Content uploaded by Bernhard Wohner
Author content
All content in this area was uploaded by Bernhard Wohner on Jun 11, 2020
Content may be subject to copyright.
Environmental and economic assessment of food-packaging systems
with a focus on food waste. Case study on tomato ketchup
Bernhard Wohner
a,
⁎,Viktoria Helene Gabriel
a,b
, Barbara Krenn
a
,Victoria Krauter
a
, Manfred Tacker
a
a
Section Packaging and Resource Management, University of Applied Sciences Campus Vienna, Helmut-Qualtinger-Gasse 2/2/3, Vienna 1030, Austria
b
Circular Analytics TK GmbH, Otto-Bauer-Gasse 3/13, Vienna 1060, Austria
HIGHLIGHTS
•Ketchup waste due to poor emptiability
ranged from 3.85% (±0.41) to 28.80%
(±3.30).
•Emptiability of ketchup in glass packag-
ing is better than in polypropylene bot-
tles.
•Glass packaging has greater environ-
mental impacts than polypropylene
bottles.
•Including packaging-related FLW can
alter the ranking of products.
•Poor emptiability increases costs to the
consumer but also economic value
added.
GRAPHICAL ABSTRACT
abstractarticle info
Article history:
Received 14 January 2020
Received in revised form 25 May 2020
Accepted 29 May 2020
Available online 01 June 2020
Editor: Deyi Hou
Keywords:
Life cycle assessment
Multi-criteria decision analysis
Circular economy
Food packaging
Value added
Food waste
In this paper, a sustainability evaluation method for food-packaging systems is proposed. First,food waste due to
poor emptiability was determined. Then, these quantities were included in life cycle assessments (LCA) and life
cycle costing (value added, VA) of the products. Finally, LCA and VA results were combined using multi-criteria
decision analysis, Technique for Order by Similarity to Ideal Solution (TOPSIS), in order to identify themost sus-
tainable food packaging system.
As a case study, four different ketchup products were examined. For ketchup in polypropylene bottles, FLW
resulting from poor emptiability ranged from 13.12% (±2.05) to 28.80% (±3.30) respectively, while this was
only 3.85% (±0.41) for ketchup packaged in glass. After integrating the emptiability results into lifecycle assess-
ments, this resulted in greenhousegas emissions of 5.66 to 9.16 kg CO
2eq
per 3.80 kg consumed ketchup, the av-
erage consumption per capita in Austria. Importantly, poor emptiability of the examined products led to greater
environmental impacts than the associated packaging. While greater product loss also pushes up the costs for
consumers, it contributes to more value added to the economic system, which is in stark contrast to the goal of
decoupling the economy from resource consumption.
© 2020 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://
creativecommons.org/licenses/by/4.0/).
Science of the Total Environment 738 (2020) 139846
Abbreviations: AC, acidification; CC, climate change; CNV, conventional agriculture; CRITIC, Criteria Importance through Intercriteria Correlation; EMPT, emptiability; FEU,
eutrophication,freshwater; FLW,food losses and waste;FRD, resourceuse, fossils; FU,functional unit;GL, glass; LCA, lifecycle assessment;MCDA, multi-criteria decision analysis; PP, poly-
propylene;ORG, organic agriculture; PEF,product environmental footprint; PM, particulate matter;TOPSIS, techniquefor order by similarityto ideal solution; VA,value added; WU, water
use.
⁎Corresponding author.
E-mail address: bernhard.wohner@fh-campuswien.ac.at (B. Wohner).
https://doi.org/10.1016/j.scitotenv.2020.139846
0048-9697/© 2020 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Contents lists available at ScienceDirect
Science of the Total Environment
journal homepage: www.elsevier.com/locate/scitotenv
1. Introduction
Today, the world's economy is mainly based on a linear model. Re-
cent studies suggest that globally, only 9% of all raw materials are
reused, recycled or composted after their use (de Wit et al., 2018).
Concerning the European Union (EU), only 67% of packaging and 46%
of municipal waste is currently recycled (eurostat, 2019a). As a result,
initiating a transformation towards a circular economy by adopting
the ‘Circular Economy Package’has become one of the top priorities of
the EU (European Commission, 2019b).
This package includes goals suchas requiring full recyclability or re-
usability of packaging (European Commission [DG ENV - Directorate C],
2018), increased recycling quotas of packaging as well as halving food
waste by 2050 or 2030 respectively (The European Parliament and the
Council, 2018). In Austria, only 25% of plastic packaging is currently
recycled (Altstoff Recycling Austria AG, 2018), meaning that this must
be approximately doubled by 2030 to fulfill the mandatory quota of 55%.
As a possible solution, in addition to increasing the recyclability of
plastic packaging, a general reduction of plastic is highly discussed.
Such a reduction has further gained fresh prominence due to increasing
public disdain concerning plastic. This has been addressed by several
Austrian food retailers, who declared the reduction of plastic packaging
in their mission statements (HOFER, 2018;REWE Group, 2018;SPAR,
2019). Furthermore, the reduction of plastic packaging by 20% to 25%
was officially declared a goal of the Austrian government in 2018
(Bundeskanzleramt, Bundesministerium Öffentlicher Dienst und Sport,
Bundesministerium Nachhaltigkeit und Tourimus, 2018). However, en-
vironmental benefits of reducing the quantity of plastic packaging could
even lead to greater environmental impacts when it is substituted by
other materials such aspaper, glass or metal (Pilz et al., 2010). Further-
more, a reduction or substitution of plastic packaging could increase the
generation of food loss and waste (FLW) (Pauer et al., 2019;Wohner
et al., 2019a). While muchresearch has been carried out on the evalua-
tion of direct environmental impacts of packaging by conducting life
cycle assessments (LCA), there is still very little scientific understanding
of indirect effects (Molina-Besch et al., 2018;Wohner et al., 2019a).
Since protecting food is in fact the main function of packaging
(Lindh et al., 2016;Pauer et al., 2019), sustainability evaluations of
packaging should not be carried out without considering its impact
on the filling good and thus of holistic evaluations of food together
with its associated packaging (food-packaging systems) (Pauer
et al., 2019). With 14% of food being lost between post-harvest and
retail level (FAO, 2019) together with older estimates of 30% being
lost across the whole supply chain (Gustavsson et al., 2011), it is
clear however that FLW and therefore indirect effects of packaging
are a pressing concern. Several authors already focus on assessing
FLW by using LCA (Beretta and Hellweg, 2019;Scherhaufer et al.,
2018), with an increasing number of authors integrating FLW into
the LCA of packaging (Molina-Besch et al., 2018). Among other as-
pects, this includes (i) FLW related to packaging being difficult to
empty (Meurer et al., 2017;Williams et al., 2012;Williams and
Wikström, 2011;Wohner et al., 2019b), (ii) calculation of break-
even rates between the volume of packaging material and FLW
(Bacenetti et al., 2018;Yokokawa et al., 2018) or (iii) modelling the
quantity of FLW based on shelf life (Conte et al., 2015).
According to Pauer et al. (2019), evaluations of food-packaging sys-
tems should include direct and indirect environmental effects, in addi-
tion to circularity assessments, yet without proposing a combined
evaluation method. Niero and Kalbar (2019) already combined direct
environmental effects (LCA results) of packaging and circularity metrics
using multi-criteria decision analysis (MCDA). In context of this re-
search, however, we argue that circularity parameters such as recycled
content or recycling quotas may affect LCA results, thus violating the
rules of using only independent attributes in MCDA (Belton and
Stewart, 2003).
In summary, the aim of the present paper is to analyze packaging-
related FLW of food-packaging systems in order to integrate it into en-
vironmental and economic assessments. Against this background, a
case study on tomato ketchup is conducted. Emptiability is quantified,
which is then integrated into LCA and life cycle costing (LCC) of the
products. Finally, the most sustainable product is identified by using
multi-criteria decision analysis (MCDA).
Fig. 1. Ketchup products chosen as illustrative examples. a) Conventional ketchup, produced in Austria, 450 g indicated filling quantity, 29.99 g colored polypropylene (PP) bottle with
10.81 g colored PP cap, 0.28 g multilayer seal (assuming a composition of 52% polyethylene, 25% polyethylene terephthalate, 17% adhesive and 6% aluminum) and 0.97 g PP labels.
172 g tomatoes per 100 g ketchup. Sales price: 1.99 €(PP-450-CONV). b) Organic ketchup, produced in Austria, 380 g indicated filling quantity, 22.30 g clear transparent PP bottle
with 4.36g colored PP cap, 0.29g multilayer seal and0.63 g PP labels. Salesprice: 2.99 €(PP-380-ORG). c) Organic ketchup,produced in the CzechRepublic, 550 g indicatedfilling quantity,
30.96 g clear transparent PP bottlewith 9.79 g coloredPP cap, 0.32 g multilayer sealand 1.27 g paper labels.210 g tomatoesper 100 g ketchup. Sales price:1.99 €(PP-550-ORG).d) Organic
ketchup, produced in Italy, 480 g indicated filling quantity, 236.61 g flint packaging glass with 4.88 g tinplate screw cap and 1.29 g paper labels. 225 g tomatoes per 100 g ketchup. Sales
price: 1.45 €(GL-480-ORG).
2B. Wohner et al. / Science of the Total Environment 738 (2020) 139846
2. Materials and methods
In this section, we first present the case study. Based on this, selected
criteria and their quantification is discussed. Finally, the selection and
calculation of a suitable method for the sustainability evaluation is
presented.
2.1. Case study: tomato ketchup
Tomato ketchup was chosen as a case study. Ketchup is made from
fresh tomatoes or tomato puree, sugar and/or sweetener, spices and
seasoning, salt and vinegar. The final product must have a minimum
of 28% dry mass (Bundesministerium für Arbeit, Soziales, Gesundheit
und Konsumentenschutz, 2015). In Austria, 3.8 kg of ketchup is con-
sumed per capita and year (Statista GmbH, 2019).
The following products of different brands were purchased at vari-
ous supermarket chains (Fig. 1):
2.2. Life cycle assessment
Life cycle assessment is a well-known method to assess environ-
mental impacts across the life cycle of a product, frequently used in
food and food packagingstudies (Fraval et al., 2019). LCA for this article
was based on ISO 14040 (ISO, 2006a) with additional guidance from the
Product Environmental Footprint (PEF) (European Commission, 2017),
which is being currently developed by the European Commission. In
contrast to ISO 14040, the PEF guidance includes stricter recommenda-
tions. For this study, the PEF guidance was used for:
•Selection of life cycle impact categories
•Identification of the most relevant life cycle impact categories
•Default transport distances
•Allocation regarding input and output of secondary materials
Calculations were performed using OpenLCA and the Ecoinvent 3.5
database. LCA for the case study was limited to secondary data only.
This type of LCA method can be considered as ‘streamlined LCA’,
which has the benefit of reducing the expenditure of time and resources
(Speck et al., 2015).
2.2.1. Functional unit, reference flow and system boundaries
The functional unit (FU) was defined as ‘consumption of 3.8 kg
ketchup’. This led to different reference flows for the examined prod-
ucts, determined by the loss of ketchup due to poor emptiability. As
an example, if 50% of food loss and waste(FLW) occurs at the consumer,
all environmental impacts up to the point of loss are doubled
(Wikström et al., 2014). System boundaries and the resulting presented
life cycle stages include:
•Packaging: Raw materials, manufacturing of glass and plastic bottles,
transport of empty bottles to the ketchup production site, disposal
of packaging
•Ketchup processing: Cultivation of tomatoes and sugar, thermal and
electrical energy used in the production of ketchup
•Transport of the final product to an Austrian supermarket
•Transport of the final product from the supermarket to the home of
the consumer
•Food loss and waste: Calculated as the difference between provi-
sioned and consumed ketchup
2.2.2. Life cycle inventory of packaging manufacturing
Ketchup bottles were first emptied (see Section 2.2.8) before the
packaging was disassembled and weighed. Packaging manufacturing
was then modelled using Ecoinvent datasets, taking the respective
datasets for the raw materials and their manufacturing processes. No
recycled content was assumed for plastic packaging and 40% for flint
glass bottles (European Commission, 2019a). Transport distances be-
tween the packaging manufacturers to the ketchup production site
were assumed to be (i) 230 km by truck, (ii) 280 km by train and (iii)
87 km by ship for plastic bottles. For glass bottles a transport of
(i) 350 km by truck, (ii) 39 kmby train and (iii) 87 km by ship was cho-
sen (European Commission, 2017).
2.2.3. Life cycle inventory of agricultural production
For the life cycle inventory of ketchup, the quantity of tomatoes used
in processing was taken from the label. From this, the quantity of added
sugar was calculated after subtracting the stated sugar content from the
sugar contained in the tomatoes, assuming a sugar content of 2.6% and a
water content of 95% of the average fruit (USDA, 2019). Among the ex-
amined products wereones of organic and conventional agriculture. Or-
ganic farming is often associated with reduced farm inputs and higher
soil carbon sequestration, therefore reducing environmental impacts
compared to conventional agriculture. However, there is an ongoingde-
bate concerning the actual sustainability of organic agriculture, since
this agricultural practice often leads to lower yields, which increases
greenhouse gas emissions in some cases (Smith et al., 2019). Regarding
tomatoes, organic agriculture may have lower (He et al., 2016;Ronga
et al., 2019) or higher yields (Stanhill, 1990), which in turn leads to
lower (He et al., 2016) or higher (Ronga et al., 2019;Vermeulen and
CJM, 2011) environmental impacts compared to conventionaltomatoes.
Moreover, comparative LCA studies of organic and conventional agricul-
ture are not always able to capture the differences (Meier et al., 2015).
For this paper, it was assumed that organic agriculture is a beneficial
concerning sustainability due to it having multiple ecological and social
benefits, such as greater biodiversity and fewer potential negative ef-
fects on human health (Shennan et al., 2017). Nonetheless, there is no
Ecoinvent dataset available for organic tomatoes. Since the impact of or-
ganic agriculture could not be considered in the LCA, it wasincluded as
an additional criterion. Quantification of organic agriculture was carried
out by assigning a value of ‘1’for products of organic, and a value of ‘0’
for products of conventional agriculture. Other ingredients of tomato
ketchup such as vinegar and spices were excluded from the analysis
due to their small and unknown quantities.
2.2.4. Life cycle inventory of ketchup processing
In the manufacturing process of ketchup, tomatoes are heated with
steam to up to 99 °C (Amón et al., 2015). Thermal energy consumption
of this process was calculated as the product of the latent heat of vapor-
ization of water at 100 °C(2.26 MJ/kg)and the volumeof water needed
to be evaporated to achieve the final water content of the respective
ketchup. This water content was estimated as the difference between
100% and the sum of carbohydrates, fat, protein and assumed average
ash content of 3% (Sharoba et al., 2005). It was assumed that waste
heat is not recovered (Amón et al., 2015). The electricity consumption
of ketchup manufacturing was taken from existing literature
(Andersson et al., 1998). Country-specific electricity mixes and trans-
port distances to Austria were considered, with a modal split of 75%
lorry and 25% freight train (eurostat, 2019b)(eurostat, 2019b)forinter-
national transports. The following distances for the transport of the final
products between productionssites and Austrian retail were estimated:
•Ketchup produced in Austria: 200 km
•Ketchup produced in the Czech Republic: 375 km
•Ketchup produced in Italy: 950 km
2.2.5. Transports of final products
The transport of the final products between the supermarket and the
home of the consumer was assumed to be 5 km, of which 62% were al-
located to a passenger car with a trunk of load of 200 l, 5% to a van and
33% were not allocated (Castellani et al., 2018;European Commission,
3B. Wohner et al. / Science of the Total Environment 738 (2020) 139846
2019a). As a result, the distribution of 1 l ketchup is associated with
0.0155 km driven by passenger car.
A summary of data concerning the modelled foreground system is
presented in Table 1.
2.2.6. Selection of impact categories
Initially, all 16 impact categories recommended by PEF (Castellani
et al., 2018) were calculated. Then, the PEF guidance was followed for
the selection of the most relevant impact categories.
First, all impact categories were normalized, meaning that their
magnitude of relative to a reference information (ISO, 2006b) (in the
context of PEF the impacts of an average world citizen per year) were
calculated. Next, the normalized values were weighted using the values
provided by the PEF guidance. Accordingly, the three toxicity impact
categories shall not be used for benchmarking with assigned weights
of 0%, since their methodology is not yet considered as robust enough.
Finally, the most relevant impact categories were identified based on
the ones that contribute at least 80% to the total sum (European Com-
mission, 2017). Relevant impact categories were the same for all prod-
ucts. This is also true for their order of contribution except for GL-480-
ORG, where the ranks of particulate matter and acidification are
swapped (Table 2).
Consequently, results of the most relevant impact categories per
functional unit are used as criteria in the MCDA. Normalized and
weighted results were only used for the procedure of selecting the
most relevant impact categories. Results of all impact categories, their
respective contribution to the total, as well as normalization and
weighting factors are listed in the supplementary material.
2.2.7. End-of-life and allocations
The use of recycled content and the disposal of the packaging was
modelled according to the Circular Footprint Formula listed in the PEF
guidance (European Commission, 2017). Energy savings of 2.5% per
10% recycled content are assumed for the production of glass bottles
(Stettler et al., 2016). Life cycle inventory data of plastic recycling pro-
cesses in Austria was taken from literature (van Eygen et al., 2018b),
with quality factors of recyclate of 1.00 for glass and metal (European
Commission, 2019a), as well as 0.67 for polypropylene (calculated as
the average ratio of market prices between September 2018 and 2019
(plasticker et al., 2019)).
For this article, it was assumed that PP bottles contaminated with
ketchup can be recycled. However, this might not be true since ketchup
residues may affect the sorting and/or recycling process as has been
shown for PET bottles (Boesveld, 2011). It was assumed that all PP bot-
tles consist of 5% by weight of ethylene vinyl alcohol (Hedenqvist,
2018), which still allows the bottle to be recycled (FH Campus Wien,
2019). Consequently, the only non-recyclable packaging components
were multilayer seals and paper labels.
Recycling rates in Austria are 14% for polypropylene bottles (van
Eygen et al., 2018a), 84% for glass and 86% for metal packaging
(eurostat, 2019c). Polypropylene caps are currently not recycled in
Austria (van Eygen et al., 2018a). Due to landfill restrictions in Austria
(Bundesministerium für Land- und Forstwirtschaft, Umwelt und
Wasserwirtschaft, 2008), only non-recycled quantities of metal and
glass packaging were assumed to be landfilled, while non-recycled plas-
tic packaging was assumed to be incinerated.
2.2.8. Indirect environmental effects due to FLW
Quantifying packaging-related FLW is challenging (Wohner et al.,
2019a) and therefore often omitted in studies of food-packaging sys-
tems (Molina-Besch et al., 2018). In a previous study we proposed a
method for testing dairy products on their ‘technical emptiability’and
its integration in LCA studies (Wohner et al., 2019b) as a possibility to
measure packaging-related FLW. For the present case study, not only
technical but also practical emptiability was tested. Finally, the results
of practical emptiability were taken to calculate the respective reference
flows of the investigated products associated with the functional unit.
Practical emptiability simulates an average emptying behavior by
the consumer. For plastic bottles, first the bottles were shaken three
times and squeezed until air was released. Next, the bottles were
swiveled and then squeezed again until air was released. This step
was repeated three times. Glass bottles were shaken three times and
then held upside down for 2 min. Subsequently, the bottles were shaken
three times and then held upside for 1 min.
Technical emptiability represents thebest possibleemptying proce-
dure without damaging the packaging. For this, both glass and plastic
bottles including their caps were scraped with a dedicated ketchup
spoon (length of 24.5 cm) after practical emptiability tests.
Finally, the emptiability index was expressed as the ratio of ketchup
left in the bottle to the orig inal filling quantity. Testingwas performed at
room temperature (22 °C ± 1). Based on previous studies, a sample size
of 6 was taken to assure significant results (Meurer et al., 2017;Wohner
et al., 2019b).
2.3. Economic assessment
Life cycle costing is an approach often used for the economic evalu-
ation of a product. ‘Conventional’LCC represents the historic practice of
economic assessments, which includes costs associated with a product
and which are generally presented only from one, the producer's or
consumer's, perspective (Hunkeler et al., 2008). Further, conventional
LCC is often performed not all along the entire supply chain, often ex-
cluding End-of-Life operations. In contrast, ‘environmental LCC’is per-
formed alongside LCA, using the same system boundaries and models
and thus covering thewhole life cycle of aproduct.Moreover, by includ-
ing the full life cycle, environmental LCC enables the economic evalua-
tion of a product from a system's perspective. Therefore, according to
Hunkeler et al. (2008), environmental LCC should be the approach of
choice for sustainability assessments. Hence, the economic evaluation
in this paper is conducted taking the ‘value added’approach (VA). Gen-
erally, the revenues (R) for selling a product are higher than its produc-
tion costs (C) (Heijungs et al., 2013), resulting in a margin which is
referred to as “added value”, given in a monetary unit, in this study
Euro (€).
VA ¼R−C
Consequently, the total life cycle cost is the “sum of all value added
over the life cycle”(Moreau and Weidema, 2015). Since environmental
impacts are already covered by theLCA, their associated costs are not in-
cluded in VA, as this would be considered as double-counting.
In this paper, VA is calculated following the same principles as for
the LCA. Therefore, the final VA result is the sum of value added by the
production and disposal of ketchup, its packaging and all related trans-
port, with additional consideration of the final sales price. This can be
expressed as follows:
VATotal ¼VAIN−CIN þVAEN−CEN þVAPA −CPA þVATR −CTR þRPU −CPU
þVAEoL
where:
•VA
Total
: Total VA of the respective product
•VA
IN
: VA of agricultural production of ingredients (tomatoes and
sugar) (calculated as the totalof the difference between costs for pro-
ducing and revenues of selling tomatoes or sugar, and the VA for all
upstream processes)
•C
IN
: Costs to the ketchup producer for purchasing ingredients
•VA
EN
: VA of thermal and electrical energy production (calculated as
the total of the difference between costs for producing and revenues
of selling energy, and the VA for all upstream processes)
•C
EN
: Costs to the ketchup producer for purchasing energy
4B. Wohner et al. / Science of the Total Environment 738 (2020) 139846
•VA
PA
: VA of packagingproduction (calculated as the total of the differ-
ence between costs for producing and revenues for selling packaging,
and the VA for all underlying processes)
•C
PA
: Costs to the ketchup producer for purchasing packaging
•VA
TR
: VA of transports (calculated as the total of difference between
costs and revenues for providing transport, and the VA for all up-
stream processes)
•C
TR
: Costs to the ketchup producer for the transport of products
•R
PU
: Revenue to the ketchup producer for selling ketchup to the con-
sumer
•C
PU
: Costs to the consumer for purchasing ketchup from the producer
•VA
EOL
: VA of disposal of ketchup and packaging (calculated as thetotal
of the difference between costsand revenuesof recycling or incinera-
tion of ketchup or packaging, and the VA for all upstream processes)
For the calculation, default values available in the Ecoinvent 3.5 da-
tabase version of OpenLCA were taken (Ciroth, 2016a). In OpenLCA,
prices already contained but hidden in several Ecoinvent datasets
were made visible by the software publisher, with information on
costs added to further datasets(C iroth, 2016b). Similar to the conducted
LCA, a major limitation is that possible differences between organic and
conventional tomatoes could not be considered due to a lack of data in
Ecoinvent.
2.4. Multi-criteria decision analysis
2.4.1. Selection and calculation
The examined products show different results between LCA impact
categories, as well as between LCA and VA results in general. Hence,
the need for a method to decision making tool arises, able to solve
multi-dimensional issues. In this context, multi-criteria decision analy-
sis methods are increasingly used to identify the best possible solution
out of several alternatives (Wątróbski et al., 2019a). Based on the listed
criteria, a suitable MCDA method was defined as being able to (i) take
different weights into account, (ii) compare criteria on a quantitative
scale and (iii) generate a ranking. Using the MCDA tool (Wątróbski
et al., 2019b), TOPSIS (Hwang et al., 1993) was identified as a method
meeting these requirements. The following terms are defined for better
readability and are frequently used in MCDA:
•Alternative: Several predetermined,limited and independent alterna-
tives. For this study, these are the four examined products (Alinezhad
and Khalili 2019).
•Criterion: A particular perspective according to which alternatives
may be compared (Belton and Stewart, 2003). In the context of this
study, these are comprised of the six chosen LCA impact categories
and the VA.
•Attribute: a “quantitative or qualitative measure of performance asso-
ciated with a particular criterion”(Belton and Stewart, 2003), which
can be either beneficial (with the goal of maximization) or non-
beneficial (with the goal of minimization). In this study, theattributes
are the results of VA and the chosen LCA impact categories, with the
former considered as being beneficial, and the latter as being non-
beneficial.
•Normalization: Converting attributes into non-dimensional form to
allow their aggregation into a final score (Jahan and Edwards, 2015;
Vafaei et al., 2016)
The general calculation steps of TOPSIS can be summarized as fol-
lows (ÇELEN, 2014;Hwang et al., 1993;Kumar et al., 2017):
Table 1
Summary of data for modelling the foreground system. Abbreviations for products represent (i) the packagingmaterial as polypropylene (PP) or glass (GL), (ii) the content of bottles of
380, 450, 480or 550 g and (iii) if the ketchup is a productof conventional(CNV) or organic (ORG)agriculture. Data are given per kg produced and distributedketchup. Remaining abbre-
viations represent: PP, polypropylene; vkm, vehicle-kilometer; tkm, ton-kilometer.
Unit PP-450-CNV PP-380-ORG PP-550-ORG GL-480-ORG
Ingredients Tomatoes kg 1.72 2.85 2.10 2.25
Added sugar kg 0.14 0.15 0.09 0.16
Energy consumption for processing Electricity MJ 0.38 0.38 0.38 0.38
Thermal energy (steam) MJ 2.34 4.82 2.97 3.62
Packaging PP bottle (blow moulded) g 66.50 59.17 55.60 0
Glass bottle g 458.84 0 0 0
PP cap (injection moulded) g 23.97 11.57 17.58 0
Tinplate cap g 0 0 0 9.46
Multilayer seal g 0.62 0.77 0.57 0
PP label g 2.15 1.67 0 0
Paper label g 0 0 2.28 2.50
Transport from manufacturer to retail Lorry tkm 0.22 0.21 0.30 1.03
Freight train tkm 0 0 0.10 0.37
Transport from retail to consumer Passenger car vkm 0.014 0.019 0.014 0.014
Van tkm 0.0001 0.0001 0.0001 0.0002
Table 2
Most relevant life cycle impact categories, in descending order of their relevance.
Impact category Indicator Unit Life cycle impact assessment method
Climate change (CC) Radiative forcing as Global Warming Potential (GWP100) kg CO
2eq
IPCC 2013 (IPCC, 2013)
Resource use, fossils (FRD) Abiotic resource depletion –fossil fuels (ADP-fossil) MJ CML 2002 (Bruijn et al., 2004)
Water use (WU) User deprivation potential (deprivation-weighted water
consumption)
m
3
world
eq
Available Water Remaining (AWARE) (UNEP,
2016)
Eutrophication, freshwater
(FEU)
Fraction of nutrients reaching freshwater end compartment (P) kg P
eq
EUTREND model (Goedkoop et al., 2013)
Acidification (AC) Accumulated Exceedance (AE) mol H +
eq
Accumulated Exceedance (Posch et al., 2008)
Particulate matter (PM) Impact on human health Disease
incidence
PM method (UNEP, 2016)
5B. Wohner et al. / Science of the Total Environment 738 (2020) 139846
1. Creation of a decision matrix
X¼xij
mxn
consisting of malternatives (A
1
,A
2
,…,A
m
)andncriteria (C
1
,C
2
,…,C
n
),
with the intersection of each alternative and criteria given as x
ij
.
2. Normalization of the decision matrix:
rij ¼xij
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
∑m
k¼1x2
kj
q
where i=1,2,…,mand j = 1, 2, …,n
3. Calculation of the weighted normalized decision matrix by multipli-
cation of the normalized matrix with the attribute's weights (w
j
):
vij ¼wjrij;
i=1,2,…,mand j=1,2,…,nwhere wj¼Wj
∑n
i¼1Wi
j = 1, 2, …,n
4. Determination of worst alternative A
w
(or negative ideal solution)
and best alternative A
b
(or positive ideal solution):
Aw¼max vijji¼1;2;…;m
jj∈J−i;minð
hvij i¼1;2;…;m
jÞj j∈Jþ
≡vwj
j¼1;2;…;n
Ab¼min vijji¼1;2;…;m
jj∈J−i;maxðh vij i¼1;2;…;mjÞjj∈Jþ
≡vbj
j¼1;2;…;n;
where for beneficial attributes:
Jþ¼j¼1;2;…;n
fj
j;
and for non-beneficial attributes:
J−¼j¼1;2;…;nfjj
5. Calculation of the Euclidean distance of each alternative to the worst
(d
iw
) and best solution (d
ib
):
diw ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
X
n
j¼1
vij−vwj
2
v
u
u
t
dib ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
X
n
j¼1
vij−vbj
2
v
u
u
t
where i=1,2,…,m
6. Calculation of the relative closeness (CC
i
) of each alternative to the
ideal solution:
CCi¼diw
diw þdib
7. Ranking of the alternatives according to CCi (i =1,2,…,m)
Individual calculation steps of TOPSIS for the case study are listed in
the supplementary material.
2.4.2. Determination of weights
Determination of criteria weights is equally crucial and controversial
since there is an abundant number of methods regarding this procedure
which all produce different results and thus considerably influence the
outcome of an MCDA. Such methods can be classified either (i) a priori,
where weights are determined before data is collected, or (ii) a
posteriori, were the determination of weights occurs after data collec-
tion. While a priori weights are generally elicited by expert interviews
or questionnaires, a posteriori weights are calculated based on the col-
lected data for each alternative (Kao, 2010).
For this paper, three weighting sets were calculated and used for
TOPSIS, namely (i) equal weighting, (ii) Criteria Importance through
Intercriteria Correlation (CRITIC) (Diakoulaki et al., 1995) and (iii) en-
tropy (Li et al., 2011), similar to a sustainability assessment of biodiesel
(Anwar et al., 2019).
2.4.2.1. Equal weighting. Equal weighting is the simplest type of
weighting method, in which each criterion is given the same impor-
tance. In this study, 8 criteria were selected, which results in a weight
(w
j
) of 12.5% per criteria.
wj¼1
8
2.4.2.2. Weights of criteria using CRITIC. Calculating weights using CRITIC
is performed by characterizing each vector by its standard deviation and
a subsequent construction of a symmetric matrix with linear correlation
coefficients between the vectors (Alinezhad and Khalili 2019).
First, the decision matrix is normalized as follows:
xij ¼rij−r−
i
rþ
i−r−
i
xij ¼rij−rþ
i
r−
i−rþ
i
;
where i=1,…,mand j=1,…,nand x
ij
representing the normalized
value for alternative iand attribute j,with
rþ
i¼max r1;r2;…;rm
ðÞ
r−
i¼min r1;r2;…;rm
ðÞ
Then, the correlation coefficient between attributes is calculated as
follows:
ρjk ¼∑m
i¼1xij−xj
xik−xk
ðÞ
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
∑m
i¼1xij−xj
2∑m
i¼1xik−xk
ðÞ
2:
q
with xjandxkrepresenting the mean of jth and kth attributes, calculated
as
xj¼1
nX
n
j¼1
xij
xk¼1
nX
n
k¼1
xik;
where i = 1, 2, …,m.
After that, the standard deviation of each attribute is calculated as
σj¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1
n−1X
n
j¼1
xij−xj
2
v
u
u
t;
where i=1,…,m.
6B. Wohner et al. / Science of the Total Environment 738 (2020) 139846
Next, the index (C) is calculated as:
Cj¼σjX
n
k¼1
1−ρjk
Finally, the weight of attributes is derived by:
wj¼Cj
Pn
j¼1Cj
2.4.2.3. Weights of criteria using entropy. First, the decision matrix is nor-
malized as follows:
rij ¼rij
∑m
i¼1rij
;
where j=1,2,…,nandrij is the normalized value ofthe decision matrix.
Then, the degree of entropy is determined:
Ej¼−1
lnmX
m
i¼1
rij lnrij;
where j=1,2,…,nand 0bE
j
b1.
Next, the deviation rate is calculated by:
dj¼1−Ej;
where j=1,2,…,n.
Finally, weights of attributes are derived by:
wj¼dj
∑n
j¼1dj
3. Results and discussion
3.1. Emptiability
Practical emptiability of the examined bottles ranges from 3.85% (±
0.41) to 28.80% (±3.30), while this can be substantially reduced to be-
tween 3.37% (±0.29) and 7.08% (±0.61) when a spoon is used (‘techni-
cal emptiability’)(Fig. 2). Variability was calculated as 95% confidence
intervals.
In previous studies, the emptiability index of ketchup was reported
as 0.5% to 26% (Andersson et al., 1998) in PP bottles and 30% to 52%
(Boesveld, 2011) in PET bottles, which shows that the quantity of
ketchup remaining in the package can even be higher.
From the figure above (Fig. 2), it is apparent that the product in a
glass bottle (GL-480-ORG) has the best emptiability. In contrast, PP-
380-ORG has the poorest. Important to emphasize is that emptiability
is a function of both product and packaging, thus not allowing the gen-
eralization of glass being better than plastic packaging, since the prod-
ucts in different packages were not identical. Emptiability is mainly
influenced by the packaging geometry, the surface tension of food and
packaging, and particularly by the viscosity of food (Schmidt, 2011). Be-
sides processing conditions, viscosity of ketchup increases with its to-
mato content. Since the product with the highest tomato content
yielded the worst emptiability, this may result in being one of the
major drivers of FLW. One major limitation here is that the portioning
behavior of the products could not be considered. With the glass bottle,
dosing may be more difficult than with the plastic bottles. This could
lead to the consumer emptying more ketchup than required which
may ultimately result in disposing of it.
Statistical analysis was performed using one-way ANOVA (Fisher's
with Tukey post hoc test for samples with equality of variances and
Welch's with Games-Howell post hoc test for samples without equality
of variances), after testing for normality with Shapiro Wilk tests. All sta-
tistical tests were performed with the software ‘Jamovi’(version 1.1.7)
(The jamovi project, 2019) and can be found in the supplementary
material.
3.2. LCA results
Climate change results of all products (Fig. 3a) range from 5.66 to
9.16 kg CO
2eq
per functional unit (FU) respectively. Packaging is respon-
sible for 24% to 26% of the total for PP-450-CNV, PP-550-ORG and GL-
480-ORG, but only 12% for PP-380-ORG due to its high tomato content
and poor emptiability (Fig. 3a-f). In other impact categories, plastic
packaging contributes 7% to 13% and glass packaging 29% to 31% to
the overall result. Obviously, direct environmental impacts of glass
packagingare associated with greater environmental impacts than plas-
tic bottles, which is well in lin e with results of other LCA studies (Boesen
et al., 2019;Humbert et al., 2009;Niero and Kalbar, 2019). Nonetheless,
this is compensated for by its good emptiability.
Concerning the total LCA results, the most influential factors are
FLW, the tomato content and the resulting thermal energy required
for water vaporization. Regarding water use, cultivation of tomatoes is
almost solely responsible for environmental impacts. Taken together,
production and loss of food is substantially more relevant than its asso-
ciated packaging concerning environmental impacts. By contrast, trans-
port is of relatively low importance. One interesting outcome is that LCA
results of PP-550-ORG are better than PP-450-CNV, which would not be
the case if FLW would have been excluded. This finding underlines the
value of quantifying and integrating packaging-related FLW into life
cycle assessments.
Detailed LCA results and results of the remaining calculated impact
categories are listed in the supplementary material.
3.3. Value added results
Value added results for the investigated products (Fig. 4)showa
similar picture to that of the LCA results with the important difference
that here, higher values are considered as beneficial. Therefore, VA re-
sults are in fact diametrically opposed to most of the impact categories
of the performed LCA. This arises mostly from the effect that a greater
material intensity leads to more value added along the supply chain,
PP-450-CNV PP-380-ORG PP-550-ORG GL-480-ORG
Practical emptiability 20.47% 28.80% 13.12% 3.85%
Technical emptiability 7.08% 6.70% 5.12% 3.37%
0.00%
5.00%
10.00%
15.00%
20.00%
25.00%
30.00%
35.00%
xedni ytilibaitpmE
Practical emptiability Technical emptiability
Fig. 2. Emptiability results of examinedketchup products. Bars represent the mean,while
error bars are 95% confidence intervals (n= 6). Abbreviations for products represent
(i) the packaging material as polypropylene (PP) or glass (GL), (ii) the content of bottles
of 380, 450 or 550 g and (iii) if the ketchup is a product of conventional (CNV) or
organic (ORG) agriculture.
7B. Wohner et al. / Science of the Total Environment 738 (2020) 139846
which contradicts the goal of eco-economic decoupling (European
Commission, 2011).
Consequently, since the sales price of a product is higher than its
production costs, poorer emptiability also leads to a greater VA result.
For PP-380-ORG, this is particularly clear, since it has the highest tomato
content as well as the poorest emptiability. Furthermore, the calculated
margin regarding the sales price for this product is substantially greater
compared to the others. This is confirmed by other studies indicating
that smaller packages generally generate higher revenues than larger
ones (Yonezawa and Richards, 2016).
In contrast, GL-480-ORG, is not only the one with the lowest sales
price per kg, but also the one with the best emptiability, leading to the
worst VA results in comparison. Using conventional LCC and taking
the consumer's perspective, the results would be exactly the other
6.15E+00
9.16E+00
5.66E+00
6.54E+00
0.00E+00
2.50E+00
5.00E+00
7.50E+00
1.00E+01
PP-450-CNV PP-380-ORG PP-550-ORG GL-480-ORG
kg CO2eq/FU
Climate change
a
9.40E+01
1.37E+02
8.62E+01
9.65E+01
0.00E+00
2.50E+01
5.00E+01
7.50E+01
1.00E+02
1.25E+02
1.50E+02
PP-450-CNV PP-380-ORG PP-550-ORG GL-480-ORG
MJ/FU
Resource use, fossils
b
1.30E+01
2.15E+01
1.28E+01 1.29E+01
0.00E+00
5.00E+00
1.00E+01
1.50E+01
2.00E+01
2.50E+01
PP-450-CNV PP-380-ORG PP-550-ORG GL-480-ORG
UF/³m
Water use
c
1.40E-03
2.11E-03
1.26E-03
1.58E-03
0.00E+00
5.00E-04
1.00E-03
1.50E-03
2.00E-03
2.50E-03
PP-450-CNV PP-380-ORG PP-550-ORG GL-480-ORG
kg Peq/FU
Eutrophicaon, freshwater
d
3.90E-02
6.06E-02
3.54E-02
4.95E-02
0.00E+00
1.00E-02
2.00E-02
3.00E-02
4.00E-02
5.00E-02
6.00E-02
7.00E-02
PP-450-CNV PP-380-ORG PP-550-ORG GL-480-ORG
UF/
q
e+H lom
Acidificaon
f
3.02E-07
4.72E-07
2.82E-07
4.51E-07
0.00E+00
1.00E-07
2.00E-07
3.00E-07
4.00E-07
5.00E-07
PP-450-CNV PP-380-ORG PP-550-ORG GL-480-ORG
[disease incidence/FU]
Parculate maer
g
Fig. 3. Life cycle assessment resultsof most relevant impactcategories forketchup. Abbreviations forproducts represent(i) the packagingmaterial as polypropylene (PP)or glass (GL), (ii)
the content of bottles of 380, 450, 480 or 550 g and (iii) if theketchup is a product of conventional (CNV) or organic (ORG) agriculture.
8B. Wohner et al. / Science of the Total Environment 738 (2020) 139846
way around. Costs to the consumer for eating 3.8 kg ketchup would be
42.35 €for PP-380-ORG, but only 11.12 €for GL-480-ORG. In turn,
from the manufacturer's point of view, a higherloss would be preferable
as the quantity sold would increase. As Wood and Hertwich (2013)
point out, life cycle costing results should generally be maximized
from society's perspective to generate economic growth but minimized
from an individual's perspective to save costs. Consequently, we agree
with Heijungs et al. (2013) who raised the question: “What do we in
fact want to learn from life cycle costing”?
We conclude that taking a system's perspective is more relevant in
the context of sustainability assessments than taking an individual's
perspective. Thus, despite its limitations, we still consider VA as a suit-
able method for performing environmental LCC together with LCA.
Nonetheless, ff this debate is to be moved forward, methods portraying
a broader economic scope should be developed. Previous research has
already demonstrated how not only economic growth, butalso charac-
teristics such as consumer satisfaction, business diversity or long-term
investments could be considered in new methods concerning life cycle
costing (Neugebauer et al., 2016).
3.4. Sustainability evaluation using TOPSIS
After determining LCA and VA results, the decision matrix for TOPSIS
was created (Table 3).
Next, weights were calculated based on the approaches of equal
weighting, CRITIC and entropy (Table 4)describedinSection 2.4.2.
Using CRITIC, VA and organic agriculture are given more, LCA results
less weight compared to equal or entropy weights.
Finally, after following the calculation steps laid out in Section 2.4.1,
the final closeness values using TOPSIS were determined, with the most
sustainable food-packaging system being the one closest to ‘1.00’
(Fig. 5).
Closeness values ofthe products differ greatlydepending on the cho-
sen weighting set. Nonetheless, PP-550-ORG performs best concerning
all three weighting sets, which is followed by GL-480-ORG. The most
striking observation is the difference in performance of PP-380-ORG
and PP-450-CNV, which is the consequence of the higher importance
of LCA results in the entropy and organic agriculture in the CRITIC
weighting set. As discussed in Section 3.3, VA increases with material in-
tensity and FLW. If TOPSIS were calculated with life cycle costs from the
consumer's perspective, this would have a positive impact on theresults
of GL-480-ORG and a negative impact on PP-380-ORG.
Since the study was limited to the use of secondary data, generaliza-
tion of these results is limited. Furthermore, these results areonly appli-
cable to Austria, due to recycling rates of packaging and costs of these
products are only viable for this country. Depending on the country of
marketing, the evaluation could change substantially. Furthermore,
the difference of organic and conventional agriculture could not be cap-
tured in the calculation of LCA and VA, which however was addressed
by considering it as an additional criterion in the MCDA.
4. Conclusions
The main aim of this study was to combine environmental and eco-
nomic assessments of food-packaging systems, including and putting
the focus on indirect effects of food loss. Historically, most LCA studies
of packaging did not consider FLW (Molina-Besch et al., 2018), predom-
inantly due its quantification being challenging (Wohner et al., 2019a).
In this study, FLW was quantified by testing the emptiability of prod-
ucts, which was then integrated into the LCA and VA calculations of
the examined products. As a result, environmental impacts increased,
and more surprisingly, also the value added to the economy, which is,
however, inherent in the respective method (Wood and Hertwich,
2013).
A further limitation is the exclusion of criteria of taste or qual-
ity. A point could be made that PP-380-ORG is the product with
the highest tomato content and thus the one with the highest
quality. However, this is highly subjective and would have to be
the subject of sensory testing which was outwith the scope of
this study.
We conclude and agree with authors of similar previous studies
that TOPSIS assists in overcoming the limitations inherent in LCA
27.08
42.78
23.69
20.15
7.02
17.39
3.61 0.82
-16.77
-30.15
-13.58
-10.69
-4.32
-12.20
-2.05 -0.43
13.02
17.83
11.67 9.85
-40.00
-30.00
-20.00
-10.00
0.00
10.00
20.00
30.00
40.00
50.00
PP-450-CNV PP-380-ORG PP-550-ORG GL-480-ORG
€/FU
Value Add ed
Provision of original quanty Provision of FLW-related quanty Purchase of original quanty
Purchase of FLW-related quanty SUM
Fig. 4. Value added results. Original quantity is 3.8 kg of ketchup, while the quantity due to food loss and waste (FLW) is generated by the respective emptiability of the products.
Abbreviations for products represent (i) the packaging material as polypropylene (PP) or glass (GL), (ii) the content of bottles of 380, 450, 480 or 550 g and (iii) if the ketchup is a
product of conventional (CNV) or organic (ORG) agriculture.
9B. Wohner et al. / Science of the Total Environment 738 (2020) 139846
studies (Maxim, 2014;Niero and Kalbar, 2019), such as only consid-
ering environmental performance, while excluding assessments of
other sustainability dimensions (Zimek et al., 2019) or compliance
with environmental regulations (Levy, 2017). The proposed
sustainability assessment of food-packaging systems can solve
multi-dimensional issues, particularly of conflicting sustainability
goals. TOPSIS provides a single score and therefore an easy to un-
derstand indication of the best possible solution. However, it is
not without its limitations. TOPSIS does not provide a ‘final word’
since the selection of criteria and weights strongly influence the re-
sults, again shown in this study. Furthermore, sustainability may be
considered as a social construct and, arguably, weighting sets
should then only be determined subjectively (Mollayosefiet al.,
2019). While this may be a benefit due to it being highly adaptable
to the preferences of one decision maker, it is then challenging to
compare the results of one such study to those of others (Maxim,
2014). A natural progression of this work would be to apply this
method to an increasing number of different food-packaging sys-
tems. Furthermore, future studies could incorporate social life
cycle assessments to depict all three pillars of sustainability. Addi-
tionally, the economic assessment could be enhanced by developing
environmental LCC methods which cover a more extensive scope of
economic sustainability. Finally, while admittedly challenging, a
greater focus on quantifying FLW besides emptiability and the inte-
gration into such assessments would produce a better and broader
insight into the sustainability of food-packaging systems.
CRediT authorship contribution statement
Bernhard Wohner:Conceptualization, Formal analysis, Validation,
Writing - original draft, Writing - review & editing, Visualization.
Viktoria Helene Gabriel:Conceptualization, Writing - review &
editing.Barbara Krenn:Conceptualization, Formal analysis.Victoria
Krauter:Conceptualization, Validation, Writing - review & editing, Su-
pervision.Manfred Tacker:Conceptualization, Validation, Resources,
Writing - review & editing, Supervision.
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgments
Reinhard Zeilinger assisted with the statistical analysis. Vivienne
Nieuwenhuizen created figures. Mary Wallis provided comments on
the manuscript.
Table 3
Decisionmatrix of TOPSIS for case study. Abbreviationsfor products represent (i) thepackaging materialas polypropylene(PP) or glass (GL), (ii)the content of bottles of 380, 450,480 or
550 g and (iii) if the ketchup is a product of conventional (CNV) or organic (ORG) agriculture.
Abbreviations for criteria represent, beneficial (B) or non-beneficial (NB): CC (climate change), FRD (resource use, fossils), WU (water use), FEU (Eutrophication, freshwater), AC (acid-
ification), PM (Particulate matter), and VA (Value Added).
Type of
criterion
Unit
PP-450-CNV
PP-380-ORG
PP-550-ORG
GL-480-
ORG
LCA
CC
NB
kg CO2eq/FU
5.97E+00
9.16E+00
5.66E+00
6.54E+00
FRD
NB
MJ/FU
9.40E+01
1.37E+02
8.62E+01
9.65E+01
WU
NB
m³eq/FU
1.23E+01
2.15E+01
1.28E+01
1.29E+01
FEU
NB
kg Peq/FU
1.40E-03
2.11E-03
1.26E-03
1.58E-03
AC
NB
mol H+eq/FU
3.90E-02
6.06E-02
3.54E-02
4.95E-02
PM
NB
disease
incidence/FU
3.02E-07
4.72E-07
2.82E-07
4.51E-07
Organic
agriculture
B
yes (1) / no (0)
0
1
1
1
Value added
B
€/FU
13.02
17.83
11.67
9.85
Table 4
Weights of criteria, calculated using equal weighting (‘EQUAL’), CRITIC and entropy. Ab-
breviations for criteria represent: CC (climate change), FRD (resource use, fossils), WU
(water use),FEU (eutrophication, freshwater), AC (acidification),PM (particulate matter),
and VA (value added).
Category Criteria Equal Critic Entropy
Life cycle assessment CC 12.5% 6.8% 14.4%
FOSSILS 12.5% 7.5% 13.9%
WATER 12.5% 8.4% 17.3%
FW_EUTROPH 12.5% 6.8% 14.3%
FW_ACID 12.5% 8.0% 14.2%
RESP 12.5% 15.2% 13.3%
Organic agriculture Yes/no 12.5% 32.2% 7.5%
Economic assessment VA 12.5% 15.2% 5.1%
0.50
0.77
0.33
0.49
0.23
0.65
0.79
0.85
0.90
0.63
0.77
0.65
0.00
0.25
0.50
0.75
1.00
Weights: Equal Weights: CRITIC Weights: Entropy
Ideal soluon = 1.00
PP-380-ORG PP-450-CNV PP-550-ORG GL-480-ORG
Fig. 5. Relative closeness values of products.Abbreviations for productsrepresent (i) the
packaging material as polypropylene (PP) or glass (GL), (ii) the content of bottles of
380, 450, 480 or 550 g and (iii) if the ketchup is a product of conventional (CNV) or
organic (ORG) agriculture.
10 B. Wohner et al. / Science of the Total Environment 738 (2020) 139846
Funding
This research did not receive any specific grant from funding agen-
cies in the public, commercial, or not-for-profitsectors.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.scitotenv.2020.139846.
References
Alinezhad, A., Khalili, J., 2019. CRITIC method. In: Alinezhad, A., Khalili, J. (Eds.), New
Methods and Applications in Multiple Attribute Decision Making (MADM). Springer
International Publishing, Cham, pp. 199–203.
Altstoff Recycling Austria AG, 2018. Transparenzbericht 2018: Die Zukunft der
Kreislaufwirtschaft. Austria, Vienna.
Amón, R., Maulhardt, M., Wong, T., Kazama, D., Simmons, C.W., 2015. Waste heat and
water recovery opportunities in California tomato paste processing. Appl. Therm.
Eng. 78, 525–532.
Andersson, K., Ohlsson, T., Olsson, P., 1998. Screening life cycle assessment (LCA) of to-
mato ketchup: a case study. J. Clean. Prod. 6 (3–4), 277–288.
Anwar, M., Rasul, M.G., Ashwath, N., 2019. The efficacy of multiple-criteria design matrix
for biodiesel feedstock selection. Energy Convers. Manag. 198, 111790.
Bacenetti, J., Cavaliere, A., Falcone, G., Giovenzana, V., Banterle, A., Guidetti, R., 2018.Shelf
life extension as solution for environmental impact mitigation: a case study for bak-
ery products. Sci. Total Environ. 627, 997–1007.
Belton, V., Stewart, T.J., 2003. Multiple Criteria Decision Analysis: An Integrated Approach.
2nd ed. Kluwer Acad. Publ, Boston, Mass.
Beretta, C., Hellweg, S., 2019. Potential environmental benefits from food waste preven-
tion in the food service sector. Resour. Conserv. Recycl. 147, 169–178. http://www.
sciencedirect.com/science/article/pii/S0921344919301284.
Boesen, S.,Bey, N., Niero, M., 2019. Environmental sustainability of liquidfood packaging:
is there a gap betweenDanish consumers’perceptionand learnings from life cycleas-
sessment? J. Clean. Prod. 210, 1193–1206.
Boesveld, E., 2011. The Recycling of Heinz PET Packaging in Europe. Bachelor thesis. En-
schede, Netherlands. https://essay.utwente.nl/62202/4/Verslag_Els_Boesveld.pdf,
Accessed date: 7 November 2019.
Bruijn, H., Duin, R., Huijbregts, M.A.J., Guinee, J.B., Gorree, M., Heijungs, R., et al., 2004.
Handbook on Life Cycle Assessment: Operational Guide to the ISO Standards. Kluwer
Academic Publishers, Dordrecht.
Bundeskanzleramt, Bundesministerium Öffentlicher Dienst und Sport,
Bundesministerium Nachhaltigkeit und Tourimus, 2018. Maßnahmen der
österreichischen Bundesregierung zur Reduktion von Plastikabfällen in Österreich.
38/18.
Bundesministerium für Arbeit, Soziales, Gesundheit und Konsumentenschutz, 2015.
Österreichisches Lebensmittelhandbuch. Codexkapitel/B 24/Gemüse und
Gemüsedauerwaren.
Bundesministerium für Land- und Forstwirtschaft, Umwelt und Wasserwirtschaft, 2008.
Bundesrecht konsolidiert: Gesamte Rechtsvorschrift für Deponieverordnung 2008,
Fassung vom 08.11.2019: DVO 2008.
Castellani, V., Diaconu, E., Fazio, S., Sala, S., Schau, E.M., Secchi, M., et al., 2018. Supporting
Information to the Characterisation Factors of Recommended EF Life Cycle Impact As-
sessment Methods: New Methods and Differences with ILCD. Publications Office of
the European Union, Luxembourg.
ÇELEN, A., 2014. Comparative analysis of normalization procedures in TOPSIS method:
with an application to Turkish deposit banking market. Informatica 24 (2), 185–208.
Ciroth, A., 2016a. Life Cycle Costing and Calculation of Value Added in openLCA. https://
www.openlca.org/wp-content/uploads/2016/08/Life-Cycle-Costing-in-openLCA.pdf,
Accessed date: 15 April 2020.
Ciroth, A., 2016b. Life Cycle Costing in openLCA 1.5 with Ecoinvent 3.2 Extended. https://
www.openlca.org/wp-content/uploads/2016/08/Webinar-Life-Cycle-Costing-in-
openLCA-1.5-with-ecoinvent-3.2-extended.pdf, Accessed date: 14 May 2020.
Conte, A., Cappelletti, G.M., Nicoletti, G.M., Russo, C., Del Nobile, M.A., 2015. Environmen-
tal implications of food loss probability in packaging design. Food Research Interna-
tional (Ottawa, Ont.) 78, 11–17.
Diakoulaki, D., Mavrotas, G., Papayannakis, L., 1995. Determining objective weights in
multiple criteria problems: the critic method. Comput. Oper. Res. 22 (7), 763–770.
European Commission, 2011. Communication from the Commission to the European Par-
liament, the Council, the European Economic and Social Commitee and the Commit-
tee of the Regions: Roadmap to a Resource Efficient Europe. European Commission,
Brussels.
European Commission, 2017. PEFCR Guidance Document: Guidance for theDevelopment
of Product Environmental Footprint Category Rules (PEFCRSs). version 6.3. http://ec.
europa.eu/environment/eussd/smgp/pdf/PEFCR_guidance_v6.3.pdf, Accessed date:
17 April 2019.
European Commission,2019a. Product Environmental Footprint Guidance: Annex C –List
of Default Values for A, R1, R2, R3 and Qs/Qp. https://eplca.jrc.ec.europa.eu/
permalink/Annex_C_Transition_CFF.xlsx, Accessed date: 29 April 2020.
European Commission, 2019b. Circular Economy: Implementation of the Circular Econ-
omy Action Plan. https://ec.europa.eu/environment/circular-economy/index_en.
htm, Accessed date: 11 December 2019.
EuropeanCommission (DG ENV - Directorate C), 2018. A European Strategy for Plastics in
a Circular Economy: Communication from the Commission to the European Parlia-
ment. the Council, the European Economic and Social Committee of the Regions,
Brussels, Belgium.
eurostat, 2019a. Circular Economy in the EU: Record Recycling Rates and Use of Recycled
Materials in the EU. https://ec.europa.eu/eurostat/documents/2995521/9629294/8-
04032019-BP-EN.pdf/295c2302-4ed1-45b9-af86-96d1bbb7acb1, Accessed date: 7
January 2020.
eurostat, 2019b. Freight Transport Statistics - Modal Split - Statistics Explained. https://ec.
europa.eu/eurostat/statistics-explained/index.php?title=Freight_transport_statis-
tics_-_modal_split#Modal_split_in_the_EU, Accessed date: 2 D ecember 2019.
eurostat, 2019c. Statistics | Eurostat: Recycling Rates for Packaging Waste. https://ec.eu-
ropa.eu/eurostat/databrowser/view/ten00063/default/table?lang=en, Accessed
date: 7 November 2019.
van Eygen, E., Laner, D., Fellner, J., 2018a. Circular economy of plastic packaging: current
practice and perspectives in Austria. Waste Management (New York, N.Y.)72, 55–64.
van Eygen, E., Laner, D., Fellner, J., 2018b. Integrating high-resolution material flow data
into the environmental assessment of waste management system scenarios: the
case of plastic packaging in Austria. Environmental Science & Technology 52 (19),
10934–10945.
FAO, 2019. The State of Food and Agriculture 2019: Moving Forward on Food Loss and
Waste Reduction. Rome. .
FH Campus Wien, 2019. Circular Packaging Design Guideline: Design Recommendations
for Recyclable Packaging. https://www.fh-campuswien.ac.at/fileadmin/redakteure/
Veranstaltungen/Circular_Packaging_Day/FH-Campus-Wien_Circular-Packaging-De-
sign-Guideline_V02.pdf, Accessed date: 8 November 2019.
Fraval, S., van Middelaar, C.E., Ridoutt, B.G., Opio, C., 2019. Life cycle assessment of food
products. Encyclopedia of Food Security and Sustainability. Elsevier, pp. 488–496.
Goedkoop, M., Heijungs, R., Huijbregts, M., An, D.S., Jaap, S., vZ, Rosalie, 2013. ReCiPE
2008: A Life Cycle Impact Assessment Method which Comprises Harmonised Cate-
gory Indicators at the Midpoint and the Endpoint Level.
Gustavsson, J., Cederberg, C., Sonesson, U., 2011. Global Food Losses and Food Waste: Ex-
tent, Causes and Prevention; Study Conducted for the International Congress Save
Food! At Interpack 2011, [16–17 May], Düsseldorf, Germany. Rome, Food and Agri-
culture Organization of the United Nations.
He, X., Qiao, Y., Liu, Y., Dendler, L., Yin, C., Martin, F., 2016. Environmental impact assess-
ment of organic and conventional tomato production in urban greenhouses of Beijing
city, China. J. Clean. Prod. 134, 251–258.
Hedenqvist, M.S., 2018. Barrier packaging materials. In: Hedenqvist, M. (Ed.), Handbook
of Environmental Degradation of Materials. Elsevier, pp. 559–581.
Heijungs, R., Settanni, E., Guinée, J., 2013. Toward a computational structure for life cycle
sustainability analysis: unifying LCA and LCC. Int. J. Life Cycle Assess. 18 (9),
1722–1733.
HOFER, 2018. Medieninformation: HOFER packt aus: Diskonter startet Offensive zur
Reduktion von Verpackungsmüll. https://www.hofer.at/de/unternehmen/presse/
presseaussendungen/projekt-2020/verpackungsmission/, Accessed date: 7 Novem-
ber 2019.
Humbert, S., Rossi, V., Margni, M., Jolliet, O., Loerincik, Y., 2009. Life cycle assessment of
two baby food packaging alternatives: glass jars vs. plastic pots. Int. J. Life Cycle As-
sess. 14 (2), 95–106.
Hunkeler, D., Lichtenvort, K., Rebitzer, G., 2008. Environmental Life Cycle Costing. CRC
Press, Hoboken.
Hwang, C.-L., Lai, Y.-J., Liu, T.-Y., 1993. A new approach for multiple objective decision
making. Comput. Oper. Res. 20 (8), 889–899.
IPCC, 2013. Climate Change 2013: The Physical Science Basis: Contribution of Working
Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate
Change. Cambridge University Press, Cambridge.
ISO (Ed.), 2006a. Environmental management —Life cycle assessment —Requirements
and guidelines: 14040:2006(en), 2006th ed. ISO.
ISO (Ed.), 2006b. Environmental Management —Life Cycle Assessment —Requirements
and Guidelines: 14044:2006, 2006th ed. ISO. (13.020.10).
Jahan, A., Edwards, K.L., 2015. A state-of-the-art survey on the influence of normalization
techniques in ranking: improving the materials selection process in engineering de-
sign. Materials & Design (1980–2015) 65, 335–342.
Kao, C., 2010. Weight determination for consistently ranking alternatives in multiple
criteria decision analysis. Appl. Math. Model. 34 (7), 1779–1787.
Kumar, A., Sah, B., Singh, A.R., Deng, Y., He, X., Kumar, P., et al., 2017. A review of multi
criteria decision making (MCDM) towards sustainable renewable energy develop-
ment. Renew. Sust. Energ. Rev. 69, 596–609.
Levy, M., 2017. Life cycle analysis: strengths and limitations of LCA. Encyclopedia of Sus-
tainable Technologies. Elsevier, pp. 233–236.
Li, X., Wang,K., Liu, L., Xin, J.,Yang, H., Gao, C., 2011. Application of the entropy weightand
TOPSIS method in safety evaluation of coal mines. Procedia Engineering 26,
2085–2091.
Lindh, H., Williams, H., Olsson, A., Wikström, F., 2016. Elucidating the indirect contribu-
tions of packaging to sustainable development: a terminology of packaging functions
and features. Packag. Technol. Sci. 29 (4–5), 225–246.
Maxim, A., 2014. Sustainability assessment of electricity generation technologies using
weighted multi-criteria decision analysis. Energy Policy 65, 284–297.
Meier, M.S., Stoessel, F., Jungbluth, N., Juraske, R., Schader, C., Stolze, M., 2015. Environ-
mental impacts of organic and conventional agricultural products—are the differ-
ences captured by life cycle assessment? J. Environ. Manag. 149, 193–208.
Meurer, I.R., Lange, C.C., Hungaro, H.M., Valenzuela Bell, M.J., de Carvalho dos Anjos, V.,
Antonio de Sá Silva, C., et al., 2017. Quantification of whole ultra high temperature
UHT milk waste as a function of packages type and design. J. Clean. Prod. 153,
483–490.
11B. Wohner et al. / Science of the Total Environment 738 (2020) 139846
Molina-Besch, K., Wikström, F., Williams, H., 2018. The environmental impact of packag-
ing in food supply chains—does life cycle assessment of food provide the full picture?
Int. J. Life Cycle Assess. 21, 492 Part 3.
Mollayosefi, M.M., Hayati, B., Pishbahar, E., Nematian, J., 2019. Selecting weighting meth-
odologies for evaluating agricultural sustainability in Iran. Proceedings of the Institu-
tion of Civil Engineers - Engineering Sustainability 172 (4), 207–217.
Moreau, V., Weidema, B.P., 2015. The computational structure of environmental life cycle
costing. Int. J. Life Cycle Assess. 20 (10), 1359–1363.
Neugebauer, S., Forin, S., Finkbeiner, M., 2016. From life cycle costing to economic life
cycle assessment—introducing an economic impact pathway. Sustainability 8 (5),
428.
Niero, M., Kalbar, P.P., 2019. Coupling material circularity indicators and life cycle based
indicators: a proposal to advance the assessment of circular economy strategies at
the product level. Resour. Conserv. Recycl. 140, 305–312.
Pauer, E., Wohner, B., Heinrich, V., Tacker, M., 2019. Assessing the environmental sustain-
ability of food packaging: an extended life cycle assessment including packaging-
related food losses and waste and circularity assessment. Sustainability 11 (3), 925.
Pilz H, Brandt B, Fehringer R. The Impactof Plastics on Life Cycle Energy Consumptionand
Greenhouse Gas Emissions in Europe: Summary Report, 2010. https://www.
plasticseurope.org/application/files/9015/1310/4686/september-2010-the-impact-
of-plastic.pdf (accessed April 15, 2020).
plasticker, Markt, Preise, 2019. https://plasticker.de/preise/preise_monat_single.php,
Accessed date: 7 November 2019.
Posch, M., Seppälä, J., Hettelingh, J.-P., Johansson, M., Margni, M., Jolliet, O., 2008. The role
of atmospheric dispersion models and ecosystem sensitivity in the determination of
characterisation factors for acidifying and eutrophying emissions in LCIA. Int. J. Life
Cycle Assess. 13 (6), 477–486. https://link.springer.com/content/pdf/10.1007/
s11367-008-0025-9.pdf.
REWE Group, 2018. REWE International AG will “Raus aus Plastik”: BILLA, MERKUR,
PENNY und ADEG starten großangelegte Initiative für umweltfreundliche
Verpackungen. https://www.rewe-group.at/de/newsroom/pressemitteilungen/riag-
1817-Raus-aus-Plastik, Accessed date: 7 November 2019 ).
Ronga, D., Gallingani, T., Zaccardelli, M., Perrone, D., Francia, E., Milc, J., et al., 2019. Carbon
footprint and energetic analysis of tomato production in the organic vs the conven-
tional cropping systems in southern Italy. J. Clean. Prod. 220, 836–845.
Scherhaufer, S., Moates, G., Hartikainen, H., Waldron, K., Obersteiner, G., 2018. Environ-
mental impacts of food waste in Europe. Waste Manag. 77, 98–113.
Schmidt, M.C., 2011. Untersuchung und Verbesserung des Entleerungsverhaltens von
Füllgut-Verpackungssystemen. Dissertation. Universität Stuttgart, Stuttgart.
Sharoba, A.M., Senge, B., El-Mansy, H.A., Bahlol, H.E., Blochwitz, R., 2005. Chemical, sen-
sory and rheological properties of some commercial German and Egyptian tomato
ketchups. Eur. Food Res. Technol. 220 (2), 142–151.
Shennan, C., Krupnik, T.J., Baird,G., Cohen, H., Forbush, K., Lovell, R.J., et al., 2017. Organic
and conventional agriculture: a useful framing? Annu. Rev. Environ. Resour. 42 (1),
317–346.
Smith, L.G., Kirk, G.J.D., Jones, P.J., Williams, A.G., 2019. The greenhouse gas impacts of
converting food production in England and Walesto organic methods. Nat. Commun.
10 (1), 4641.
SPAR, 2019. Gemeinsam Plastik sparen mit Spar. https://www.spar.at/nachhaltigkeit/
produkte/verpackungen, Accessed date: 7 November 2019.
Speck, R., Selke, S., Auras, R., Fitzsimmons, J., 2015. Choice of life cycle assessment soft-
ware can impact packaging system decisions. Packag. Technol. Sci. 28 (7), 579–588.
Stanhill, G., 1990. The comparative productivity of organic agriculture. Agric. Ecosyst. En-
viron. 30 (1–2), 1–26.
Statista GmbH, 2019. Ketchup - Österreich | Statista Marktprognose. https://de.statista.
com/outlook/40070100/128/ketchup/oesterreich, Accessed date: 24 October 2019.
Stettler, C., Rüttimann, S., Kägi, T., 2016. Ökobilanz Verwertungen von Altglas -
ökologischer Nutzen der Sammlung von Verpackungsglas: Grundlagenbericht für
Empfehlungen zum Verteilschlüssel der vorgezogenen Entsorgungsgebühr (VEG).
https://carbotech.ch/cms/wp-content/uploads/%C3%96kobilanz_Verwertung_von_
Altglas_de.pdf, Accessed date: 7 November 2019.
The EuropeanParliament and the Council, 2018. Directive (EU) 2018/851 of the European
Parliament and of the Council of 30 May 2018 amending Directive 2008/98/EC on
waste: EU 2018/851.
The jamovi project. jamovi, 2019. https://www.jamovi.org/ (accessed November 15,
2019).
UNEP, 2016. Global guidance for life cycle impact assessment indicators. https://www.
lifecycleinitiative.org/download/5746/, Accessed date: 9 March 2020.
USDA, 2019. FoodData central: tomatoes, red, ripe, raw, year round average. https://fdc.
nal.usda.gov/fdc-app.html#/food-details/170457/nutrients, Accessed date: 29 Octo-
ber 2019.
Vafaei, N.,Ribeiro, R.A., Camarinha-Matos, L.M.,2016. Normalizationtechniques for multi-
criteria decision making: analytical hierarchy process case study. In: Camarinha-
Matos, L.M., Falcão, A.J., Vafaei, N., Najdi, S. (Eds.), Technological Innovation for
Cyber-Physical Systems. Springer International Publishing, Cham, pp. 261–269.
Vermeulen PCM, van der Lans CJM. Combined heat and power (CHP) as a possible
method for reduction of the Co2 Footprint Of Organic Greenhouse Horticulture. In:
Dorais M, Bishop SD, Meijer RJM. Proceedings of the First International Conference
on Organic Greenhouse Horticulture; Bleiswijk, The Netherlands, October 11–14,
2010: International Society for Horticultural Science;International Conference on Or-
ganic Greenhouse Horticulture. Leuven: Secretariat of ISHS Int. Soc. for Horticultural
Science; 2011.
Wątróbski, J., Jankowski, J., Ziemba, P., Karczmarczyk, A., Zioło, M., 2019a. Generalised
framework for multi-criteria method selection. Omega 86, 107–124.
Wątróbski, J., Jankowski, J., Ziemba, P., Karczmarczyk, A., Zioło, M., 2019b. MCDA method
selection tool: An innovative tool for selecting MCDA method tailored to the decision
problem. http://mcda.it/.
Wikström, F., Williams, H., Verghese, K., Clune, S., 2014. The influence of packaging attri-
butes on consumer behaviour in food-packaging life cycle assessment studies - a
neglected topic. J. Clean. Prod. 73, 100–108.
Williams, H., Wikström, F., 2011. Environmental impact of packaging and food losses in a
life cycle perspective: a comparative analysis of five food items. J. Clean. Prod. 19 (1),
43–48.
Williams, H., Wikström, F., Otterbring, T., Löfgren, M., Gustafsson, A., 2012. Reasons for
household food waste withspecial attention to packaging. J. Clean. Prod. 24, 141–148.
de Wit, M.,Hoogzaad, J., Ramkumar, S., Friedl, H., Douma, A., 2018.The Circularity GapRe-
port: An Analysis of the Circular State of the Global Economy. https://bfc732f7-80e9-
4ba1-b429-7f76cf51627b.filesusr.com/ugd/ad6e59_
733a71635ad946bc9902dbdc52217018.pdf, Accessed date: 11 December 2019.
Wohner, B., Pauer, E., Heinrich, V., Tacker, M., 2019a. Packaging-related food losses and
waste: an overview of drivers and issues. Sustainability 11 (1).
Wohner, B., Schwarzinger, N., Gürlich, U., Heinrich, V., Tacker, M., 2019b. Technical
emptiability of dairy product packaging and its environmental implications in
Austria. PeerJ 7 (12), e7578.
Wood, R., Hertwich, E.G., 2013. Economic modelling and indicators in life cycle sustain-
ability assessment. Int. J. Life Cycle Assess. 18 (9), 1710–1721.
Yokokawa,N., Kikuchi-Uehara, E., Sugiyama, H.,Hirao, M., 2018. Framework for analyzing
the effects of packaging on food loss reduction by considering consumer behavior.
J. Clean. Prod. 174, 26–34.
Yonezawa, K., Richards, T.J., 2016. Competitive package size decisions. J. Retail. 92 (4),
445–469.
Zimek, M., Schober, A., Mair, C., Baumgartner, R.J., Stern, T., Füllsack, M., 2019. The third
wave of LCA as the “decade of consolidation”. Sustainability 11 (12), 3283.
12 B. Wohner et al. / Science of the Total Environment 738 (2020) 139846
