clothing
libraries, as more customers mean more transport to and from the
store (offline scenario) or pickup point (online scenario). So the
frequency of garment transactions, and the number of uses per
customer, matter greatly. These factors can be influenced by the
setup of the clothing library membership system, for example by
the number of clothing pieces the customer can borrow within a set
time period, and the length of that time period. Fewer items and a
longer time period could give an incentive to reduce the number of
customers per garment life cycle, which would be environmentally
preferable. Also, the payment system could matter. For example, if
the customer has to pay for each clothing transaction, that could
reduce the number of transactions and the associated environ-
mental impacts. Overall, the setup of the membership system, and
its influence on the frequency of transactions, is particularly
important for clothing libraries in locations that induce user
transportation with high environmental impact (e.g. remote areas
with poor access to public transportation). Similarly, the setup of
the membership system is less important in locations that induce
user transportation with low environmental impact (e.g. down-
town areas with good access to public transportation).
Previous studies show that reusing garments can reduce the
environmental burden of clothing since the environmental burden
associated with reuse of garments is insignificant in comparison
with the saving due to replacement of virgin materials (Woolridge
et al., 2006; Farrant et al., 2010). Moreover Woolridge et al. (2006)
show the energy consumption of the retailing and distribution
activities associated with second hand clothes are much lower than
the energy consumption associated with the production of virgin
materials. Previous studies with a specific focus on collaborative
consumption discussed the idea that leasing garments can poten-
tially contribute to more efficient resource use (Agrawal et al., 2011;
Leismann et al., 2013). In our study, the results of different envi-
ronmental impact categories show there are potential benefits (per
garment use) of implementing clothing libraries if the garments’
service life is prolonged, but it is important to achieve a substantial
increase of service life for environmental gains to occur. Moreover,
Leismann et al. (2013) hypothesized the existence of a negative
environmental rebound effect in collaborative consumption due to
increased transportation. This hypothesis is confirmed by our
quantified results showing that increased customer transportation
in some scenarios (i.e. offline, high impact customer
Produc on
Wholesale and
retail
Use phase,
transport
Use phase,
laundry
End of life
Scenario 3, 44 Scenario 3, 11
customers customers
transportation) can offset the environmental gains from reduced
production. This is a classic example of problem shifting between
life cycle phases, which underlines the importance of adopting a life
cycle perspective when studying the environmental implications of
various business models.
Since the geographical focus of our study was Sweden, further
investigation is needed for evaluating the outcome of environ-
mental impacts associated with setting up clothing libraries in
other countries. Since the supplier countries we used in our
assessment dominate global clothing supply, the key point of dif-
ferentiation when this type of assessment is applied to other cities
or countries may well be parameters such as transportation and
laundry impacts, as these can be expected to be different in
different countries. The range of per kilometre transportation im-
pacts in our study may be similar in other urbanised western
communities, but transport distances can be very location-specific.
Laundry habits may differ elsewhere and the impacts will be higher
in countries with carbon-intense electricity supplies, so that the
laundry may play a larger role in the overall climate impact
assessment.
In this study, we focused on the controlling factors which can
directly affect the environmental impacts of different clothing li-
brary scenarios, such as the extent to which the service life is
actually prolonged, the number of customers per garment service
life, and the transportation modes of the customer’s journeys. Then
there are other, indirect factors influencing these direct environ-
mental drivers, related to the customers’ attitude and adaptation
towards these new business models, as studied e.g. by Pedersen
and Netter (2015). Evidence from their study show there are
some limiting factors in relation to expansion of clothing libraries
including limited human and financial resources and conventional
owner-ship consumption patterns. Such factors were outside the
scope of the present paper.
4. Conclusions
One of the current concerns in relation to fast fashion is that
customers discard garments before the end of their technical life-
span. In this context, interventions that increase the practical ser-
vice life of garments are expected to be environmentally beneficial.
In this study two research questions were addressed:
1. What are the environmental advantages and disadvantages of
clothing libraries in relation to conventional business models?
 B. Zamani et al. / Journal of Cleaner Production 162 (2017) 1368e1375
2. What are the key controlling factors influencing the environ-
mental impact of clothing libraries?
The research questions were addressed through studying 12
clothing library scenarios and comparing them with scenarios
reflecting regular fashion business models. The results of different
environmental impact categories showed potential environmental
benefits (per garment use) of implementing clothing libraries if the
garments’ service life is substantially prolonged.
In order to illustrate the influence of key variables on different
elements of the life cycle, the contribution of each life cycle phase
to the global warming potential was measured. The results quan-
titatively demonstrated the potential risk of problem shifting:
increased customer transportation can offset the benefits gained
from reduced production. This highlights the need for accounting
for the logistics when implementing collaborative consumption
business models.
Acknowledgements
The research was funded by Mistra, the Swedish Foundation for
Strategic Environmental Research, through the Mistra Future
Fashion research programme.
References
Agrawal, V.V., Ferguson, M., Toktay, B., Thomas, V.M., 2011. Is leasing greener than
selling? Manag. Sci. 58 (3), 523e533.
Armstrong, C.M., Lang, C., 2013. Sustainable product service systems: the new
frontier in apparel retailing? Res. J. Text. Appar. 17 (1), 1e2.
Belk, R., 2014. You are what you can access: sharing and collaborative consumption
online. J. Bus. Res. 67 (8), 1595e1600.
Botsman, R., Rogers, R., 2011. What’s Mine is Yours: How Collaborative Consump-
tion is Changing the Way We Live. Collins, London.
Byun, S.-E., Sternquist, B., 2011. Fast fashion and in-store hoarding: the drivers,
moderator, and consequences. Cloth. Text. Res. J. 29 (3), 187e201.
Carlsson, A., Hemstro €m, K., Edborg, P., Stenmarck, Å., So
€rme, L., 2011. Kartla
€ggning
av m€ angder och flo €den av textilavfall. Sveriges Meteorologiska och Hydro-
logiska Institut (SMHI), Norrko €ping, Sweden.
Farrant, L., Olsen, S.I., Wangel, A., 2010. Environmental benefits from reusing
clothes. Int. J. Life Cycle Assess. 15 (7), 726e736.
Frischknecht, R., Tuchschmid, M., Faist-Emmenegger, M., Bauer, C., Dones, R., 2007.
Strommix. In: Dones, R. (Ed.), Sachbilanzen von Energiesystemen: Grundlagen
für den o €kologischen Vergleich von Energiesystemen und den Einbezug von
€
Energiesystemen in Okobilanzen für die Schweiz. Swiss Centre for Life Cycle
Inventories, Dübendorf, CH.
Frischknecht, R., Büsser Kno €pfel, S., 2013. Swiss Eco-factors 2013 According to the
Ecological Scarcity Method. In: Methodological Fundamentals and Their
Application in Switzerland. Environmental studies no. 1330, Bern.
Gansky, L., 2010. The Mesh: Why the Future of Business is Sharing. Penguin.
Goedkoop, M., van Halen, C., te Riele, H., Rommens, P., 1999. Product Service Sys-
tems, Ecological and Economic Basics. Pre consultants, The Netherlands.
€nbrink, A.K., Roos, S., Johansson, T., Granberg, H., 2015. Consumer
Granello, S., Jo
Behaviour on Washing, a Mistra Future Fasion Project 7 Report. Copenhagen.
Gwozdz, W., Netter, S., Bjartmarz, T., Reisch, L.A., 2013. Report on Survey Results on
Fashion Consumption and Sustainability Among Young Swedes. Copenhagen.
1375
Heinrichs, H., 2013. Sharing economy: a potential new pathway to sustainability.
GAIA Ecol. Perspecties Sci. Soc. 22 (4), 228e231.
Hischier, R., 2003. Life Cycle Inventories of Packaging and Graphical Paper. Final
report ecoinvent 2000 No. 11. Dübendorf.
Hu, Z.-H., Li, Q., Chen, X.-J., Wang, Y.-F., 2014. Sustainable rent-based closed-loop
supply chain for fashion products. Sustainability 6 (10), 7063e7088.
IPCC, 2013. In: Stocker, T.F., Qin, D., Plattner, G.-K., Tignor, M., Allen, S.K.,
Boschung, J., et al. (Eds.), Climate Change 2013: the Physical Science Basis.
Working Group I Contribution to the Fifth Assessment Report of the Intergov-
ernmental Panel on Climate Change. Cambridge University Press, Cambridge,
United Kingdom and New York, NY, USA. Available at: http://www.ipcc.ch/repo
rt/ar5/wg1/ (Accessed January 2016).
ISO, 2006. Environmental Management e Life Cycle Assessment e Principals and
Framework. International Standard ISO 14040. International Organization for
Standardization e ISO, Geneva, Switzerland.
Leismann, K., Schmidt, M., Rohn, H., Baedeker, C., 2013. Collaborative consumption:
towards a tesource-saving consumption culture. Resources 2 (3), 184e203.
Mont, O.K., 2002. Clarifying the concept of product-service system. J. Clean. Prod. 10
(3), 237e245.
Moore, C.M., Fernie, J., 2004. International Retail Marketing. In: International Retail
Marketing: a Case Study Approach. Routledge.
Pedersen, E.R.G., Netter, S., 2015. Collaborative consumption: business model op-
portunities and barriers for fashion libraries. J. Fash. Mark. Manag. 19 (3),
258e273.
Piscicelli, L., Cooper, T., Fisher, T., 2014. The role of values in collaborative con-
sumption: insights from a product-service system for lending and borrowing in
the UK. J. Clean. Prod. 97, 21e29.
Roos, S., 2016. Advancing Life Cycle Assessment of Textile Products to Include
Textile Chemicals: Inventory Data and Toxicity Impact Assessment. Doctoral
thesis. Chalmers University of Technology. series 4202, ISSN: 0346e718X,
Gothenburg, Sweden.
Roos, S., Peters, G., 2015. Three methods for strategic product toxicity assessment e
the case of the cotton t-shirt. Int. J. Life Cycle Assess. 20 (7), 903e912.
Roos, S., Sandin, G., Zamani, B., Peters, G., 2015. Environmental Assessment of
Swedish Fashion Consumption: Five Garments e Sustainable Futures. A Mistra
Future Fashion report, Gothenburg. Available at: http://www.mistrafuturefashio
n.com (Accessed January 2016).
Roos, S., Zamani, B., Sandin, G., Peters, G.M., Svanstro € m, M., 2016. Will clothing be
sustainable? Clarifying sustainable fashion. In: Muthu, S.S. (Ed.), Textiles and
Clothing Sustainability e Implications in Textiles and Fashion. Springer,
Singapore.
Rosenbaum, R.K., Bachmann, T.M., Swirsky Gold, L., Huijbregts, M.A.J., Jolliet, O.,
Juraske, R., et al., 2008. USEtox e the UNEP-SETAC toxicity model: recom-
mended characterisation factors for human toxicity and freshwater ecotoxicity
in life cycle impact assessment. Int. J. Life Cycle Assess. 13 (7), 532e546.
Spielmann, M., Bauer, C., Dones, R., Tuchschmid, M., 2007. Transport Services, Data
v2.0. Swiss Center for Life Cycle Inventories, Switzerland.
Statistics Sweden, 2015. Statistics Database. Available at: http://www.scb.se/sv_/Hi
tta-statistik/Statistikdatabasen.
Struijs, J., Beusen, A., van Jaarsveld, H., Huijbregts, M.A.J., 2009. Aquatic eutrophi-
cation. In: Goedkoop, M., Heijungs, R., Huijbregts, M.A.J., De Schryver, A.,
Struijs, J., Van Zelm, R. (Eds.), ReCiPe 2008. A Life Cycle Impact Assessment
Method Which Comprises Harmonised Category Indicators at the Midpoint and
the Endpoint Level. Report I: Characterisation Factors, first ed.
Thinkstep, 2015. GaBi LCA Software [WWW Document]. http://www.thinkstep.co
m/software/gabi-lca/.
Tukker, A., 2004. Eight types of producteservice system: eight ways to sustain-
ability? Experiences from SusProNet. Bus. Strateg. Environ. 1384, 246e260.
Woolridge, A.C., Ward, G.D., Phillips, P.S., Collins, M., Gandy, S., 2006. Life cycle
assessment for reuse/recycling of donated waste textiles compared to use of
virgin material: an UK energy saving perspective. Resour. Conserv. Recycl. 46
(1), 94e103.