Content uploaded by Mattias Eriksson
Author content
All content in this area was uploaded by Mattias Eriksson on Nov 04, 2015
Content may be subject to copyright.
Content uploaded by Mattias Eriksson
Author content
All content in this area was uploaded by Mattias Eriksson on Nov 03, 2015
Content may be subject to copyright.
Supermarket food waste
Prevention and management with the focus on reduced
waste for reduced carbon footprint
Mattias Eriksson
Faculty of Natural Resources and Agricultural Sciences
Department of Energy and Technology
Uppsala
Doctoral Thesis
Swedish University of Agricultural Sciences
Uppsala 2015
Acta Universitatis agriculturae Sueciae
2015:119
ISSN 1652-6880
ISBN (print version) 978-91-576-8436-3
ISBN (electronic version) 978-91-576-8437-0
© 2015 Mattias Eriksson, Uppsala
Print: SLU Service/Repro, Uppsala 2015
Cover: Romanesco in a degraded state
(photo: Klaus Pichler)
Supermarket food waste - Prevention and management with the
focus on reduced waste for reduced carbon footprint
Abstract
Food waste occurs along the entire food supply chain and gives rise to great financial
losses and waste of natural resources. The retail stage of the supply chain contributes
significant masses of waste. Causes of this waste need to be identified before potential
waste reduction measures can be designed, tested and evaluated. Therefore this thesis
quantified retail food waste and evaluated selected prevention and valorisation
measures, in order to determine how the carbon footprint of food can be reduced by
decreasing food waste in supermarkets.
Food waste was quantified in six supermarkets in the Uppsala-Stockholm region of
Sweden. Data were recorded over five years between 2010 and 2014 by the retail
company in a daily waste recording procedure. In addition, suppliers contributed data
on deliveries and rejections. The main suppliers contributed data on wholesale pack
size and shelf-life, which allowed the relationship between these and their effect on
waste to be analysed. Life cycle assessment was used to investigate the carbon footprint
associated with production and distribution of food and managing the waste.
The wasted mass was dominated by fresh fruit and vegetables and rejection on
delivery was the main reason for this food being wasted. Expressed in terms of carbon
footprint rather than mass, the relative importance of meat waste increased and that of
fruit and vegetables decreased.
A reduction in storage temperature to prolong shelf-life proved to have the potential
to reduce waste in all supermarket departments studied. However, when the
temperature reduction was achieved by extended use of the current electricity mix, a net
lowering of carbon footprint was only found for the meat department. For food
products with a high carbon footprint, e.g. beef, there was much greater potential to
lower the carbon footprint by preventing waste through source reduction than by
upgrading the waste management option. If food waste cannot be prevented, donation
to charity and anaerobic digestion of the waste were found to have the greatest potential
to reduce the carbon footprint, depending on the substituted bread value and biogas
potential, respectively. This follows the EU waste hierarchy, although there are
variations from the trend of more favourable options at higher levels of the hierarchy.
Keywords: Food waste, Supermarket, Prevention, Valorisation, Life cycle assessment
Author’s address: Mattias Eriksson, SLU, Department of Energy and Technology,
P.O. Box 7032, 750 07 Uppsala, Sweden
E-mail: Mattias.Eriksson@slu.se
Dedication
To all the food waste geeks out there
We cannot solve our problems with the same
thinking we used when we created them.
Albert Einstein
Contents
List of Publications 7
Abbreviations 9
1 Introduction 10
1.1 The food waste problem 10
1.2 The role of supermarkets in the food supply chain 11
2 Objectives and structure of the thesis 13
2.1 Objectives 13
2.2 Structure of the thesis 13
2.3 Other publications by the author relating to the thesis 14
3 Background 16
3.1 Definitions of food waste in the literature 16
3.2 Waste and losses in the food supply chain 18
3.3 Carbon footprint of food production and waste handling 20
3.4 The waste hierarchy 21
3.5 Structuring waste reduction efforts 23
3.5.1 Quantities 23
3.5.2 Causes and risk factors 24
3.5.3 Measures 27
4 Material and Methods 30
4.1 Classification and definition of food waste 31
4.2 Collection and analysis of store data 33
4.2.1 Data collection for recorded waste and rejections 33
4.2.2 Data collection for unrecorded waste 34
4.2.3 Data collection for delivered and sold mass 35
4.2.4 Analysis of waste data 36
4.2.5 Identification of systematic causes and risk factors of waste 36
4.3 Carbon footprint of processes related to food waste 38
4.3.1 Carbon footprint associated with cradle to retail emissions 38
4.3.1 Carbon footprint associated with waste management options 39
4.4 Waste prevention and valorisation framework 40
5 Results 42
5.1 Quantities 42
5.1.1 Quantities of wasted perishable food 42
5.1.2 Mass balance of fresh fruit and vegetables 47
5.2 Risk factors for food waste and causes of discarding food 48
5.3 Measures 51
5.3.1 Prevention measures 51
5.3.2 Valorisation measures 54
5.3.3 Comparison of valorisation and prevention measures 55
6 Discussion 57
6.1 Quantities of food waste 57
6.1.1 Use of different units for quantification 57
6.1.2 Data quality and selection of study objects 58
6.1.1 Uncertainties in carbon footprint of food 59
6.1.2 Issues regarding data quality for fruit and vegetables 61
6.1.3 Comparison of indicator values of waste generation 63
6.2 Waste reduction measures 65
6.2.1 Perspectives on waste prevention and valorisation 65
6.2.2 Factors influencing the evaluation of waste reduction measures 67
6.3 Potential to increase sustainability by reducing food waste 68
7 Conclusions 71
8 Future research 73
References 75
Acknowledgements 83
Appendix I. Store department level results. 84
Appendix II. Food category level results. 85
Appendix III. Food product level results 88
Appendix IV. Article level results 93
7
List of Publications
This thesis is based on the work contained in the following papers, referred to
by Roman numerals in the text:
I Eriksson, M., Strid, I. & Hansson, P-A. (2012). Food losses in six Swedish
retail stores - wastage of fruit and vegetables in relation to quantities
delivered. Resources, Conservation and Recycling 68, 14-20.
II Eriksson, M., Strid, I. & Hansson, P.-A. (2014). Wastage of organic and
conventional meat and dairy products - a case study from Swedish retail.
Resources, Conservation and Recycling 83, 44-52.
III Scholz, K., Eriksson, M. & Strid, I. (2015). Carbon footprint of supermarket
food waste. Resources, Conservation and Recycling 94, 56-65.
IV Eriksson, M., Strid, I. & Hansson, P.-A. (2015). Food waste reduction in
supermarkets – net costs and benefits of reduced storage temperature.
(Submitted manuscript).
V Eriksson, M., Strid, I. & Hansson, P.-A. (2015). Carbon footprint of food
waste management options in the waste hierarchy - a Swedish case study.
Journal of Cleaner Production 93, 115-125.
Papers I-III and V are reproduced with the permission of the publishers.
8
The contribution of Mattias Eriksson to Papers I-V was as follows:
I Planned the paper in cooperation with the co-authors. Performed data
collection, observations in stores, physical measurements and analysis of
data. Interpreted the data and wrote the manuscript together with the co-
authors.
II Planned the paper in cooperation with the co-authors. Performed data
collection, calculations and analysis of data. Interpreted the data and
wrote the manuscript with input from the co-authors.
III Planned the paper together with the co-authors. Supervised the data
collection, calculations and analysis of data. Provided input to the writing
of the manuscript and was corresponding author.
IV Planned the study. Performed data collection, calculations and analysis of
data. Interpreted the data and wrote the manuscript with input from the
co-authors.
V Planned the study with input from the co-authors. Performed data
collection, calculations and analysis of data. Interpreted the data and
wrote the manuscript with input from the co-authors.
9
Abbreviations
CF
Carbon footprint
CO
2
e
Carbon dioxide equivalents
EAN
European Article Number
FAO
Food and Agricultural Organisation of the United Nations
FFV
Fresh fruit and vegetables
GWP
Global Warming Potential
LCA
Life Cycle Assessment
MFA
Material/mass flow analysis
MLR
Multiple linear regression
MOS
Minimum order size
PLU
Price look-up
SL
Shelf-life
T
Turnover
WFD
Waste Framework Directive
10
1 Introduction
Providing enough food for the world’s growing population is easy, but doing
this at an acceptable cost to the planet is more challenging (Nature, 2010). This
challenge requires changes in the way food is produced, stored, processed,
distributed and consumed. Godfrey et al. (2010) suggest five major strategies
to meet these challenges: Closing the yield gap; increasing production limits by
genetic modification; expanding aquaculture; dietary changes; and reducing
waste. These all involve utilising the full potential of the production system so
that more food can be consumed without increased resource demand at the
same rate. Reducing waste is unique in this context, since it focuses on food
that is already produced, but not consumed for various reasons. Since reduced
waste of edible food is also one of the least controversial ways to make the
food supply chain more productive, it has the potential to be used immediately
to decrease the competition for natural resources that could be saved for future
production to avoid a future food crisis (Nellemann et al., 2009).
1.1 The food waste problem
Waste, loss or spoilage of food is an efficiency issue that has attracted
increasing attention from the media, researchers, politicians, companies and the
general public in recent years. Although food waste seems like a simple
problem, the solution “to just stop throwing it away” is much more complex
than would appear at first glance. This is because food waste is not just a
problem, but also a solution to other problems, such as public health or
economic profit, which are often a higher priority. Food is also wasted for a
large number of reasons, which makes it difficult to find a ‘quick fix’ to reduce
food waste once and for all. In many countries the food waste in itself creates a
problem if it is dumped in landfill and generates methane. In other countries,
Sweden included, landfilling of organic waste is prohibited and surplus food is
considered a resource that can be used for biogas production or for feeding
11
people in need. It is therefore not the wasted food that should be the main
concern, but the wasteful behaviour that results in unnecessary food
production.
The complexity of the food waste issue also links it to the three parts of
sustainable development: economics, social issues and environmental impact.
This does not mean that reduced food waste automatically results in sustainable
development, but reducing unnecessary food waste has the potential to make
an important contribution and also has a high symbolic value. Food waste can
be related to waste of money (FAO, 2013) and natural resources (Steinfeldt et
al., 2006; Garnett, 2011), but also has moral implications in relation to food
security (Stuart, 2009; FAO, 2012). The political will to work on food waste
reduction can be seen as rational and positive, since there are few good
arguments for keeping on wasting food. This has resulted in several goals on
waste reduction among companies (Tesco, 2014), states (Rutten, 2013) and
international organisations (EC, 2011). As pointed out by Garnett (2011),
reducing food waste is not the only way to make the food supply chain more
sustainable, but it has the potential to save money too and is less controversial
than e.g. reducing meat consumption.
One of the problems closely associated with food waste is food security and
the moral implications of throwing away food while people in parts of the word
are starving (Stuart, 2009). However, just finishing off the food on one’s plate
will not make a starving person any happier, since the problem of starvation is
also connected to the global economy and how resources are distributed around
the world. Therefore a reduction in food waste in a supermarket in Sweden will
not necessarily lead to less starvation in the world, but may have an indirect
influence due to reduced demand for the finite resources needed for food
production.
1.2 The role of supermarkets in the food supply chain
The loss of food is a problem along the whole food supply chain but since
more value, in terms of both money and resources, is added for every step in
the food supply chain, waste represents more loss of value at the end of the
chain when more subprocesses have been in vain (Eriksson & Strid, 2013;
Strid et al., 2014). This means that the potential economic benefits of reducing
waste per unit mass are higher in later stages of the value chain (SEPA, 2012).
However for some products, especially those of animal origin, much of the life
cycle emissions are generated already at farm level (Röös, 2013) and food
waste reduction will therefore have the same high reducing effect along the
whole supply chain after farm level.
12
Supermarkets are located close to the end of the supply chain and also
collect large quantities of food in a limited number of physical locations.
Therefore these are potentially good targets for waste reduction measures, even
though supermarkets contribute a relatively small share of waste in comparison
with other stages in the food supply chain (Jensen et al., 2011a; FAO, 2011;
Göbel et al., 2012). Recent studies of food waste in supermarkets mostly focus
on describing the quantity of waste, problems causing it and how it could be
given to charity in order to avoid waste (Alexander & Smaje, 2008; Buzby et
al., 2009; 2011; Lee & Willis, 2010; Gustavsson & Stage, 2011; Stenmarck et
al., 2011; Lebersorger & Schneider, 2014). There is therefore a need to take
this problem one step further and investigate waste prevention and waste
valorisation measures, and the potential to reduce the environmental, social and
economic impacts related to food waste.
This thesis focuses on waste quantification in order to move further towards
finding potential ways of preventing food waste in supermarkets or, when
prevention is not possible, reducing the negative outcome regarding the carbon
footprint of handling food waste. Such knowledge could be used to reduce the
negative impact of the food supply chain and thereby contribute to sustainable
development for future generations.
13
2 Objectives and structure of the thesis
2.1 Objectives
The overall aim of this thesis was to provide new information on how to reduce
food waste and the carbon footprint associated with wasted food. Specific
objectives were to describe the quantity of wasted food in supermarkets in
terms of mass and carbon footprint, analyse some risk factors that can increase
waste and perform a theoretical evaluation of various waste valorisation and
prevention measures.
2.2 Structure of the thesis
In order to fulfil these objectives, the thesis was structured according to the
four steps of waste reduction shown in Figure 1. The first step is to quantify the
extent of the problem and potential hotspots. The quantities defined can then be
analysed to find causes and risk factors influencing waste generation. With this
information, efficient measures can be designed to reduce the risk factors.
When effective measures have been introduced, they can be evaluated in terms
of how much they save by reducing waste and how much they cost.
Papers I and II focus on the wasted mass in supermarkets, concentrating on
fruit and vegetables and organic meat, deli, cheese and dairy products. The
carbon footprint associated with production and distribution of the wasted food
is the main focus in Paper III.
Paper II also examines causes relating to turnover, shelf-life and minimum
order size. This relationship is further developed in Paper IV, where it is used
to design and theoretically evaluate a waste prevention measure of increasing
the shelf-life by reducing the storage temperature.
Several waste valorisation options are evaluated in Paper V, together with
theoretic measures regarding donation of surplus food and its use in animal
feed.
14
2.3 Other publications by the author relating to the thesis
During the work of this thesis, a number of ideas or problems in need of further
investigation were identified. It was not possible to explore all of these ideas in
Papers I-V and therefore a number of other papers have been written based on
the material and experience collected. These related publications are included
together with Papers I-V in Figure 1 and are listed with a short description in
Table 1. Only publications where the author of this thesis was co-author and
where the investigation centred on waste in the food supply chain are shown.
Figure 1. Schematic diagram of Papers I-V in this thesis and related publications. The vertical
levels illustrate different steps in the waste reduction process and the horizontal segments
illustrate different stages in the food supply chain.
15
Table 1. Brief summary of other publications related to the work in this thesis
Reference
Type of
publication
Short description
Eriksson & Strid (2011)
Technical report
Pre-study of Paper I quantifying in-store
waste of fruit and vegetables, cheese, dairy,
deli and meat during 2010.
Marklinder et al. (2012)
Technical report
The 2011 mass experiment of the Swedish
version of researchers’ night, where school
children were engaged to measure the
temperature in several places in domestic
refrigerators.
Marklinder & Eriksson (2012)
Conference paper
Marklinder & Eriksson (2015)
Research paper
Eriksson (2012)
Licentiate thesis
Summary of the findings of Papers I and II.
Strid & Eriksson (2013)
Conference paper
Evaluation of a pilot test where
supermarkets froze down meat cuts and
sold them to a restaurant.
Eriksson & Strid (2013)
Technical report
Describing and calculating the potential
savings and cost of six food waste
reduction measures in supermarkets.
Strid et al. (2014)
Technical report
Investigating losses in Swedish production
and distribution of iceberg lettuce.
Strid & Eriksson (2014)
Conference paper
16
3 Background
Food is wasted in all stages of the food supply chain, but since the food
distribution system is large and complex, there are significant variations in
quantities over time, between products and between different types of
businesses. Due to the complexity of the food supply chain there is a need for
many large studies to fully cover the quantities of waste generated and the
underlying causes, and ultimately what could be done to reduce the negative
consequences of food waste. This chapter presents some existing knowledge
about food waste in general, but with the emphasis on food waste in
supermarkets.
3.1 Definitions of food waste in the literature
In order to quantify food waste, there is first a need to define what the
quantification should include. Since food consists of a large and diverse group
of products, it is complicated to find an easy definition that fits all purposes.
Moreover, waste and the process that turns food into waste include many
situations and perspectives. Therefore the literature is full of expressions such
as “food loss” (e.g. FAO, 2011; Strid & Eriksson, 2014), “food waste” (e.g.
DEFRA, 2010), “post-harvest loss” (e.g. Hodges et al., 2011), “food and drink
waste” (e.g. Griffin et al., 2009; Lee & Willis, 2010) and “spoilage” (e.g.
Lundquist et al., 2008). According to Östergren et al. (2014), the list may even
be much longer. Some of these expressions are overlapping and some are used
to define different type of waste.
One problem with developing the definition of food waste, as explained by
Schneider (2013b), is the commonly used EU definition of food (EC, 2002).
17
This legal definition
1
excludes plants prior to harvesting. Therefore plants
which are not harvested due e.g. to low market price are not counted as food
waste (Schneider, 2013). This creates a problem, since the food waste issue
does not necessarily start at harvest. Therefore Östergren et al. (2014) propose
a definition that includes products prior to harvest, which is a clear distinction
from many other studies. Their definition of food waste
2
uses a definition of
the food supply chain
3
, which includes products ready for harvest or slaughter,
not just products defined as food by EC (2002). Since the definition by
Östergren et al. (2014) also includes inedible parts of food products, it covers
as subcategories other commonly used categorisations such as “avoidable”,
“possibly avoidable” and “unavoidable” food waste (EC, 2010; WRAP, 2011).
The definition used is of course a matter of opinion and as long as it is
clearly stated in publications, it does not create problems. Problems appear,
however, when quantities of food waste based on different definitions are
merged together and used as if defined similarly. An example of this is the
Institution of Mechanical Engineers (2013) statement that 30-50% (or 1.2-2
billion metric tonnes (tons)) of all food produced never reaches a human
stomach, based on FAO (2011) and Lundquist et al. (2008). The problem with
this is that Lundquist et al. (2008) compare the basic production with what is
1
REGULATION (EC) No 178/2002, Article 2, Definition of ‘food’:
For the purposes of this Regulation, ‘food’ (or ‘foodstuff’) means any substance or product,
whether processed, partially processed or unprocessed, intended to be, or reasonably expected to
be, ingested by humans.
‘Food’ includes drink, chewing gum and any substance, including water, intentionally
incorporated into the food during its manufacture, preparation or treatment. It includes water after
the point of compliance as defined in Article 6 of Directive 98/83/EC and without prejudice to the
requirements of Directives 80/778/EEC and 98/83/EC.
‘Food’ shall not include: (a) feed; (b) live animals unless they are prepared for placing on the
market for human consumption; (c) plants prior to harvesting; (d) medicinal products within the
meaning of Council Directives 65/65/EEC (1) and 92/73/EEC (2); (e) cosmetics within the
meaning of Council Directive 76/768/EEC (3); (f) tobacco and tobacco products within the
meaning of Council Directive 89/622/EEC (4); (g) narcotic or psychotropic substances within the
meaning of the United Nations Single Convention on Narcotic Drugs, 1961, and the United
Nations Convention on Psychotropic Substances, 1971; (h) residues and contaminants.
2
Food waste is any food, and inedible parts of food, removed from the food supply chain to be
recovered or disposed of (including composted, crops ploughed in/not harvested, anaerobic
digestion, bio-energy production, co-generation, incineration, disposal to sewer, landfill or
discarded to sea).
3
The food supply chain is the connected series of activities used to produce, process, distribute
and consume food. The food supply chain starts when the raw materials for food are ready to
enter the economic and technical system for food production or home-grown consumption. This is
a key distinction, in that any products ready for harvest or slaughter being removed are within
scope, not just those harvested and subsequently not used. It ends when the food is consumed or
‘removed’ from the food supply chain.
18
eaten to estimate the waste, which means that animal feed is included in waste.
FAO (2011), on the other hand, defines food waste and losses as food that was
originally meant for human consumption but which unfortunately leaves the
human food chain (even if directed to a non-food use). Inclusion of animal feed
as a food waste or not has a large impact and could explain the difference
between 30% and 50% waste. Stating these values as a range clearly gives the
reader a false impression of the size of the waste problem, since the waste can
actually be both 30% and 50% at the same time.
3.2 Waste and losses in the food supply chain
Several studies in recent years have attempted to estimate parts of the global
food waste and its consequences. According to FAO (2011), approximately
one-third of the food produced in the world is wasted, corresponding to 1.3
billion tons of food waste every year. To put this figure into context, FAO
(2013) also estimates that this food waste gives rise to greenhouse gases
corresponding to 3.3 billion tons of carbon dioxide
equivalents (CO
2
e) every
year, costs around $750 billion annually and guzzles a volume of water
equivalent to the annual flow of Russia's Volga River. These figures are of
course rough estimates associated with both large variations and insecure data,
but clearly much of the food produced in the world is not consumed as
intended.
There seems to be a trend in the waste pattern of the whole food supply
chain for much of the waste to occur during primary production and in the
consumer stage (FAO, 2011; Jensen et al., 2011a; Göbel et al., 2012). The
stages in between, including processing, wholesale and retail, contribute
smaller amounts in this perspective, which could be the reason why consumers
are often the target of waste reduction campaigns and other efforts to reduce
food waste (NFA, 2015; WRAP, 2015). However, even if the waste occurring
in the retail stage of the supply chain is less than in some other stages, the
amounts involved are still enormous, e.g. approximately 70 000 tons per year
in Sweden (SEPA, 2013) and 4.4 million tons per year in the EU-27 (EC,
2010).
The contribution of the retail sector to waste in the Swedish food supply
chain (excluding agriculture) is estimated to be 39 000 tons per year,
corresponding to 3.8% (Jensen et al., 2011a). However, that estimate is based
only on the organic waste fraction and therefore Stare et al. (2013) investigated
the mixed waste fraction and upgraded the amount to 67 000 tons per year,
corresponding to 6.1% of the whole food supply chain (excluding agriculture).
The values presented in Figure 2 are based on data from Jensen et al. (2011a)
19
and Stare et al. (2013), updated by SEPA (2013) to represent the year 2012.
These figures, which can be considered the official Swedish food waste
statistics, show that 70 000 tons of food per year are wasted in Swedish
supermarkets. Göbel et al. (2012) estimated that the retail stage of the German
food supply chain contributes 3% of its food waste. This seems low in
comparison with the Swedish estimate of 6.1% (Stare et al., 2013), but Göbel
et al. (2012) include agriculture and if food waste from Swedish primary
production were to be included, it is likely that the Swedish value would be at a
similar level.
The retail sector of the food supply chain is not the largest contributor of
food waste, but the amounts are still high and the share of unnecessary waste is
also high (Figure 2), which makes it an important issue. Other aspects are that
food waste becomes concentrated in a limited number of physical locations,
making food rescue measures feasible. Supermarkets also represent an
important link between producers and consumers, with potential influence over
large parts of the food supply chain. This makes it possible for retailers to
communicate with consumers in order to increase their environmental
awareness and also to choose suppliers and producers that fulfil their corporate
responsibility. Retailers are particularly important for the Swedish food supply
chain, since the market is extremely concentrated and is completely dominated
by just a few large companies (Eriksson, 2012). For example, the market share
Figure 2. Estimated volumes of food waste generated in Sweden in 2012 (SEPA, 2013).
20
of the five largest food retailing companies in Sweden amounted to 94.7% in
2002, which was the highest in Europe, where the average level was 69.2%
(Vander Stichele et al., 2006). These five companies also own or control large
parts of the distribution chain and, via private brands, some of the production.
3.3 Carbon footprint of food production and waste handling
Life cycle assessment (LCA) is a method for analysing the environmental
impact of a product or service by analysing different aspects such as land use,
water use, eutrophication, climate impact and acidification. Since many
different aspects are included, a substantial review of environmental impact
can be assessed. The problem is of course that it requires large reasearch
resources to make a full LCA with many impact categories for a variety of
products or services, with many geographical regions and production systems
that need to be considered. Carbon footprint (CF) assessment provides a
limited perspective, since only the global warming potential (GWP) is
included. However, a less extensive assessment can instead allow analysis of a
larger number of scenarios or a more extensive product range, using the same
research resources.
A large number of studies on the GWP or CF of food products have been
performed (Roy et al., 2009; Röös, 2012). As pointed out by Röös (2013), the
results vary widely between different food products, but also for a particular
food product depending on factors such as production system and
methodological choices in the assessment. However, one pattern which has
emerged is that products of animal origin generally have a considerably larger
CF than products of vegetable origin (EC, 2006), and that this footprint are
generated already at farm level. Meat, particularly lamb and beef, has an
exceptionally high CF, followed by cheese, due to the contribution of methane
(CH
4
) from enteric fermentation in ruminants. Meat from monogastric animals,
such as pigs and poultry, has lower CF values than products from ruminants,
but still higher than most foods of plant origin, due to the large amount of feed
needed in livestock production and emissions from manure handling. Some
fruit and vegetables can have a considerably high CF if produced in heated
greenhouses, transported by air or produced in low-yielding systems (Stoessel
et al., 2012). For many food products, nitrous oxide (N
2
O) emissions from soil
also contribute significantly to the CF.
Losses in the food supply chain are seldom included in the CF of food
products, possibly due to lack of data. If the wasted part were to be included,
the CF of some food products could increase significantly, since surplus
production is needed to cover both the fraction consumed and the fraction
21
wasted. If food waste is managed properly, it could be used as a byproduct that
can replace other virgin materials and thereby, to some extent, reduce the CF.
However, according to Hanssen (2010), producing biogas from food waste
only saves approximately 10% of the emissions generated during the
production of surplus food, so the recovery of food waste can be considered a
small part of the life cycle of food.
Even though waste management only can recover a small fraction of the
resources invested in food production, it is still important to consider waste
management due to the large quantity of waste generated. According to many
review studies (e.g. Bernstad & la Cour Jansen, 2012; Laurent et al., 2013a;
2013b), the CF of food waste could be reduced by shifting from less favoured
options in the EU waste hierarchy (EC, 2008) to higher priority options.
According to Laurent et al. (2013a), the most common order in the waste
hierarchy is landfilling as least favourable, followed by composting, thermal
treatment and anaerobic digestion as the most favourable. However, there is
great variation due to differences in local contexts, but also the use of different
methodology to assess the different waste management systems (Bernstad & la
Cour Jansen, 2012; Laurent et al., 2013a; 2013b).
3.4 The waste hierarchy
The EU waste hierarchy is set in the European Waste Framework Directive
(WFD), which ranks waste prevention and management options in order of
priority (EC, 2008). The WFD also obliges member states to encourage options
that deliver the best overall environmental outcome from a life cycle
perspective, even when this differs from the waste hierarchy. However, since
the environmental outcome is not defined in the WFD, this goal can be
achieved in many ways. Addressing GWP is one way to do so, but GWP alone
offers only a limited perspective on the overall environmental outcome,
although to some extent it can act as an indicator of other environmental
impact categories (Röös et al., 2013).
Early versions of the waste hierarchy have been part of European policy
since the 1970s (EC, 1975). While it has been developed and amended (EC,
2008), it still provides only very general guidelines for all waste, including the
priority order from prevention, re-use and preparation for re-use, recycling,
recovery and, last and least favourable, dumping in landfill. Guidelines relating
specifically to food waste have therefore been devised. Examples of such
systems are the Moerman ladder in the Netherlands (Dutch Ministry of
Economic Affairs, Agriculture and Innovation, 2014), the Food Recovery
Hierarchy in the United States (USEPA, 2015) and the Food Waste Pyramid in
22
the United Kingdom (Feeding the 5000, 2014). All these systems prioritise
prevention, since all other waste management options include downcycling and
loss of the intended product. Despite the order of priority in the waste
hierarchy, only a few studies measure waste prevention in the context of waste
management (Laurent et al., 2013a). This omission may be due to the
methodical difficulties in measuring something that is not there (Zorpas &
Lasaridi, 2013) or, as discussed by van Ewijk & Stegemann (2015), to
prevention being fundamentally different from waste management.
The US Food Recovery Hierarchy, which is shown in Figure 3 (USEPA,
2015), agrees with the general principles of the EU waste hierarchy (EC,
2008), but has one important difference in that it separates the prevention stage
into what can be seen as two sublevels. The more preferred sublevel is source
reduction and the less preferred sublevel is feeding hungry people. This is
important, since it implies that even though the food is eaten in the latter
option, which corresponds to its intentional use, it is better to be proactive and
reduce food production.
Figure 3. The Food Recovery Hierarchy developed by the USEPA (2015).
23
Feeding hungry people is also limited by the fact that food waste can only
be donated to charity if it is surplus food still fit for human consumption
(Papargyropoulou et al., 2014). Since the food hygiene or biosecurity
requirements increase at higher levels in the waste hierarchy, there is a
decreasing likelihood that the whole waste flow will be suitable for the same
type of waste management if using a more preferred method. This creates a
need for more complex systems where part of a food waste flow is developed
and used for higher priority waste treatments, while the rest is treated with a
lower priority, more general method (Vandermeersch et al., 2014).
3.5 Structuring waste reduction efforts
In organisations and companies, waste reduction is often sought by copying the
best practice within the organisation or by taking inspiration from other
successful examples of waste reduction measures (EC, 2010; Lagerberg
Fogelberg et al., 2011). Whether the suggested measures actually reduce the
waste and by how much are seldom reported, and thus it is difficult to compare
different measures and decide on the most efficient methods to reduce waste.
Therefore, in this thesis a more analytical approach was adopted, based on the
Deming cycle (also known as the plan-do-check-act methodology) used for
environmental management systems in order to reduce waste (ISO, 2010). This
strategy was suggested by Eriksson (2012) and involves:
1. Quantification of waste.
2. Analysis of causes.
3. Introduction of measures.
4. Evaluation of measures.
The steps to reducing waste involve describing the problem and the
underlying reasons for risky behaviour, testing solutions and then evaluating
how well the solutions actually reduce the problem and how much they cost.
3.5.1 Quantities
Retail food waste has been quantified in a few previous studies (Table 2). In all
these studies, different system boundaries, methods and units have been used.
In addition, different products have been studied, making comparisons
difficult, although the results from the studies do not vary widely. The results
indicate that retail food waste for different product groups is often in the range
0-10%. Many previous studies have focused on fresh fruit and vegetables
24
(FFV), which often give high percentage waste, e.g. 10% for the European
retail distribution sector according to FAO (2011).
No previous publication states the percentage of waste originating from the
retail sector in Sweden. However, if the wasted 70 000 tons per year reported
by SEPA (2013) are divided by the 3.5 million tons per year delivered to
Swedish supermarkets, approximated from Jensen et al. (2011b), these
supermarkets waste approximately 2% of the mass delivered. This is well in
line with the 1-2% waste reported for Finnish supermarkets (Katajajuuri et al.,
2014).
Table 2. Brief review of studies in the literature quantifying food waste in supermarkets
Reference
Country
Data collection
method
Reference base
Product group
Relative
waste (%)
Katajajuuri et al.
(2014)
Finland
Interviews
Not specified
Retail sector
1-2
Göbel et al.
(2012)
Germany
Analysis of
national statistics
Delivered mass
Retail sector
1
Buzby et al.
(2009)
USA
Supplier records
Supplier
shipment data
Fruit
Vegetables
8.4 - 10.7
8.4 - 10.3
Buzby & Hyman
(2012)
USA
Analysis of
national statistics
Food supply
value
FFV
9
Beretta et al.
(2013)
Switzerland
Estimate from
store records
Volumes of
sales
FFV
8 – 9
Fehr et al. (2002)
Brazil
Quantification at
retailer
Delivered mass
FFV
8.8
Stensgård &
Hanssen (2014)
Norway
Store records
Sales value
Fruit
Vegetables
4.5
4.3
Lebersorger &
Schneider (2014)
Austria
Store records
Sales in cost
price
FFV
4.3
Mattsson &
Williams (2015)
Sweden
Store records
Sold mass
FFV (only in-
store waste)
1.9
Buzby & Hyman
(2012)
USA
Analysis of
national statistics
Food supply
Value
Dairy products
9
Lebersorger &
Schneider (2014)
Austria
Store records
Sales in cost
price
Dairy products
1.3
Stensgård &
Hanssen (2014)
Norway
Store records
Sales value
Milk products
Cheese
0.8
0.9
3.5.2 Causes and risk factors
Food can be wasted for a large variety of reasons, which makes the food waste
issue difficult to solve with one single solution. Common reasons for food
25
being discarded in supermarkets are expired shelf-life or visual defects that
make food unsellable (at least at full price). However, as pointed out by
Lindbom et al. (2014), it is important to identify not just the reason for food
being discarded but also the underlying root cause of the problem. However,
such identification is problematic, since there are so many potential root causes
of e.g. expired shelf-life, such as too short shelf-life, too large inflow of
products, unexpected lack of demand, or a combination of all of these. Since it
is very difficult to identify a single root cause, risk factors are used here since
they better capture the multiplying effect when several risk factors are present
and include factors not necessarily leading to food waste, but just increasing
the risk of waste. Possible risk factors can be low demand, short shelf-life,
unsuitable packaging or storage conditions and inappropriate handling by staff
and customers.
In an extreme perspective, an inflow of food that is unbalanced with regard
to the outflow required can even be assumed to be the only root cause of food
waste. If so, all problems that prevent a supermarket from selling the food are
risk factors. These risk factors can also have an effect on the inflow, since the
supermarket will try to order just the right amount of all products, but anything
that creates variation will make this forecast more difficult. Thus to summarise,
if the forecast is just right there will be no waste and no empty shelves, but
everything that introduces variation will make forecasting more difficult and
increase the risk of food waste (or empty shelves).
There are several activities and problems introducing variation. One is
increased product variety (Lindbom et al., 2014), since having more different
types of products decreases turnover for each and makes forecasting more
difficult. On the other hand, providing a large variety of products also means
freedom for customers, which supermarkets might use as a competitive
advantage to differentiate them from their competitors. Since larger variety
might thus be expected to increase profits, it might be something that the
retailers are unwilling to alter, and waste is simply a part of the price they have
to pay for the larger range of products sold.
Promotions have a similar effect on food waste since they temporarily shift
the turnover of products and make forecasting more difficult. According to
Hernant (2012), some promotions prompt the customer to buy the promoted
product, but to reject other similar products as a consequence. Since
forecasting of sales is more difficult when there are many aspects to consider,
temporary shifts in sales can be difficult for retailers to predict accurately. This
leads to a larger than necessary stock of not promoted products and, since the
store must not run out of the promoted product, also a surplus of the promoted
product. The result of the campaign is increased waste of the promoted product
26
and also increased waste of other similar products. Added to the cost of the
waste is the lack of profit that arises when the store sells products at a lower
margin than usual. Thus promotions can really seem a waste of effort (Hernant,
2012), but they are unlikely to disappear since they are there to attract
customers and thereby increase overall profits. Promotions can therefore be
viewed as a marketing cost and waste as simply part of that cost.
In many cases the food waste does not appear in the same organisation that
caused it. If customers decide to stop buying a certain product, this product is
likely to end up as food waste if the supplier cannot stop its production fast
enough or find an alternative market. If this change in purchasing behaviour is
made by a single customer it might not affect the food logistics system at all,
but when many customers unexpectedly change their behaviour the food
supply chain simply cannot react fast enough to prevent overproduction and
eventually food waste. A fast reaction from a customer group might also cause
a chain reaction along the value chain that increases the effect and, in the end,
creates large amounts of food waste in primary production. According to
Taylor (2006), there are a number of actions in the supermarket that can lead to
a “bullwhip effect”, where the amplitude of the customer reaction increases
from retail to wholesale, from wholesale to industry and from industry to
primary production and everyone along the chain increases/decreases
production and increases/decreases stock in order to compensate for the
customer reaction. Increased communication along the logistics chain so that
primary producers get their signals directly from the end customers could be
one way to deal with this problem. Another way to decrease the risk of a
bullwhip effect could be by reducing the activities that increase variation.
According to Taylor (2006), these activities include promotions, large numbers
of products and/or actors in the logistics chain, and ordering and production in
large batches with large stocks. Therefore the same risk factors for food waste
can be problematic both within supermarkets and in other parts of the food
supply chain.
Most types of waste and losses are unintentional, but since several risk
factors are accepted as a normal part of any activity, waste must also be
accepted as something natural. A common reason for accepting the presence of
risk factors is that they are too expensive or too difficult to prevent. There can
also be a conflict of interest between waste reduction and increased profit or
public health, with waste reduction being likely to be a lower priority. To put
this simply, there are a large number of problems causing food waste that are
not interesting to solve because the potential benefits are believed to be less
than the cost of change. On the other hand, there are also many problems that
could easily be economically justified and therefore should be dealt with in
27
order to reduce food waste (Eriksson & Strid, 2013). The problem is knowing
which problems have low required management intensity (Garrone et al.,
2014), meaning that they are cheap and/or easy to solve. With this knowledge,
a countermeasure to reduce risk factors can be designed so the potential
savings can be compared with the expected cost of the intervention.
3.5.3 Measures
In order to reduce food waste in supermarkets, there is a need for measures that
solve the basic problems which cause waste. Waste quantification and cause
identification are often performed in order to design measures. These can be
seen as necessary pre-studies in order to identify where to target a measure, but
also to select the measures with the largest potential for reduction and/or the
lowest cost.
Food waste reduction measures can be categorised in several different
ways, but the main distinction is between prevention and valorisation
measures. Prevention measures aim to reduce the production of food, while
valorisation measures aim to create value from the waste occurring and thereby
reduce the negative effect of the waste. Donation to charity can be considered a
prevention measure, since the food is eaten by humans, but also a valorisation
measure, since it handles the surplus food rather than reducing the production
of food. Valorisation in this case can be considered in strictly monetary terms,
as done by Eriksson & Strid (2013), who only considered measures that use the
food for human consumption. Value in this case can have a wider meaning, i.e.
including any byproduct that reduces the negative effects of the waste
(Vandermeersch et al., 2014), but it can also just apply to food (and uneatable
parts of food) sent to animal feed, bio-material processing or other industrial
uses (Östergren et al., 2014). In their wider meaning, valorisation measures can
include any waste management option that recovers nutrients, energy or
byproducts from the food waste. It can also include waste management options
that give rise to less emissions or less general problems then the worst option,
e.g. landfill or even illegal dumping.
Most previous studies on waste management methods for food waste, or
organic waste including food waste, describe and sometimes compare landfill,
incineration, composting and anaerobic digestion (Bernstad & la Cour Jansen,
2012; Laurent et al., 2013a; 2013b). However, all these options occur within
the less prioritised part of the waste hierarchy defined by the European Waste
Framework Directive (EC, 2008). Some studies also include animal feed in the
comparison (e.g. Lee et al., 2007; Menikpura et al., 2013; Vandermeersch et
al., 2014), but none includes comparisons with the highest levels in the waste
hierarchy, such as donation and prevention. However, some studies describe
28
the environmental benefits of preventing food waste. For example, Gentil et al.
(2011) concluded that there are significant benefits of reducing food waste,
especially wasted meat, by 20% in a food waste stream. However, those
authors do not specify how this reduction should be achieved, or the cost of
doing so. Williams & Wikström (2011) & Williams et al. (2008) investigated
whether waste reduction can justify the increased use of packaging material
and found that it could do so for resource-consuming products such as cheese
and beef. However, those studies did not specify how large the potential
reduction could be if the packaging was redesigned. Another prevention study,
by Salhofer et al. (2008), regarded prevention as being equal to donation, but
did not quantify the actual potential in this measure. Moreover, Schneider
(2013a) valued donated food by its emissions during production, instead of the
produce that could be replaced. The lack of studies quantifying higher levels of
the waste hierarchy with a method comparable to the lower levels makes it
difficult to evaluate the actual environmental benefits of donation and
prevention in relation to other waste management options. Without such an
extended analysis, the life cycle perspective described in the WFD will not
actually be considered when selecting waste management options.
Among the large number of publications reviewed by Laurent et al. (2013a;
2013b), a pattern emerged in studies comparing different waste management
alternatives. The least favourable option was landfill, followed by composting
and thermal treatment, and the most favourable was anaerobic digestion.
However, not all studies fitted this pattern. Therefore Laurent et al. (2013a)
concluded that local infrastructure is essential for the outcome, making it more
difficult to generalise results.
Despite the order of priority in the waste hierarchy, only a few studies have
measured waste prevention in the context of waste management (Laurent et al.,
2013a). This omission may be due to the methodical difficulties of measuring
something that is not there (Zorpas & Lasaridi, 2013) or, as discussed by van
Ewijk & Stegemann (2015), to prevention being fundamentally different from
waste management. One of the differences that make it fundamentally different
is that waste management options are carried out by professions handling waste
management facilities, such as a municipal department, but prevention
measures can only be handled by staff in the supermarket or by logistic
departments in retail and wholesale companies. This means that supermarket
staff have little influence over what happens with the food waste after it leaves
the supermarket and that waste management professionals have little influence
over what happens with the food before it becomes waste.
Prevention of food waste relates more to resource management than to
waste management and therefore it is important to achieve source reduction,
29
i.e. reduced production, and not just prevent the food entering the supermarket.
However, there is no guarantee that the waste will not just move to an earlier
stage in the food supply chain and sub-optimisations like this reduce the effect
of the prevention measure. From an environmental perspective, it is not a
solution to move the waste as a way to prevent it occurring, even though when
waste occurs earlier in the food supply chain some sub-processes such as
transportation, storage and packaging might still be avoided (Strid & Eriksson,
2014; Strid et al., 2014). From an economic perspective, it might be enough to
reduce the inflow of food into the supermarket, although the food will then be
wasted at the supplier or producer, as long as the supermarket does not have to
pay. Moreover, the producer may increase the price of the food supplied in
order to cover the waste cost and if so, the supermarket will have to pay for the
waste anyway.
Swedish supermarkets are likely to use the local infrastructure available for
waste management, which means that if they do not prevent food waste or
donate it to charity, they send it to incineration, composting or anaerobic
digestion. Since it has been illegal to dump organic matter in landfill in
Sweden since 2005 (Ministry of the Environment and Energy, 2001), it is very
unlikely that any of the Swedish supermarket food waste is disposed of in this
way. According to Jensen (2011a), 22% of the food wasted in Swedish
supermarkets is managed with biological treatment, while the rest can be
assumed to be incinerated for production of district heating.
30
4 Material and Methods
The work presented in this thesis is based on case studies performed in the
context of six supermarkets located in Stockholm and Uppsala in Sweden.
Paper I used the data to quantify wasted fruit and vegetables and Paper II
quantified waste of organic food from the cheese, dairy, deli and meat
departments and analysed causes of this waste. Through an extended literature
review, Paper III added the perspective of CF associated with the wasted
quantities. Paper IV combined the causes analysis in Paper II and the CF
analysis of wasted food from Paper III with a literature review to examine
shelf-life extension potential and energy consumption at reduced storage
temperature. To extend this perspective, Paper V investigated different waste
management options that could be used for the fractions of the food waste that
cannot be prevented.
The six supermarkets investigated are owned, and were selected for the
study, by the head office of Willy:s, which is a major actor on the Swedish low
price retail market. The stores were selected within a specified region close to
the university performing the research and to provide a representative view of
the whole retail chain with regard to factors such as turnover, percentage waste
and profit. Within these supermarkets, the fresh fruit and vegetables, dairy,
cheese, meat and deli departments were selected for in-depth study, in
consultation with the retail company, due to their large contribution to food
waste and the expected high environmental impact of this waste. The bread
department also makes a large waste contribution, but this is managed
separately by the suppliers and was therefore not included in the quantification
studies. Wasted bread is considered in Paper V, but using only assumptions
regarding the wasted mass.
The material and methods used for data collection are described in detail in
Åhnberg & Strid (2010), Eriksson & Strid (2011; 2013) and Eriksson (2012).
In the study by Eriksson (2012), material flow analysis (MFA) was used as a
31
method to investigate the incoming and outgoing flows of food within a group
of supermarkets (Brunner & Rechberger, 2005). Variations on the MFA
approach (including life cycle inventories) were used in Papers I-V in order to
establish the mass of each type of food leaving the supermarkets either as any
type of waste or as sold food. In Paper I the use of MFA was most extensive,
since a full mass balance was performed for the waste in the FFV department.
More extensive LCA was performed in Paper III regarding the CF from cradle
to retail, in Paper IV regarding the cost and benefits of reducing waste through
reduced storage temperature and in Paper V where different waste management
options for the food waste were assessed (Figure 4).
4.1 Classification and definition of food waste
The definition of food waste used in this thesis is that proposed by Östergren et
al. (2014): “Food waste is any food, and inedible parts of food, removed from
the food supply chain to be recovered or disposed of (including composted,
crops ploughed in/not harvested, anaerobic digestion, bio-energy production,
co-generation, incineration, disposal to sewer, landfill or discarded to sea)”.
Since supermarkets sell food products that have not yet been separated into
their edible and inedible parts, the waste consists of a mix of avoidable,
possibly avoidable and unavoidable food waste (EC, 2010; WRAP, 2011).
Since e.g. a banana is sold in the supermarket with the peel on, it is also wasted
with peel and a categorisation like that suggested by EC (2010) and WRAP
(2011) is only applicable at a stage in the FSC where the banana is consumed.
Food waste from supermarkets can be divided into several categories
depending on system boundaries (Östergren et al., 2014) but, as described in
Figure 4. The food supply chain with the system perspective from each of the papers (III-V)
using LCA as a method. Paper IV also relies heavily on results from Paper III that are not
included in the diagram.
32
Paper I, food waste (or retail food waste) was defined in this thesis as products
discarded in the supermarkets studied, irrespective of whether they belonged to
the supplier or the supermarket. This meant that losses of mass due to theft or
evaporation were not considered food waste and are therefore included in a
separate category (missing quantities) in Figure 5.
Pre-store waste consisted of items rejected by the supermarket at delivery
due to non-compliance with quality requirements. This waste belongs to the
supplier in accounting terms, since it is rejected by the supermarket, but is
usually discarded at the supermarket. Pre-store waste is defined through
documented complaints to suppliers, which according to the rules must be done
within 24 hours of delivery. This waste is on rare occasions sent back to the
supplier for control, but is still wasted.
Recorded in-store waste was defined as food waste occurring after purchase
from the supplier. This waste is sorted out and discarded by supermarkets when
there is little or no possibility of selling the products. This could be due to
exceeded best-before date or product deterioration for unpackaged fresh fruit
and vegetables.
Unrecorded in-store waste consisted of food waste that was discarded but
not recorded. This means that it had the potential to be either pre-store waste or
recorded in-store waste if recorded in any of these categories. Unrecorded in-
store waste originated from two sources: underestimated mass when recording
unpackaged waste; and unrecorded of wasted items. The latter can occur in
Figure 5. Flow chart with an overview of the waste categorisation used and the physical flow of
food marked with arrows.
33
error or as a deliberate act, e.g. it is not cost-effective to record small amounts
of waste.
The three food waste categories all contributed to fill up the waste
containers of the supermarkets studied, but there was also a category of
missing quantities. This was due to loss of mass between outgoing and ingoing
flows, the two main reasons for which are believed to be theft and mass loss
due to evaporation. Stolen food is considered not to be an environmental
problem, since it is believed to be eaten. Evaporation losses are also not
primarily food waste, since the food items are left, but with a higher dry matter
content and smaller mass. However, when visible, this might act as a secondary
effect, leading to losses of food in one of the waste categories.
4.2 Collection and analysis of store data
4.2.1 Data collection for recorded waste and rejections
Food that was sorted out and discarded was recorded as part of a daily routine
normally performed by the stores and established years before this
investigation (Åhnberg & Strid, 2010). This routine was not introduced by the
author, only used in order to collect data. The routine starts with an inventory
in the morning where products considered unsellable are sorted out. Products
are considered unsellable if they have passed their best-before or use-by date.
Since some FFV are sold without a date label, the sorting of these products is
based on visual appearance and the unsellable limit is defined by each staff
member based on whether they would buy the product themselves (Willy:s,
2010).
Products from the deli, meat, dairy and cheese departments are recorded
directly with a mobile scanner connected to the company database and then
discarded. Waste due to poor quality at delivery is economically reimbursed by
the supplier if the member of staff presses a one-digit code on the mobile
scanner to indicate whether the waste is charged to the supermarket, the main
supplier (DAGAB) or other suppliers.
Discarded fruit and vegetables are placed in the storage room until the end
of the shift, when the staff record the waste. Recording is often done by the
team leader or other experienced member of staff using the mobile scanner for
waste at the supermarket’s expense. Waste due to rejections is registered first
on paper and then transferred to the website of the logistics company (SABA)
delivering all fruit and vegetables to the supermarkets. Since all products are
owned by Axfood when handled by SABA, the data on rejections are then
transferred to a database within Axfood (Figure 6).
34
The records on wasted products are stored in the retail company database.
Data on rejections are stored by DAGAB and Axfood and were provided in the
form of weekly reports to the author.
4.2.2 Data collection for unrecorded waste
From observations and interviews with the staff, it became clear that the
recording of wasted fruit and vegetables is not completely accurate. To
quantify the missing part of the waste, a control measurement of the waste was
performed. This method was closely related to the data collection methods
used for household waste surveillance (Ventour, 2008; Andersson, 2012), with
the distinction that the waste was not allowed to enter the waste container
before recording. This manual recording of otherwise unrecorded waste was
the only data collection process that could not harvest data from an existing
system within the supermarkets.
The data collection was performed after the staff had recorded the waste,
when instead of dumping the waste they left it together with printouts of the
record. All fruit and vegetables in the pile were then measured on a set of
scales to check the masses, which were compared with the masses recorded
earlier.
During the first measurement of unrecorded waste, which lasted for two
weeks, only differences between recorded and measured mass were quantified.
It then became clear that some items were discarded without being recorded at
all, and that some items were recorded without being found in the pile of
waste, possibly discarded directly by mistake. Therefore a second
Figure 6. Flow chart with an overview of the companies involved in supplying food to the
supermarkets investigated here.
35
quantification was performed during three days taking into account items
discarded but not recorded, and vice versa. The absence of some items from the
waste pile was tracked by asking the staff about every missing item to
determine whether the item was expected to be in another location than the
waste pile at that time, e.g. if some items were supposed to be discarded later
or had already been discarded. All items that the staff did not expect to be in
the pile were excluded from the study.
4.2.3 Data collection for delivered and sold mass
Sold products from all five departments investigated are recorded by the
cashier at the pay point in the supermarket, or at a self-scanning pay point.
These data are then stored in the financial records that the company is obliged
to keep. Most products are recorded with the European Article Number (EAN)
code on the packages, but some products, mostly fruit and vegetables sold
unpackaged, are weighed at the pay point and identified by a four-digit price
look-up (PLU) code typed in by the cashier. Mistakes in self-scanning or with
the PLU codes are likely to create uncertainty in the data. The extent of this
problem is unknown, but can be assumed to have no significant effect on the
results presented in this thesis.
Delivered fruit and vegetables are recorded by the supplier as part of the
financial records. These data were used in Paper I in order to calculate the
missing quantities.
The supermarket departments studied are defined by the retail chain. The
meat department sells fresh meat from terrestrial animals, mainly beef, pork
and chicken, but also lamb and game meat. It also sells grilled chicken, raw
sausages and some frozen meat. In the deli department, processed meat
products such as sausages, meatballs and cold cuts, as well as black pudding
and pâté, are sold. Besides dairy products such as milk, cream, butter and
yoghurt, the dairy department also carries eggs and beverages based on fruit,
vegetables or grain. The cheese department sells various cheeses, mainly hard
or semi-hard cheese, soft cheese and cream cheese, but also tofu. The fruit and
vegetable department sells a wide range of domestic and imported fresh
produce.
All food products sold in the departments investigated can be aggregated at
several levels. The lowest level of aggregation is the article level, where each
article is defined by individual article number (EAN code). Some of these
articles may have the same name, but different brands or package sizes. If the
article code was changed over time without any change to the article, it was
still considered as two separate articles in this thesis. The articles sold in the
stores are grouped into categories defined by the supermarkets. These
36
categories are grouped into departments, which in this thesis included the five
departments cheese, dairy, deli, meat and fresh fruit and vegetables, all
belonging to the division of perishable food. Since the store has no level for
apples or oranges, an aggregation level between article and category, called
product level, was created. The definition of products was not as robust as the
other aggregation levels set by the supermarket, since there are several possible
sublevels where the product level can be set and this level differs between
different products. This can be exemplified by granny smith apples, which
have more than one article number. In this thesis the product level was set to
apple, but not to granny smith apple, which could also have been a possibility.
4.2.4 Analysis of waste data
Articles sold piecemeal were allocated a mass based on the mass stated on the
package when this was possible. For articles sold without packaging (only
FFV), the mass was set using the estimates used by the supplier for each
article. (All masses stated as tons in this thesis refers to metric tons.)
Relative waste (RW) was calculated either in relation to the actual mass
delivered (D) (Equation 1) or in relation to estimated mass delivered (Equation
2). The sum of sold products (S), pre-store waste (PW) and in-store waste (IW)
was used as estimated mass delivered. The difference between the equations is
the lack of a ‘missing goods’ term in Equation 2.
(1)
(2)
Equation 2 was mostly used in this thesis due to the lack of data on actual
delivered mass of cheese, dairy, deli and meat. The exception was in Paper I,
where Equation 1 was used since delivery data were available for the fresh fruit
and vegetables department.
For unrecorded in-store waste, the difference between measured waste and
recorded waste was calculated for each supermarket studied. The percentage
difference was then used to calculate the difference for a whole year for each
store, which gave the mass of unrecorded in-store waste.
4.2.5 Identification of systematic causes and risk factors of waste
The causes of food waste can be divided into systematic causes, which are
often small but happen over a long time or on many occasions, and occasional
causes, which are often the outcome of mistakes or rarely occurring events.
Three systematic causes or risk factors, short shelf-life, low turnover and
large minimum order size, were analysed in more depth in Paper II. Shelf-life
37
(SL) was defined as the time between the production (or packing) date and the
best-before date (or use-by date) or, in the case of eggs, from the production
date to the last legal sale date (Persson, 2015). Turnover (T) was defined as the
average number of items sold per week in weeks when the product was sold.
Minimum order size (MOS) was defined as the minimum number of items a
store can order on a single occasion. This was assumed to equal the wholesale
pack size, which is the number of items delivered together in some kind of
distribution package.
The waste risk factors were analysed in Paper II with the focus on organic
products, which are often found to have high waste ratios. To test the
hypothesis that low turnover, in combination with fluctuating demand, leads to
wasted products, waste quantifications were supplemented with data on
minimum order size and shelf-life for those deli products for which DAGAB
had available data. The data on MOS (number of items) and SL (weeks) were
combined with data on weekly turnover T (number of items per week) for each
store to calculate the β-indicator (β), as shown in Equation 3.
(3)
The β-indicator was used to explain part of the organic food waste in the
dairy, cheese, deli and meat departments (Paper II), but since the data for both
conventional and organic waste were used, the β-indicator can be applied to
other products, especially those with low turnover. The β-indicator was
developed in Paper II, but multiple linear regression (MLR) was also
performed to confirm this method. The method of using MLR to obtain an
equation describing how waste depends on T, SL and MOS was further
developed in Paper IV, where it was used to simulate the outcome of prolonged
shelf-life.
To establish the connection between reduced food waste and extended
shelf-life, the model first presented in Eriksson (2012) and further developed in
Eriksson & Strid (2013), Paper II, Björkman (2015) and Persson (2015) was
used. This model employs multiple linear regression to provide an equation
describing how the relative waste depends on T (sold items per week), SL
(days) and MOS (number of items) from products where data on all parameters
are available. In the MLR, the analysis was limited to only include food items
with a shelf-life shorter than 85 days. The result was based on 984 articles
consisting of 92 cheese articles, 258 dairy articles, 333 deli articles and 311
meat articles and Equation 4 were created from the MLR results, with an
adjusted R
2
value of 0.666. The reduction in relative waste depending on
increased shelf-life was calculated with Equation 4 and then applied to the
recorded waste of each product in Paper IV.
38
(4)
4.3 Carbon footprint of processes related to food waste
Life cycle assessment (LCA) (ISO, 2006a; 2006b) was used to calculate the
global warming potential (GWP) associated with cradle to retail emissions in
Paper III, emissions related to cold storage in Paper IV and different waste
management options in Paper V. The functional unit used was always 1 kg of
food, but due to the different contexts both production or prevention of 1 kg
food delivered to the supermarket (Papers III and IV) and removal of 1 kg food
(waste) from supermarket (Paper V) were used.
In LCA, emissions relating to waste are normally allocated to the product or
service assessed. Therefore food waste cannot have a carbon footprint (CF) by
itself, but just increases the CF of the consumed product. The food waste CF
used in this thesis should therefore be interpreted as the CF of the food before
it became waste, even though waste was not the intended product. From this, it
follows that if this waste were to be avoided, the life cycle emissions of that
specific product would also be avoided.
4.3.1 Carbon footprint associated with cradle to retail emissions
In all papers, CF was used synonymously with GWP
100
. The CF was expressed
in terms of carbon dioxide equivalents (CO
2
e). The CO
2
, N
2
O and CH
4
emissions were included, where the GWP of N
2
O and CH
4
was expressed
relative to CO
2
according to the IPCC values (Solomon et al., 2007).
In order to analyse the carbon footprint pattern of retail food waste, the CF
of cradle to retail was calculated for different food products. Waste
management of the food waste was not included, due to the low impact
described in Nilsson (2012) and Paper V. The waste carbon footprint was
defined as the specific CF value of a product, comprising emissions associated
with the production and distribution up to delivery to the supermarket,
multiplied by the total mass that was wasted in the stores (including pre-store
waste) of the respective product. The specific CF values were determined
based on existing literature, but the literature values were modified regarding
transportation in order to better fit the distance from the actual country of
origin to the supermarket located in Stockholm, as described in Scholz (2013).
The CF from cradle up to delivery to the retailer of all products was
calculated based on information from the literature. These CF values and the
literature consulted are listed in the appendix to Paper III. When more than one
study on a specific product existed, the study that best represented the product
at the store in terms of country of origin and production method and which
39
used most current data was selected. Where the scope of the available literature
did not exactly fit the purpose of the present study, assumptions or calculations
were made as described in more detail in Paper III. In general, the most
commonly included emissions associated with primary production, as well as
emissions caused by processing and transportation up to the retailer, were
considered. Potential emissions from land use change were not included.
Emissions associated with store operations and packaging were also not
included, since data availability was not sufficient and their impact was
considered to be relatively low (Cederberg et al., 2009; Stoessel et al., 2012).
4.3.1 Carbon footprint associated with waste management options
In Paper V, five food products with different properties were selected to
represent different waste streams that could be separated in the supermarkets.
For each of the food products, a waste management scenario was applied and
the CF associated with the management and substituted systems were
calculated. The scenarios used were landfill, incineration, composting,
anaerobic digestion, animal feed and donation, since they all represent possible
ways to treat food waste locally with existing infrastructure. The first five
waste management options have been described in several studies (Laurent et
al., 2013a) and the methodology is therefore well used. However, to the best of
my knowledge the same methodology with system expansion has not
previously been applied to food donation, with the exception of Eriksson
&Strid (2013).
In the system expansion, the donated food replaced other food products that
would otherwise have been bought by the charity and consumed by people in
need. There is a wide variety of food items that could be replaced by donated
products, but the same assumption as made by Eriksson & Strid (2013) was
used, i.e. that all donated food replaced bread based on energy content. The
reason why bread was selected as a substituted product is because it is one of
the cheapest types of food that can be bought in Sweden with regard to energy
content, and because it does not require preparation, unlike other cheap and
energy-rich products such as pasta and potatoes.
40
4.4 Waste prevention and valorisation framework
In this remainder of this thesis, the framework presented in Figure 7 is used to
describe how different waste management and prevention options relate to
food waste and to each other. This framework is inspired by Papargyropoulou
et al. (2014), Garrone et al. (2014), Östergren et al. (2014) and Eriksson &
Strid (2013), but focuses only on the supermarket perspective. Due to
supermarket specialisation, no distinction is made between avoidable and
unavoidable food waste, since these are not separated until the consumer stage.
The concept of waste prevention differs depending on the perspective. From
an environmental perspective, waste is prevented as long the food is never
produced or used for its intended purpose, i.e. eaten by humans. From an
economic perspective, it would be a waste to sell the food at a reduced price,
since that is a loss of money. With this logic, the measure of cutting the price
by 50% on the day before the best-before day may prevent food from being
wasted, but still wastes some of the value of the product. However, since a
price reduction also means that half the value is saved and since this thesis
Figure 7. A waste management framework inspired by Papargyropoulou et al. (2014), Garrone et
al. (2014), Östergren et al. (2014), Eriksson & Strid (2013) and findings within this thesis.
41
applied an environmental perspective, this type of measure was categorised as
prevention through economic valorisation (Figure 7), since the food is sold
through normal channels with a price reduction in order to save some of the
economic value and possibly the whole environmental value.
In Figure 7 there are a few important trends that follow the order of priority
in the EU waste hierarchy. First, the less prioritised measures are all general
and do not require food waste with high levels of product quality, biosecurity,
separation or storage conditions. Therefore these options are cheap and
general, but have an outcome with much lower economic value than the
original food products. In order to prevent food from being wasted (i.e. using it
for human consumption), there are high hygiene requirements that need to be
met, which makes separation and proper storage important. These options
therefore need more effort from the supermarket, but in return provide a more
valuable outcome. The problem is that the outcome of most waste management
options is profitable for society (SEPA, 2011, 2012), but not necessarily for the
supermarket.
42
5 Results
This results chapter is structured into three sections describing quantities, risk
factors and measures. The first section mainly presents results from Papers I-
III, while data for the years not included in these papers are presented in
Appendix I-IV, including department data, category data, product data and
article data. The second section describes a few problems causing food waste
which are mainly covered in Papers II and IV. In the third section, different
measures covered in Papers IV and V that reduce waste or reduce the carbon
footprint of food waste are described.
5.1 Quantities
5.1.1 Quantities of wasted perishable food
During 2010 to 2014, a total of 2.4 kton of food waste was recorded in the five
departments studied in the six selected supermarkets. The majority (84%, 2.0
kton) of the recorded mass was wasted in the fresh fruit and vegetables
department and 77% or 1.6 kton of this was recorded as pre-store waste.
A summary of waste from the different supermarket departments during
three years (2010-2012) is shown in Figure 8. Fruit and vegetables had a
dominant position when the mass of waste was considered, contributed 86% of
the waste, but only 72% of the cost of the waste and 48% of the carbon
footprint produced in vain when wasting the food. The meat department
displayed the opposite pattern to the FFV department, since meat only
contributed 4% of the wasted mass, but 12% of the cost of the waste and 30%
of the wasted CF. The deli and cheese departments followed the same trend as
the meat department, but on a slightly smaller scale.
43
In order to find hotspots in the waste data, it is not only whole departments
that need to be investigated, but also the actual products within these
departments. Paper I showed that the products making the largest contribution
to FFV in-store waste mass were everyday fruit and vegetables, which are sold
in large quantities, and not exotic fruits, which have higher percentage waste.
For organic deli products, the largest waste contribution also came from
products sold in large quantities, e.g. meatballs and Falun sausage (Paper II).
Since Paper I only shows in-store waste during 2010 in the analysis of wasted
products and Paper II only organic products, Appendices II-IV present data on
the most wasted categories, products and articles.
For each of the five departments, a few articles represented a large share of
the total wasted mass. The most extreme was the dairy department, where five
products contributed almost half (47%) of that department’s waste. In the other
departments, the top five most wasted products contributed between 34% and
41% of the waste within each department. In terms of the carbon footprint, the
concentration of waste from a few products differed and for some departments
was even higher. In the FFV department, tomatoes, bananas and lettuce made a
combined contribution of 36% to the waste carbon footprint. In the meat
Figure 8. Total waste of perishable products from different departments in the six supermarkets
studied during five years, quantified in terms of mass, purchase cost to the supermarket and
carbon footprint (CF) associated with the lifecycle from cradle to retail for the wasted products.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Mass
(86 ton/store/year)
Cost
(1.6 mSEK/store/year)
CF
(130 ton CO2e/store/year)
Share of waste from each department (%)
Fruit and vegetables Deli Meat Dairy Cheese
CF
(130 ton CO
2
e/store/year)
44
department, minced beef contributed 17% of the carbon footprint associated
with that department’s waste (Appendix III).
Since food waste can be quantified in different units, it is important to set
the goals for waste reduction using units that actually measure what is intended
to be achieved. Figure 9 exemplifies this issue when comparing the five
supermarket departments investigated in terms of wasted mass and wasted
carbon footprint. In comparison with 2010, the waste increased by 12% in
2011 in terms of mass, but decreased by 5% in terms of wasted CF. This was
due to increased FFV waste and decreased waste of mainly meat. The trend of
both increasing and decreasing waste continued to 2013, when the waste in the
FFV department also started to decrease. This led to a total decrease in food
waste, both in terms of mass and carbon footprint, from 2010 to 2014 by 21%
and 26% measured in mass and carbon footprint, respectively (Figure 9).
It is not only the use of different units that complicates food waste
quantification. The waste is often set in relation to something else in order to
make the results comparable between e.g. supermarkets of different sizes.
When the waste in the present case was related to estimated delivered mass
(Equation 2), it is clear that the main flow of delivered and sold food was
important. Since the sold mass in the six supermarkets studied decreased by
12% from 2010 to 2014, the relative waste presented in Figure 10 gives a
slightly different result compared with the absolute waste in Figure 9. When
0
5
10
15
20
25
30
0
2
4
6
8
10
12
14
16
18
20 10 20 11 20 12 20 13 20 14
Wasted CF (t CO
2
e)
Wasted mass (t)
Meat Mass
Deli Mass
Dairy Mass
Cheese Mass
FFV Mass
Meat CF
Deli CF
Dairy CF
Cheese CF
FFV CF
Figure 9. Total waste per year from different departments in the six supermarkets studied during
five years, quantified in terms of mass and carbon footprint (CF) associated with the lifecycle
from cradle to retail for the wasted products. The scale is set so the bars for 2010 are equally high
in the diagram.
45
the waste was related to the sum of sold and wasted mass in all five
departments, the relative wasted mass showed its peak value in 2012, while the
relative wasted CF peaked in 2013. It is also clear that the trend of reduced
wasted CF in 2010-2013 followed the reduced sold CF, and in relative terms
therefore did not decrease (Figure 10).
Apart from variations due to the use of different units and absolute/relative
numbers, there are also natural variations over time. In Figure 11 this is
illustrated by showing the weekly average relative waste for each store during
the whole study period. The highest weekly relative waste rate in a single
supermarket was 7.3% and the lowest 0.5%. Since there are only a few high
waste peaks in Figure 11, a long quantification period can be used to reduce the
influence of these peaks. If waste is quantified during a short period, it might
be heavily affected by an occasional high peak. If the waste is quantified for a
less aggregated level than a whole supermarket, the variation between high and
low values will increase.
0.0%
0.5%
1.0%
1.5%
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
20 10 20 11 20 12 20 13 20 14
Relative wasted CF (%)
Relative wasted mass (%)
Meat Mass
Deli Mass
Dairy Mass
Cheese Mass
FFV Mass
Meat CF
Deli CF
Dairy CF
Cheese CF
FFV CF
Figure 10. Total waste related to estimated deliveries per year from the different departments in
the six supermarkets studied during five years, quantified in terms of mass and carbon footprint
(CF) associated with the lifecycle from cradle to retail for the wasted products. The scale is set so
that the bars for 2010 are equally high in the diagram.
46
Figure 11. Mean, maximum and minimum values of the weekly relative waste in the six supermarkets studied during 2010-2014.
0%
1%
2%
3%
4%
5%
6%
7%
8%
1
7
13
19
25
31
37
43
49
55
61
67
73
79
85
91
97
103
109
115
121
127
133
139
145
151
157
163
169
175
181
187
193
199
205
211
217
223
229
235
241
247
253
259
Relative waste (%)
Week
Mean Min Max
2010
2011
2012
2013
2014
47
5.1.2 Mass balance of fresh fruit and vegetables
In Paper I, a mass flow analysis was performed to create a mass balance of
fruit and vegetables in the six supermarkets, where 94.6% of the delivered
mass was sold and the rest was distributed over the three categories of waste
(4.3%) and missing quantities (1.1%). Missing quantities were calculated
simply to achieve a balance between the inflow and outflow of food.
A similar mass flow analysis was performed with data for the following
years (Table 3) with a maximum of pre-store waste during 2012, the same year
as missing quantities reached the maximum. It is worth noting that the year
with the highest pre-store waste (2012) also had the lowest in-store waste,
while the year with the lowest pre-store waste (2014) had the highest in-store
waste. Since the unsold mass was also lowest during 2014, it appears that the
larger the share of in-store waste, the less total waste there is.
Table 3. Results of mass flow analysis of fruit and vegetables in all six supermarkets studied
during all five years. ‘Unsold mass’ corresponds to delivered mass minus sold mass; ‘Recorded
waste’ corresponds to recorded pre-store and in-store waste
Year
Delivered
mass
(%)
Pre-store
waste
(%)
In-store
waste
(%)
Sold
mass
(%)
Missing quantities and
unrecorded waste
(%)
Unsold
mass
(%)
Recorded
waste
(%)
2010
100
3.0
1.0
94.6
1.4
5.4
4.0
2011
100
4.2
0.9
94.1
0.8
5.9
5.1
2012
100
4.6
0.9
92.7
1.7
7.3
5.5
2013
100
3.6
1.2
94.1
1.1
5.9
4.8
2014
100
2.5
1.3
95.1
1.1
4.9
3.8
All
100
3.6
1.1
94.0
1.3
6.0
4.7
The missing quantities can be explained by theft or weight loss due to
evaporation and this problem was in line with the in-store waste (1.3%
compared with 1.1%). However, the small yearly variation seems to be a
coincidence when looking at the same data divided by supermarket (Figure
12). Here the variation was larger and two supermarkets even showed negative
missing quantities, which makes the theft and evaporation explanation more
problematic. Store 3 stands out, with negative missing quantities during three
of the five years, reducing the unsold mass to a lower level than recorded
waste. Store 2 stands out with negative missing quantities that equal the
recorded waste during 2012, meaning that 100% of the delivered mass was
sold, according to the records on delivered and sold products.
The other supermarkets investigated had no negative missing quantities.
Instead, they had large masses in this category, with a maximum in
supermarket 6 during 2011 of 5.7% of delivered mass, which was higher than
48
the sum of in-store and pre-store waste (5.2%). This high loss of mass also
resulted in the largest difference between delivered and sold products, where
only 89.1% was sold (Figure 12).
5.2 Risk factors for food waste and causes of discarding food
Analysis of causes and risk factors for food waste in supermarkets is important
in order to progress from identifying flows of waste to actually reducing these
by introducing measures. Often these measures solve some kind of problem
that has been causing waste or limit the potential effect of risk factors. Expired
shelf-life is one reason for discarding food and short shelf-life could therefore
be considered a risk factor for food waste. Short shelf-life was analysed in
Paper II to find out if the time span between packing date and best-before date
could explain the greater waste of organic products in comparison with
conventional products. Since shelf-life did not differ between organic and
conventional products, Paper II did not find short shelf-life to be a cause of
food waste. However, a plot of the logarithm of shelf-life against the logarithm
of relative waste (relative to the sum of wasted and sold mass) and absolute
waste revealed a trend for increasing waste as shelf-life decreased (Figure 13).
Nevertheless, the curve fit was far from perfect and therefore other risk factors
must also influence the level of food waste.
Figure 12. Results of mass flow analysis of fruit and vegetables for each of the six supermarkets
studied during each of the five years investigated. The level of net unsold mass corresponds to the
difference between sold and delivered mass.
-4%
-2%
0%
2%
4%
6%
8%
10%
12%
2010
2011
2012
2013
2014
2010
2011
2012
2013
2014
2010
2011
2012
2013
2014
2010
2011
2012
2013
2014
2010
2011
2012
2013
2014
2010
2011
2012
2013
2014
Store 1 Store 2 Store 3 Store 4 Store 5 Store 6
In-store Pre-store Missing+unrecorded Net unsold food
49
In Paper II, low turnover was also investigated as a risk factor to explain the
high relative waste of organic products. In Figure 14, the logarithm of turnover
is plotted against the logarithm of relative waste in order to identify its
potential to cause food waste in comparison with the shelf-life factor. The trend
visible in the diagram and the R
2
values of the linear fit both indicate that low
turnover to a higher extent than short shelf-life caused food waste. However,
the analysis is not complete, since low turnover could also be seen as a risk
factor rather than the only root cause, which means that stocking too many
products will produce waste since they might not all be sold before the end of
the shelf-life, but the effect of overstocking will be even worse if the turnover
is very low. Figure 14 also illustrates that the waste in absolute terms increases
with increased turnover, simply because larger volumes are handled, but at the
same time the relative waste decreases with larger turnover.
Since the minimum order size (MOS) is one parameter influencing the
inflow of products, it was also analysed to see how it corresponded to relative
waste. It was found that MOS was corresponded even less to relative waste
than turnover and shelf-life (Figure 15).
-4
-3
-2
-1
0
1
2
3
4
0.5 1 1.5 2 2.5 3
Waste (Log (Waste))
Shelf life (Log (days))
Waste (%) Waste (kg) Linear (Waste (%)) Linear (Waste (kg))
Figure 13. Logarithm of shelf-life is plotted against logarithm of relative waste (red symbols)
and absolute waste (blue symbols), with a linear trend line added to each plot.
50
-4
-3
-2
-1
0
1
2
3
4
-1 -0.5 0 0.5 1 1.5 2 2.5 3 3.5
Waste (Log (Waste))
Turnover (Log (Items/week))
Waste (%) Waste (kg) Linear (Waste (%)) Linear (Waste (kg))
-4
-3
-2
-1
0
1
2
3
4
0 0.5 1 1.5 2 2.5
Waste (Log (Waste))
Minimum order size (Log (items))
Waste (%) Waste (kg) Linear (Waste (%)) Linear (Waste (kg))
Figure 14. Logarithm of turnover plotted against logarithm of relative waste (red symbols) and
absolute waste (blue symbols), and a linear trend line added to each plot.
Figure 15. Logarithm of minimum order size plotted against logarithm of relative waste (red
symbols) and absolute waste (blue symbols), with a linear trend line added to each plot.
51
In Paper II, both large MOS and short SL were found to be risk factors for
food waste, but only turnover was found to differ between organic and
conventional products. Therefore low turnover is a risk factor explaining why
organic food has higher relative waste than conventional products. The fact
that food is organic was not found to be a risk factor, although higher
production costs can have an effect on the sell price and thereby the turnover.
5.3 Measures
In order to reduce food waste, there is a need for waste reduction measures.
Potential target areas for these measures are the largest problems found in
waste quantification, if these problems are caused by issues that can be dealt
with. The basis used for categorisation of measures was presented in the
introduction section of this thesis. The measures presented in this sub-section
focus on the prevention measures presented in Paper IV and the valorisation
measures presented in Paper V. Following this, these two fundamentally
different ways of dealing with food waste, source reduction and waste
management, are compared.
5.3.1 Prevention measures
The most efficient way of completely preventing food waste from occurring is
of course to stop overproducing food and thereby potentially cause a lack of
supply, but for obvious reasons this is not desirable. Instead, measures must
both reduce waste and not jeopardise food security, which makes achieving
complete prevention of food waste less likely. Therefore it might be more
correct to talk about waste reduction rather than waste prevention, since waste
is unlikely to completely disappear without radical changes in the food system.
Some of the waste at the end of the food supply chain can still be prevented,
however, thereby reducing the need for production. Just reducing the need for
food production does not mean that food production will actually be reduced,
but in all calculations of waste reduction benefits presented below this source
reduction was assumed to take place.
In Paper IV, food waste was reduced by prolonging shelf-life through
reducing the storage temperature for chilled products. Prolongation of shelf-life
has the potential to lead to reduced waste, but also to increased energy use. The
net effect of reducing storage temperature was calculated by deducting the cost
of increased electricity use from the potential savings from reduced food waste.
This gave a value for each supermarket department showing whether a
reduction in temperature was justified from a reduced waste perspective.
52
If the storage temperature used for cheese, deli and dairy products were to
be decreased from 8°C to 5°C, the waste associated with these products would
potentially decrease by 15%, from 7.8 to 6.7 ton/store/year. The corresponding
reduction for these products if the temperature were to be reduced from 8°C to
4°C and 2°C would be 18% and 25%, respectively. For the supermarket’s meat
department, a reduction in storage temperature from 4°C to 2°C would
potentially lead to a 19% reduction in mass of wasted meat. Taking each
supermarket department separately, the reduction potential would be in the
range 9-30% (Figure 16), with the highest reduction potential at the lowest
storage temperature in each department.
The largest net saving of carbon footprint was found in the meat
department, where the potential savings from waste reduction were larger than
the increased emissions related to reducing the storage temperature. This was
due to the comparatively high level of waste, but also to the high CF per unit
mass, which makes the need for cooling low and the potential waste reduction
high. The dairy department was the opposite to the meat department, with a net
cost of reducing storage temperature. This was due to the already low waste in
the dairy department, but also the large quantities of water in dairy products
that would be need to be chilled to reduce the storage temperature (Figure 17).
The deli and the cheese department can be described as intermediate
between dairy and meat, with a moderate carbon footprint, price per kg,
turnover and level of waste. The cheese department showed a trend for
decreased net savings when the temperature decreased and the shelf-life was
prolonged. In the deli department, the increased energy costs equalled the
reduced emissions associated with food waste, so the measure of reduced
storage temperature gave no net saving, just a shift from food-related emissions
to energy-related emissions.
53
Figure 17. Net effect of reduced storage temperature in different supermarket departments
considering the benefits, in terms of carbon footprint, of reduced waste and the cost of increased
energy demand.
-20
-15
-10
-5
0
5
10
15
0 1 2 3 4 5 6 7 8
Net saving (ton CO
2
e/store/year)
Temperature (°C)
cheese dairy deli meat
0%
5%
10%
15%
20%
25%
30%
35%
0 1 2 3 4 5 6 7 8
Wasted mass reduction potential (%)
Temperature (°C)
cheese dairy deli meat
Figure 16. Potential wasted mass reduction (%) for different perishable food departments at
different storage temperatures.
54
5.3.2 Valorisation measures
When food waste cannot be prevented, there are several different options
available to manage the waste. In Paper V, six scenarios with differing priority
in the waste hierarchy were evaluated. The results showed a trend for
decreasing levels of carbon footprint with higher priority levels in the food
waste hierarchy (Figure 18). For all five food products investigated, landfill
was the option with the highest carbon footprint. At the other end of the scale,
donation and anaerobic digestion were the alternatives with the lowest carbon
footprint from the five food products. Donation was the alternative with the
lowest emissions for grilled chicken and bread (even though incineration
proved the lowest emissions for bread), but for bananas, lettuce and beef
anaerobic digestion generated the lowest emissions.
The other scenarios did not fully agree with the waste hierarchy.
Incineration was a good option for dry food like bread and grilled chicken, but
a poor option for the wetter lettuce and bananas, for which composting
provided a better alternative (Figure 18). Similarly, anaerobic digestion was a
better alternative than animal feed, for some products better than donation.
According to these scenarios the priority order applied to bananas, grilled
chicken, iceberg lettuce, beef and bread should therefore be anaerobic
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Banana Chicken Lettuce Beef Bread
CF for WM (kg CO
2
e/kg FW)
Waste management scenario and type of food wasted
Landfill Incineration Composting Anaerobic digestion Animal feed Donation
Figure 18. Carbon footprint (CF) of each waste management (WM) scenario and food product
investigated in Paper V.
55
digestion, donation, animal feed, incineration, composting and, the least
favourable alternative, landfill, when solely considering carbon footprint and
general options. When considering specific options for specific food products,
incineration with energy recovery is a favourable alternative for dry foodstuffs.
Different foodstuffs have different features that make them more or less
suitable for different waste management scenarios. Bananas consist of a fairly
large proportion of peel that was sorted out in the donation scenario, which
meant that a lot of the wasted mass could not be used to replace bread. In the
other scenarios, however, the banana peel was managed the same way as the
rest of the banana and therefore only the donation scenario was affected. Since
the chicken was grilled it was much dryer than beef, which made the energy
content per unit mass higher and the water content lower. Grilled chicken was
therefore much better to incinerate than beef, but for the same reason it gave
rise to more methane in the landfill scenario. Grilled chicken produced less
methane in the anaerobic digestion scenario, because the product included the
whole carcass with bones. Bones were assumed either to be sorted out in the
pre-treatment or simply not to produce any methane due to the short retention
time in the biogas reactor. The bones were also considered not to be eaten in
either the donation or animal feed scenario, which reduced the outcome from
the chicken in these scenarios.
Lettuce has a low energy content and a high water content, which is why
lettuce could be treated in any of the scenarios investigated without large
differences in outcome. Bread was the opposite, with high energy content and
low water content. Because of its energy carrying capacity, bread was useful
for incineration, anaerobic digestion, animal feed and donation. However, its
biogas potential was not as high as for meat products like beef and chicken,
which resulted in less methane production in the landfill scenario and
anaerobic digestion scenario. The energy content per unit dry matter was also
higher in chicken and beef, due to a higher fat content.
5.3.3 Comparison of valorisation and prevention measures
Placing the results from Paper IV into the context of Paper V allowed
comparison of all stages in the waste hierarchy. Figure 19 shows the outcome
using all wasted beef products as an example. The combined waste was 0.95%
in relation to estimated delivered mass. In the waste prevention scenario the
waste was reduced by 20% when the storage temperature was reduced from
4°C to 2°C.
56
Even when the waste reduction was only 20% of the wasted beef, the
prevention scenario reduced the carbon footprint from the food waste more
than any other waste management scenario (Figure 19). The second best
alternative was anaerobic digestion, which reduced the carbon footprint by 0.7
kg CO
2
e/kg food waste. This is much lower than the prevention scenario,
which reduced the carbon footprint by 4.2 kg CO
2
e/kg food waste. In the
prevention scenario, 80% of the food waste was composted (in line with the
donation scenario in Paper V) but if this waste were instead sent for anaerobic
digestion, the benefit of the prevention measure would increase to a reduction
of 4.8 kg CO
2
e/kg food waste.
Figure 19. Comparison of different ways to manage all wasted beef in the supermarkets using a
combination of scenarios in Paper IV and Paper V and the effects on carbon footprint (CF). In the
prevention scenarios, 20% of the waste was prevented and 80% was managed by composting or
by anaerobic digestion.
-5
-4
-3
-2
-1
0
1
2
CF for waste management (kg CO
2
e/kg FW)
57
6 Discussion
Since only a small part of this thesis focused on causes of food waste, this
discussion chapter focuses on quantities and measures in separate sections,
while causes are included in both these sections.
6.1 Quantities of food waste
From the data presented in this thesis, it is clear that the waste composition is
dominated by fruit and vegetables. This is well in line with other studies (e.g.
Buzby & Hyman, 2012; Beretta et al., 2013; Lebersorger & Schneider, 2014),
although their dominance would probably have been reduced if bread had been
included in the study (Scherhaufer & Schneider, 2011; Lebersorger &
Schneider, 2014; Stensgård & Hanssen, 2014). Other studies do not include the
pre-store waste that dominated the fruit and vegetable waste in the present
case, but since the sum of pre-store and in-store waste from the stores
investigated (4.7%) corresponded almost exactly to the most frequent waste
level in Lebersorger & Schneider (2014), it is likely that the waste was at a
normal level and that recorded cost was just booked differently than in
supermarkets in other studies.
Even if the waste was on a similar level as other studies, there was
considerable variation in the material. First, there was large natural variation
within the stores over time and between articles. Data can also be presented
using different units or relative numbers, which increases the number of
possible perspectives although it does not actually increase variation.
6.1.1 Use of different units for quantification
Choice of analytical method had an effect on the results presented in this
thesis. For example, all results presented in terms of mass resulted in bulky
products with a high water content, e.g. fruit, vegetables and dairy products,
having a large influence on the results. When the results were presented using
58
the monetary value of the waste, more expensive products, e.g. herbs such as
basil, gained importance at the expense of e.g. potatoes. Use of carbon
footprint shifted the focus relatively more to meat and cheese products rather
than fruit and vegetables and dairy. The weakness of using units of mass in this
kind of study is that the products with a large environmental impact can be
associated with low mass, which can be interpreted as meaning that they are
not important (Strid, 2012). For this reason, the monetary value corresponds
better to environmental impact than units of mass and would therefore work
better as an indicator of environmental hotspots then analysing units of mass.
The strength of using mass values is good transparency, since the unit is well-
defined and does not change along the food supply chain, except during
processing. Both monetary values and values describing the environmental
impact need detailed definitions and have a tendency to differ over time and
along the value chain, even without processes that change the properties of
foodstuffs. For example, the value of products increases not only when they are
processed, but also when they change owners, are transported or are kept in a
cold storage.
Using a unit of mass makes the results comparable with those of other
studies. However, it is not only the units that make comparisons complicated.
Results based on monetary values are often compared with the value of sold
products, since this is the basis of income in a company and the figure against
which all costs must be compared. When percentage waste is as low as it was
in this study, this causes no significant problems, since percentage waste of
1.00% calculated with Equation 1 corresponds to a value of 1.01% if the waste
is compared with the sold value instead. The choice of comparison is more
influential for products with higher values of percentage waste. Some of the
exotic fruits described in Paper I, with waste of above 50%, would have values
of over 100% if the waste were related to sold quantity instead of delivered
quantity.
6.1.2 Data quality and selection of study objects
The six supermarkets used in this work were selected by the parent company,
which introduces a possible bias, even though the company claimed that they
represented the average. It is unlikely that the company selected stores with
high percentage waste, since high levels of waste tend to be something
shameful and might repel customers if the information became publicly
available. Therefore the supermarkets studied can be expected to represent an
average Willy:s store or have lower percentage waste than the average Willy:s
store. The selected stores were also found to be larger than average in terms of
turnover of fruit and vegetables (Paper I), which further increases the potential
59
for them to waste less than average (Hanssen & Schakenda, 2011). However,
even if the representativity cannot be proven, all supermarkets within the
company are based on a detailed concept (Willy:s, 2010), making large
variations between individual supermarkets unlikely. The level of waste in the
six supermarkets investigated is therefore unlikely to differ greatly from the
average supermarket within the Willy:s chain.
Material flow analysis performed in Eriksson (2012) showed that the
unrecorded waste category and missing quantities differed in size between
departments. These two categories are a good indicator of the quality of
recorded data. If large quantities are lost without any reasonable explanation, a
likely cause is that the recording of waste does not function well and items are
discarded without recording. From the analysis, it is clear that data based on
EAN code scanning are more accurate than data based on estimated weights.
Therefore the results for cheese, dairy, deli and meat can be considered more
accurate than those for FFV. This is true even though efforts were made to
quantify unrecorded in-store waste of FFV by physical measurements.
6.1.1 Uncertainties in carbon footprint of food
The carbon footprint of the wasted food products in Paper III was mainly
calculated based on the existing literature on LCA studies. Although the LCA
methodology is ISO-standardised, the choice of some aspects, such as the exact
system boundary, functional unit, allocation method or use of emission factors,
is slightly open. Therefore, the results for the same product can vary and in
general the results of different studies are not directly comparable. Moreover,
for agricultural products the chosen production system and the production
country is crucial. In this thesis work, most effort was put into getting the
necessary background information and evaluating the carbon footprint that was
most representative for the wasted products. However, in some cases rather
broad assumptions had to be made.
The carbon footprint of the different deli products was calculated based on
assumptions on meat content and energy requirements. Although information
about the total meat content of the products was generally available, mostly no
information about the exact content of meat type was given. Since most
products contain beef and pork to some extent and the carbon footprint of beef
is almost five times larger than that of pork, this could have a significant
impact on the results. Moreover, the meat content of individual products can
vary, for example between different brands. The meat content was generalised
for different product categories and it was considered that the deviations were
balanced on average. It was assumed that the non-meat content does not have
any impact on the overall result, since other ingredients are usually products
60
with a much lower carbon footprint than meat, for example water or potato
starch, and the relative impact is low.
Estimating the carbon footprint of processed dairy products is difficult,
since milk intake and other activities and the associated emissions can be
allocated to different products. For example, butter fat can be seen as a
byproduct from cheese production (Cederberg et al., 2009). Here, the wastage
carbon footprint of most dairy products, including milk and other fresh dairy
products, butter, butter blends and cheese, was calculated based on results from
a study by Flysjö (2012). In that study, total emissions associated with dairy
production of a large dairy company were allocated in a top-down approach to
the different products and milk intake was calculated for the different products
based on the weighted value of fat and protein. Only a limited number of other
studies on processed products was available. Processed foods like cheese
spreads and different deli products need to be analysed in more detail to
establish more accurate carbon footprint values. When analysing meat
products, development of a method to estimate the impact of different meat
cuts and byproducts such as offal should be considered.
LCA studies on the production of fruit and vegetables often address only
one or a few production sites, so the results are specific for the particular
system. Since produce is often imported into Sweden from many other
countries, the wastage carbon footprint was calculated based on the share of the
product from its different countries of origin. Therefore, the focus was on
finding LCA studies on the countries’ typical production systems, which was
not always possible. To give a picture of variations in a product’s carbon
footprint, tomatoes are used below as an example, since they have a high waste
carbon footprint and therefore a dominant position in the waste carbon
footprint of the whole fruit and vegetables department.
Tomatoes sold in the six supermarkets investigated mainly originated from
the Netherlands and Spain. It can be assumed that most tomatoes are grown in
greenhouses and for the carbon footprint it is crucial if these greenhouses are
heated and, if so, how this heat are produced. For example, tomatoes grown in
a heated greenhouse in Sweden have an estimated CF of 2.7 (Biel et al., 2006)
or even 3.7 kg CO
2
e/kg (González et al., 2011), while according to Davis et al.
(2011) the average CF is 0.66 CO
2
e/kg due to increasing use of biofuel in
greenhouse heating. Reported values for Dutch production are between 0.78-2
kg CO
2
e/kg (Antón et al., 2010) and 2.9 kg CO
2
e/kg (Biel et al., 2006). In
Papers III and IV, the value of 2 kg CO
2
e/kg estimated by Antón et al. (2010)
was chosen for Dutch tomatoes, since it considered the use of a combined heat
and power plant, which is common in Dutch greenhouses. For Spanish
production, the values range from 0.05 kg CO
2
e/kg for production in the open
61
field (Muñoz et al., 2007) to 2.64 kg CO
2
e/kg for baby plum tomatoes in
heated greenhouses (William et al., 2008). In Paper III, it was assumed that all
Spanish tomatoes were grown in an unheated greenhouse tunnel. No distinction
was made between different tomato varieties, even though William et al.
(2008) showed that for example vine tomatoes are associated with higher
emissions due to lower yields.
The analyses for other fruit and vegetable products were based on less
complicated assumptions than were made for tomatoes, but due to lack of
literature those for many products had to be based on LCA results for similar
products or similar production systems. These assumptions potentially create
large uncertainties, but when addressing a large variety of products produced in
a variety of conditions, it is not possible to address each one of them in detail,
which makes rough assumptions necessary.
Overall, the results have to be viewed with caution. LCA studies always
include uncertainties (e.g. Röös, 2013), and for some products broad
assumptions had to be made. Nevertheless, the results can be considered to
give a good picture of the potential climate impact of food waste in the
supermarkets studied and to reveal the differences between different product
groups.
6.1.2 Issues regarding data quality for fruit and vegetables
The largest recorded mass flow of waste came from the fresh fruit and
vegetable department, which could be a target for waste reduction measures
just for this reason. In Paper I rejection was identified as the main cause of this
large mass flow of waste, but since Paper I only covered data from 2010, it is
clear that the additional results presented in this thesis make the result difficult
to interpret. The first problem that must be addressed with rejections is their
root cause. In theory, substandard quality should be the only reason for
rejection (apart from the obvious reason that the food is missing on delivery
and the store therefore refuses to pay for undelivered goods). However,
Eriksson (2012) and Eriksson & Strid (2013) found that rejections can also be a
consequence of efficient waste reduction. The example they use to illustrate
this is that when a store decreases in-store waste of bananas, pre-store waste of
bananas increases. If the shift were balanced so the total amount of wasted
bananas remained the same, this would just be a matter of accounting, but
Eriksson (2012) showed that the total banana waste increased, since the pre-
store waste increased more than the in-store waste decreased. One explanation
for this could be that supermarket staff put less effort into orders when the cost
(to the individual supermarket) of wasting products decreases. However, one
62
example cannot fully prove that this is not due to coincidence, and the result
should be interpreted with caution.
Eriksson & Strid (2013) discuss the rejection problem in the context of
waste reduction measures and suggest a limit on rejections as a way to reduce
waste, which of course would be impossible if the products were truly
unsellable. It can therefore be assumed that supermarkets use the system in a
way that increase profits, but also increases wasted mass as a consequence.
Tapper et al. (2013) describe this behaviour as opportunistic and argue that
rejections are not at all due to opportunistic individuals, but to inexperienced
staff and misunderstandings regarding quality requirements. Whether staff are
driven by an opportunistic desire for easy profits or whether they are the
victims of a system rewarding them for making decisions that lead to increased
waste is a rather philosophical question that is not answered here. However,
regardless of the reasons, product rejection is a problematic area and a hotspot
for reduction measures.
In this context it is positive to see that the initial increasing trend in
rejections of fruit and vegetables during 2011 peaked in 2012 and declined in
2013, to reach the lowest level in 2014, since total waste seemed to reflect pre-
store waste (Table 4). The in-store waste showed the opposite trend, with low
levels when the pre-store waste was high and higher levels when the pre-store
waste was lower. However, since pre-store waste was much more dominant,
the total effect was marginally influenced by the in-store waste.
There is also a problematic dimension of interpreting the results for the fruit
and vegetable department due to the findings in Figure 6, where the missing
quantities fraction was negative in four of the 30 yearly mass balances. From
an environmental perspective it is of course positive that food does not get
wasted, but from a scientific perspective it creates problems since it is difficult
to explain and reveals a potentially large quality issue in the data set.
The reason for the negative missing quantities cannot be fully clarified and
the only way to approach this problem is to discuss a broader range of possible
explanations, including why some explanations are less likely than others. The
missing quantities can of course have natural causes, like corrupt data, but
since great resources are put into data recording, great uncertainties seem
unlikely. The discrepancy could be due to deliveries that are heavier than
declared, e.g. the supplier expects a weight loss due to evaporation and
therefore loads a box with 5.5 kg tomatoes with declared weight of 5 kg. If the
evaporation is less than the expected 10%, the supermarket can actually sell
more tomatoes than it bought according to the records. This probably occurs,
but it is an unlikely explanation since the evaporation should be similar for all
63
stores, and if the storage conditions are not changed drastically the evaporation
would be expected to be the same every year.
An outflow that is larger than the inflow is not a possible explanation, but
this would be possible if data were corrupt in a way whereby the supermarket
could not only reclaim the money for some goods, but also sell them. This
would appear to be highly immoral and opportunistic, but could also have less
opportunistic reasons, such as bad routines that make overestimation of
rejections possible. One supermarket had a routine of always rejecting two
boxes of peppers for every pallet of peppers delivered. This was not because
they wanted to ‘steal’ the peppers, but because they did not want to waste time
going through all the boxes and sorting out the bad ones. Just by making an
innocent and reasonable assumption like this, it would be possible to receive a
shipment of food, reject two boxes and then sell the whole shipment, creating
an outflow that is 10 kg larger than the inflow. Whether this is immoral or not
is for others to debate, but if this is a true explanation it reduced the actual
waste of fruit and vegetables, even though the records indicated differently. It
also indicated that if waste records within fruit and vegetable departments
cannot be trusted, a better way to investigate this waste should be devised, e.g.
by comparing delivered with sold mass rather than the waste records.
6.1.3 Comparison of indicator values of waste generation
When only using the quantification data for calculating the relative waste, the
choice of method for calculating the waste appeared to be of less importance,
since Table 2 shows a large variation in methods used, but not large differences
in results. However, this can be due to the fact that most studies are based on
aggregated data, which compensate for the variations in different articles or
stores. When quantitative data from limited case studies are used for estimating
national levels of national food waste, it is clear that variations can strongly
influence the results. For example, Figure 20 compares a key figure, waste per
full time employee, for the supermarkets studied in this thesis and other
Scandinavian studies that have used this key figure to estimate the national
waste. It is clear that the mean value used varies between the studies and for
example that the waste in Denmark (Miljøstyrelsen, 2014; Landbrug &
Fødevarer, 2015) is much higher than in Sweden (Stare et al., 2013; SEPA,
2013). Why Danish supermarket waste (153 000 tons/year or 27 kg/capita) is
more than double that in Sweden (70 000 tons/year or 7 kg/capita) is not easy
to explain, since the countries could be expected to be fairly comparable.
However, since the values on which both studies base their estimates vary
widely between minimum and maximum and overlap each other, it is possible
that a different selection of supermarkets would give a very different result. In
64
Figure 20, it is clear that at least the six Swedish stores investigated in this
thesis have a waste level more similar to the Danish estimate (Miljøstyrelsen,
2014) when including pre-store waste. They also showed smaller variation than
the other studies, but the variation was still large considering that these six
stores are likely to be very similar in comparison with the variety of stores used
in the other studies.
One possible explanation for the great difference between results in Stare et
al. (2013) and the present thesis, beside the selection of stores, is the discount
profile of the stores concerned. Since labour costs are fairly high in Sweden, it
is possible that the supermarkets investigated here have been able to cut costs
more on staff than on food waste, giving a waste per employee figure that is
higher than in other stores that are more heavily staffed. This agrees with the
results from Miljøstyrelsen (2014), where discount stores had the highest level
of waste per employee of all types of store investigated.
0
1
2
3
4
5
6
Avfall Sverige
(2006)
Jensen et al
(2011a) + Stare
et al (2013)
Miljøstyrelsen
(2014)
Present study
(Excluding pre-
store waste)
Present study
(Including pre-
store waste)
Waste (ton/full time employee equivalent)
Average result
Figure 20. Comparison of three studies reporting values for waste per employee in supermarkets.
The mean, max and min values are presented, together with values from this thesis both including
and excluding pre-store waste based on five-yearly values for the six supermarkets studied. Mean
values are taken from Stare et al. (2013), but min and max values are taken from Jensen et al.
(2011a).
65
6.2 Waste reduction measures
6.2.1 Perspectives on waste prevention and valorisation
The food waste management framework presented in Figure 7 can be used to
better understand the conceptual differences of waste reduction measures. It
also gives a brief overview of how all of these measures can be evaluated with
a similar methodology in order to make the results comparable, even though
waste prevention and waste management are often regarded as completely
different processes (van Ewijk & Stegemann, 2015). The key to evaluating
prevention measures as if they were waste management options is to focus on
the substituted products in a system expansion. In the present case, the
prevented food waste replaced another identical product that was never
produced, distinguishing source reduction measures from valorisation
measures that use the food waste to substitute for products or services that are
likely to not be identical to the original product, even though they could
replace an identical product in the best case.
The environmental impact of` all waste management options, both
prevention and valorisation, depend heavily on the substituted product and the
cost of making the substitution possible. Since the original product is often the
most resource-dependent product that can be substituted, source reduction is
normally the most favourable way to manage waste. However, if the cost of
prevention is higher than that of the substituted system, the net effect will be
negative, meaning that introduction of the measure will result in a net cost. To
make this even more complicated, both the substituted system and the cost of
waste management can be measured in different values, such as money or
climate impact, since some measures can be beneficial from an environmental
perspective, but too expensive to be economically favourable (Paper IV).
The complication of valuing waste in different units is clear from the
measures of economic valorisation included in the framework in Figure 7. To
put it simply, these measures aim to recover at least some of the economic
value invested in the products instead of just throwing them away. The most
obvious example of such a measure is price reduction, where the whole product
is sold (and hopefully consumed). This means that waste is prevented, but in
comparison with selling the product at full price it represents a loss of some,
but not all, money. These measures are still considered prevention, since the
food is sold instead of being wasted, but they are fundamentally different to
other prevention measures that lead to source reduction. In order to put this
into the suggested food waste management framework (Figure 7), the
economic valorisation measures were assumed to replace food products that
are similar, but not necessarily identical, to the original product. An example of
66
this could be a customer looking for a piece of meat in a certain price range
and buying a piece of expensive meat with a price reduction instead of a
planned product. In this example, the customer would not have bought an
expensive product if the price had not been reduced, but a similar product with
a lower price. The measure therefore substituted for production of similar (and
probably cheaper) but not identical production. Another example of this is
presented in Strid & Eriksson (2013), where meat cuts were sold to a catering
company. There all meat cuts were assumed to replace a cheaper alternative,
but since all meat cuts from the same animal were associated with the same CF
per unit mass, the loss of carbon footprint was small due to this downcycling.
The other measures in Figure 7 that provide less favourable waste
management options than prevention can all be viewed as traditional waste
management methods. Most of the waste management scenarios use the food
waste to produce some kind of product that replaces another production system
with decreased value following the hierarchy down to the landfill scenario. In
that final scenario, no substitution is achieved and instead methane is produced,
which gives a higher carbon footprint than if the food had been left to degrade
in the air and just produce carbon dioxide (Paper V).
Another perspective on waste reduction measures is whether they are
specific or general, i.e. whether they can handle all kinds of food or just a
specific quality and/or specific products. The general measures are all waste
management options rather than waste prevention measures. Examples of such
general measures are incineration and composting, since all food products can
be used and transportation and management can therefore be handled at a low
cost. Examples of specific measures are donation, where only a safe quality of
the food can be given away and which requires more expensive logistics with
more frequent collections and chilled transport in order to keep the food safe
until consumption. Since the specificity of measures increases with increased
priority in the waste hierarchy, the top alternatives often require more sorting
or logistics, which makes them more expensive. On the other hand, they also
produce more valuable products, which can lead to a net benefit for the
supermarkets (Eriksson & Strid, 2013) or for society. Since the measures have
different benefits and costs, a combination of several measures can be seen as
the optimal solution, where some of the waste is reduced at source with
prolonged shelf-life (Paper IV), some is sold at a price reduction (Eriksson &
Strid, 2013), the surplus food that meets quality requirements is given to
charity (Paper V) and the rest is treated with the best available general waste
management option, such as anaerobic digestion (Paper V).
All waste reduction measures have different characteristics and some
important features, such as cost and potential savings, are often the main focus.
67
For example, reduced storage temperature for dairy products (Paper VI) has a
high waste reduction potential since it reduces the wasted mass, but the net
savings of reduced carbon footprint of the waste reduction are not enough to
cover the high cost of reduced storage temperature, which makes this an
inefficient measure, but still with the ability to reduce food waste. For meat
products, the savings from reducing storage temperature exceed the costs,
which gives a net saving, but the ability to reduce the wasted mass is lower
than for dairy products. This shows that waste reduction potential is important,
but since the actual goal of reducing food waste is to reduce the consequences
of the waste, the net benefit in terms of money or emissions must be
considered.
Even though reduced storage temperature does not have a reducing effect
on carbon footprint in all supermarket departments (Paper IV), a comparison of
beef products (Figure 19) clearly shows the potential of source reduction in
comparison with other waste management options from Paper V. Beef is the
most extreme example due to the large carbon footprint per unit mass (Paper
III), but it is clear that even a 20% reduction by far exceeds the outcome of the
best general waste management scenario, which in this case was anaerobic
digestion. The large difference in outcome from source reduction compared
with waste management was also observed by Bernstad Saraiva Schott &
Andersson (2014). This confirms that the exact order of priority of waste
management options might not fully agree with the EU waste hierarchy (Paper
V), but placing prevention in the most favourable position seems to be correct.
Efforts should therefore focus on source reduction whenever possible, even if
the waste reduction potential is small. In most cases even a small reduction will
reduce the carbon footprint from the wasted food much more than any waste
management option, and the most favourable waste management option should
therefore only be applied to waste that cannot be prevented.
6.2.2 Factors influencing the evaluation of waste reduction measures
There are several limitations with the theoretical evaluations of waste
prevention and valorisation options used in this thesis. They simplify complex
systems and use assumptions that do not necessarily represent reality, and
therefore the results should be interpreted with caution and other analyses
should be performed to either support or disprove the results. However, since
the field had not been investigated previously, this early study should be seen
as a starting point for a process of increased knowledge, rather than the end
point.
In Paper IV, many assumptions had to be made in order to link the
processes of reduced waste, increased shelf-life and increased electricity use.
68
This created many possibilities for small divergences from reality, which can
eventually add up to a large divergence. However, one assumption that had a
strong influence on the results was the carbon footprint associated with
electricity production. Depending on the electricity mix used, the results will
vary widely, from negative to positive savings, according to the sensitivity
analysis in Paper IV. Since there were also other ways to achieve the
temperature reduction, such as installing cabinet doors, there is great potential
to come up with a solution that includes significantly larger savings than
presented in Paper IV. It could therefore be questioned whether a temperature
reduction only gives net savings for fresh meat products, as concluded in Paper
IV.
Paper V also included assumptions that could be varied, thereby creating
variation in the results. The difference from Paper IV is that Paper V
investigated a series of parallel scenarios that only influenced each other to a
very small extent, while Paper IV examined a longer chain of processes where
small errors had a higher risk of adding up to large errors.
One assumption with potential influence on the results in Paper V is that the
composting scenario did not replace any product or service in the system
expansion. This assumption was made with the limitation that it should
describe the current situation. However, if the compost were to replace e.g.
peat for soil improvement, larger savings could be achieved in this scenario,
although this would require the full potential of the compost to be used to
replace peat and not just added to the soil as described in Andersen et al.
(2010). Moreover, even if the full potential of the compost is utilised, it is still
unlikely that it will be enough to drastically change the order of priority of the
scenarios.
6.3 Potential to increase sustainability by reducing food waste
The potential to reduce food waste should not only be related to the cost of
reducing waste, but also to other possible measures that lead to increased
sustainability. In order to put the problem of food waste into context, it can be
compared with other sources of carbon footprint presented in the Axfood
Sustainability Report (Axfood, 2014). Using the data in that report, the average
carbon footprint on specific services during 2012-2014 was recalculated into a
value of emissions per supermarket (of which six out of 246-259 are the stores
investigated here). This value was set to 71 ton CO
2
e/supermarket/year
originating from electricity use in the supermarkets (excluding wholesale
facilities), 3 ton CO
2
e/supermarket/year from business flights and 39 ton
CO
2
e/supermarket/year from transport of goods in the 137 company-owned
69
trucks. When these three sources of emissions were added together, they were
still lower than the average carbon footprint associated with food waste, which
was estimated to be 131 ton CO
2
e/supermarket/year (Paper III and Appendix
I). Since these are average values, they are likely to include some variation and
therefore some stores can diverge from this pattern, but since the key figure for
food waste is so much larger than the other key figures, reducing food waste
should clearly play an important role in work towards sustainable development.
In the sustainability report of the parent company of the stores investigated
(Axfood, 2014), food waste is mentioned several times, but there are no actual
goals set for reducing this waste. This could be because the company does not
want to declare exact amount for food waste, because it is difficult to measure
in comparable units such as ton or ton CO
2
e or simply because food waste
gives rise to emissions long before they reach the stores and therefore is
considered to be outside the sustainability work of the retailer. It could also be
because the company does not see the same potential in reducing this problem
as can be achieved with other problems. Just because a problem is large does
not mean that there is a large savings potential in solving it. However, the
results in Paper IV indicate that a waste reduction through a temperature
reduction (from 4°C to 2°C) and increased shelf-life in the meat department
could reduce the CF by 12 ton CO
2
e/supermarket/year. This waste reduction
could be related to the goals of reduced emissions from flights, energy use and
transport by multiplying the average emissions by the goal set (as a percentage)
to get the wanted reduction in absolute terms. However, this is a simplification,
since all goals are set in relative terms, e.g. in relation to floor area or ton
transported, so that company growth does not interfere with sustainability
goals. The goals also have different base years, so relating the goal to an
average emissions level must be considered a rough estimate, but still gives an
indication of the relative magnitude of the potential. For the supermarket
company in this thesis, a goal of 15% less emissions from business flights
corresponded to a reduction of 0.4 ton CO
2
e/supermarket/year, a 10%
reduction in transport emissions corresponded to a saving of 4 ton
CO
2
e/supermarket/year and a 25% reduction in electricity
4
consumption
correspond to savings of 18 ton CO
2
e/supermarket/year. These goals can be
related to the potential in shifting waste management method from composting
to anaerobic digestion (assuming that all waste corresponds to bananas in the
composting and anaerobic digestion scenarios in Paper V), which would
potentially save 34 ton CO
2
e/supermarket/year. It can also be related to the
potential savings of 12 ton CO
2
e/supermarket/year achievable by reducing
4
The goal in Axfood (20014) states energy consumption, but here it was assumed that this
includes a 25% reduction in electricity-related emissions.
70
storage temperature in the meat department (Paper IV). A complicating factor
is of course that the local infrastructure might not make it possible to shift
waste management to anaerobic digestion. Another is that reducing waste
through reduced storage temperature will consume more electricity, according
to the assumptions used in Paper IV, and therefore conflict with the goal of
reduced energy consumption. However, whether the temperature reduction is
achieved through better cleaning routines for the chill cabinets, as suggested by
Danielsson-Tham & Bood (2015), by using doors on vertical cabinets
(Lindberg et al., 2010) or by changing to an electricity mix with lower CF
(Paper IV), the goals of reduced energy-related emissions and reduced food
waste can both be achieved without conflict.
71
7 Conclusions
This thesis made a deeper investigation than most previous studies of food
waste in supermarkets and it provides new information about quantities of
wasted food, data quality issues, risk factors and potential net savings from
different prevention and waste management options. From this information, it
is clear that waste increases the carbon footprint of food and that waste
reduction has the potential to reduce the carbon footprint from the food supply
chain. Since there was found to be a lack of tools and definitions regarding
supermarket food waste issues, new approaches and new combinations of
methodologies were developed during the work in order to meet the objectives.
From the overall data presented, a few more general conclusions can be
drawn than are presented in the individual papers. First, the fruit and vegetable
department had the largest recorded wasted mass, with most of this recorded
waste coming from rejections. However, when this waste was measured as the
difference between delivered and sold products the figure is lower, which
reduced the significance of this problem and also the potential to reduce waste
by limiting rejections. While physical rejections may be less than supermarket
financial records indicate, the largest share of wasted mass was still found in
the fruit and vegetable department and was dominated by rejections on delivery
to the supermarket.
When the carbon footprint, instead of mass, was used to evaluate the waste,
the meat department increased in importance and the fruit and vegetables
department lost some of its dominance. Other supermarket departments with
mainly products of animal origin contributed a larger share of waste when the
carbon footprint was considered rather than the mass of waste.
Reducing the storage temperature proved to have the potential to increase
shelf-life which can lead to reduced waste. However, the way of achieving this
temperature reduction was of great importance for achieving a net saving in
terms of both carbon footprint and money. By using other alternatives than
72
increased use of the average electricity mix, a net saving for products other
than just meat would be possible. On comparing waste management options
with prevention measures, source reduction of beef waste was found to
decrease the carbon footprint much more than all valorisation measures
investigated.
For food waste that cannot be prevented, valorisation options should be
used to reduce the negative effects of food waste. This can be achieved mainly
by replacing other products in a substituted system. Target products for
efficient waste valorisation measures should therefore be foodstuffs that can be
used to replace other products or services that are expensive, resource-
demanding and/or have a high carbon footprint. When considering only carbon
footprint, bread is a good example of a high priority target for waste
valorisation due to its low water content, which allows it to carry much energy
that can be used for various purposes.
Waste reduction measures specialised for a specific food product tend to be
more environmental efficient than general waste management options, whereas
the latter instead have the ability to handle more waste with less restrictions on
quality. Since most food products have high water content, a mixed flow of
supermarket food waste will be most efficiently managed by anaerobic
digestion for production of biogas that can replace resource-demanding
products or services. However, since supermarket food waste is by nature
separated before it reaches the waste container, it has the potential to be used
differently depending on the nature of the wasted products. Therefore a
combination of prevention, economic valorisation, donation, conversion and
recovery should be practised simultaneously in order to find efficient ways to
reduce the carbon footprint of the food supply chain.
73
8 Future research
Food waste in supermarkets is a quite narrow research field, but expansion
would be desirable due to the potential to reduce both cost and environmental
impact. This thesis digs deeper than many other studies, but still just uncovers
the tip of the iceberg. Therefore it is a continuing need to advance the
knowledge frontiers. Some suggested fields that needs to be further
investigated are:
Risk factors need to quantified in terms of waste generation: Which risk
factors are relevant for supermarkets? What quantities do they generate?
and Under what circumstances?
Measures aimed at reducing food waste must be further investigated.
This includes more theoretical simulations to find promising waste
interventions, but also practical tests where the measures are evaluated
in real situations. Such theoretical and practical evaluations should
include both the costs of performing the measure and the potential waste
reduction, so a net result can be achieved.
Food waste research in supermarkets has come quite far in comparison with
that by other actors in the food supply chain. Other actors therefore have the
opportunity to learn from supermarkets in order to improve methodology,
perhaps by simply replicating the work in this thesis but with the perspective
shifted to food services, industry, household, and so on. Some suggestions are:
Continuous quantification must be performed, since waste varies widely
between different sectors, but also over time within the same sector. If
only selected periods or waste fractions are quantified, they should at
least be randomly selected so that it is not only the periods with the
lowest levels of waste that are quantified.
After establishing robust food waste quantification routines, systematic
work to reduce food waste can start. If interventions are tested without
74
sufficient quantification, monitoring will be impossible and it will be
unclear whether the measure reduces food waste at all.
75
References
Alexander, C., Smaje, C. (2008). Surplus retail food distribution: an analysis of a third sector
model. Resources, Conservation and Recycling 52, 1290-1298.
Andersen, J., Christensen, T., Scheutz, C. (2010) Substitution of peat, fertiliser and manure by
compost in hobby gardening: User surveys and case studies. Waste Management 30,2483-
2489.
Andersson, T. (2012). Från hage till mage – en studie om oundvikligt och onödigt matavfall.
Thesis 2012:02, Department of Chemical Engineering, Lund University, Lund, Sweden.
Antón, A., Torrellas, M., Montero, J.I., Ruijs, M., Vermeulen, P., Stanghellini, C. (2012)
Environmental Impact Assessment of Dutch Tomato Crop Production in a Venlo Glasshouse,
In XXVIII International Horticultural Congress on Science and Horticulture for People
(IHC2010): International Symposium, 927, 781–791.
Avfall Sverige (2006) Matavfall från restauranger, storkök och butiker - Nyckeltal med
användarhandledning, RVF report 2006:07, RVF – Svenska Renhållningsverksföreningen,
Malmö.
Axfood (2014) Sustainability program, Axfood AB, Solna.
Beretta C., Stoessl F., Baier U., Hellweg S. (2013) Quantifying food losses and the potential for
reduction in Switzerland. Waste Management 33, 764–773.
Bernstad, A., la Cour Jansen, J., 2012, A life cycle approach to the management of household
food waste – A Swedish full-scale case study, Waste Management, 31, 1879-1896.
Bernstad Saraiva Schott, A., Andersson, T. (2014) Food waste minimization from a life-cycle
perspective, Journal of Environmental Management, (article in press).
Biel, A., Bergström, K., Carlsson-Kanyama, A., Fuentes, C., Lagerberg-Fogelberg, C., Shanahan,
H., Solér, C., Grankvist, G. (2006) Environmental Information in the Food Supply System,
FOI – Swedish Defence Research Agency, Stockholm.
Björkman, J. (2015) Design av en modell för att förutsäga matsvinn, Bachelor thesis in
mathematics, Uppsala university, Uppsala. [Unpublished thesis]
Brunner P. H., Rechberger H. (2005) Practical handbook of material flow analysis, CRC Press
LLC, Boca Raton.
Buzby J.C., Farah Wells H., Axtman B., Mickey J. (2009) Supermarket loss estimates for fresh
fruits, vegetables, meat, poultry, and seafood and their use in the ERS loss-adjusted food
76
availability data. Economic Information Bulletin Number 44, United States Department of
Agriculture, Economic Research Service, Washington.
Buzby, J. C., Hyman, J., Stewart, H. & Wells, H.F. (2011). The value of retail- and costumer-
level fruit and vegetable losses in the United States. Journal of Consumer Affairs 45, 492-515.
Buzby J.C., Hyman J. (2012) Total and per capita value of food loss in the United States. Food
Policy 37, 561–570.
Cederberg, C., Flysjö, A., Sonesson, U., Sund, V., Davis, J. (2009) Greenhouse gas emissions
from Swedish consumption of meat, milk and eggs 1990 and 2005. SIK Report No 794, SIK –
the Swedish Institute for Food and Biotechnology, Gothenburg.
Danielsson-Tham, M-L., Bood, U. (2015) Det enkla är det svåra – också i butikernas kyldiskar,
Nr 7, Svensk Veterinärtidning, Stockholm.
Davis, J., Wallman, M., Sund, V., Emanuelsson, A., Cederberg, C., Sonesson, U., (2011)
Emissions of Greenhouse Gases from Production of Horticultural Products : Analysis of 17
Products Cultivated in Sweden. SIK - Institutet för livsmedel och bioteknik, Gothenburg.
DEFRA (2010) A Review of Municipal Waste Component Analyses, Department for Environment,
Food and Rural Affairs, London.
Dutch Ministry of Economic Affairs, Agriculture and Innovation (2014) Preventing food waste
and optimising residual waste streams, http://no-opportunity-
wasted.com/images/document/416.pdf, Accessed 2014-06-13
EC (1975) Council Directive 75/442/EEC of 15 July 1975 on waste, The Council of the European
Communities, Official Journal L 194, Brussels.
EC (2002). Regulation (EC) No 178/2002 of the European Parliament and of the Council of the
28 January 2002 laying down the general principles and requirements of food law,
establishing the European Food Safety Authority and laying down procedures in matters of
food safety. Official Journal of the European Communities, Brussels, Belgium.
EC (2006). Environmental Impact of Products (EIPRO): Analysis of the Life Cycle
Environmental Impacts Related to the Total Final Consumption of the EU 25, European
Commission Technical Report EUR 22284 EN.
EC (2008). Directive 2008/98/EC of the European Parliament and of the Council of 19 November
2008 on waste and repealing certain Directives. Official Journal of the European
Communities, Brussels.
EC (2010) Final Report – Preparatory Study on Food Waste Across EU27, European
Commission. DG ENV – Directorate C, Brussels.
EC (2011) Roadmap to a Resource Efficient Europe, European Commission, Brussels.
Eriksson M. (2012) Retail Food Wastage: a Case Study Approach to Quantities and Causes,
Licentiate thesis 045, Department of Energy and Technology, Swedish university of
Agricultural Science, Uppsala.
Eriksson, M., Strid, I., 2011, Food loss in the retail sector - a study of food waste in six grocery
stores. Report 035, Department of Energy and Technology, Swedish University of
Agricultural Science, Uppsala.
Eriksson M., Strid, I. (2013) Svinnreducerande åtgärder i butik - Effekter på kvantitet, ekonomi
och klimatpåverkan [Waste reducing measures in supermarkets – Effects on quantity,
77
economy and greenhouse gas emissions], Report 6594, Swedish Environmental Protection
Agency, Stockholm.
FAO (2011) Global food losses and food waste, FAO, Rome.
FAO (2012) The State of Food Insecurity in the World 2012, Economic growth is necessary but
not sufficient to accelerate reduction of hunger and malnutrition. FAO, WFP and IFAD,
Rome.
FAO (2013) Food Wastage Footprint: Impacts on natural resources, FAO, Rome.
Feeding the 5000 (2014) http://www.feeding5k.org/businesses.php, Accessed 2014-06-13.
Fehr, M., Calcado, M.D.R. & Romão, D.C. (2002). The basis of a policy för minimizing and
recycling food waste. Environmental Science & Policy 5, 247-253.
Flysjö, A. (2012) Greenhouse gas emissions in milk and dairy product chains : improving the
carbon footprint of dairy products, Aarhus University, Department of Agroecology, Science
and Technology, Aarhus.
Garnett T. (2011) Where are the best opportunities for reducing greenhouse gas emissions in the
food system (including the food chain)?, Food Policy 36, S23-S32.
Garrone P., Melacini M., Perego A. (2014) Opening the black box of waste reduction, Food
policy 46, 129-139.
Gentil E., Gallo D., Christensen T.H. (2011) Environmental evaluation of municipal waste
prevention, Waste management 31, 2371-2379.
Giuseppe A., Mario E., Cinza M. (2014) Economic benefits from food recovery at the retail stage:
An application to Italian food chains. Waste Management 34, 1306-1316.
Godfray, C., Reddington, J., Crute, I., Haddad, L., Lawrence, D., Muir, J., Pretty, J., Robinson, S.,
Thomas, S., Toulmin, C. (2010) Food Security: The Challenge of Feeding 9 Billion People,
Science 327, 812-818.
González, A.D., Frostell, B., Carlsson-Kanyama, A. (2011) Protein Efficiency Per Unit Energy
and Per Unit Greenhouse Gas Emissions: Potential Contribution of Diet Choices to Climate
Change Mitigation, Food Policy, 36, 562–570.
Griffin, M., Sobal, J., Lyson, T.A. (2009) An analysis of a community food waste stream,
Agriculture and Human Values 26, 67-81.
Gustavsson, J. & Stage, J. (2011). Retail waste of horticultural products in Sweden. Resources,
Conservation and Recycling 55, 554-556.
Göbel, C., Teitscheid, P., Ritter, G., Blumenthal, A., Friedrich, S., Frick, T., Grotstollen, L.,
Möllenbeck, C., Rottstegge, L., Pfeiffer, C., Baumkötter, D., Wetter, C., Uekötter, B.,
Burdick, B., Langen, N., Lettenmeier, M. & Rohn, H. (2012). Reducing Food Waste -
Identification of causes and courses of action in North Rhine-Westphalia. Abridged version,
University of Applied Sciences Münster, Institute for Sustainable Nutrition and Food
Production – iSuN, Münster.
Hanssen O. J. (2010) Matavfall og emballasje – hva er mulige sammenhenger?, Report OR.16.10
from the EMMA-project, Østfoldsforskning.
Hanssen O. J., Schakenda V. (2011) Nyttbart matsvinn i Norge 2011 – Analyser av status och
utvikling i matsvinn i Norge 2010-11. Rapport fra ForMat-prosjektet, Østfoldsforskning.
Hanssen, O. J., Møller, H. (2013) Food Wastage in Norway 2013: Status and Trends 2009-13,
ForMat, Østfoldsforskning.
78
Hernant, M. (2012) Presentation from the project “Reduced food wastage in grocery stores -
measures and their impact on economy and environment”, Available at
http://www.slu.se/Documents/externwebben/nj-fak/energi-och-
teknik/Matsvinn/Presentation%2020130507%20SHORT.pdf.
Hodges, R., Buzby, J., Bennet, B. (2011) Postharvest losses and waste in developed and less
developed countries: opportunities to improve reasource use, Journal of Agricultural Science
149, 37-45.
Institution of Mechanical Engineers (2013) Global Food: Waste Not, Want Not, Institution of
Mechanical Engineers Westminster, London.
ISO (2006) Environmental management – Life cycle assessment – Principles and framework, ISO
14040:2006. Swedish Standard Institute, Stockholm.
ISO (2006) Environmental management – Life cycle assessment – Requirements and guidelines,
ISO 14044:2006. Swedish Standard Institute, Stockholm.
ISO (2010) Environmental management systems – General guidelines on principles, systems and
supporting techniques, ISO 14004:2004. Swedish Standard Institute, Stockholm.
Jensen, C., Stenmarck, Å., Sörme, L. & Dunsö, O. (2011a). Matavfall 2010 från jord till bord.
ISSN 1653-8102, SMED, Swedish Meteorological and Hydrological Institute, Norrköping.
Jensen, J., Bark, P., Storhagen, NG. (2011b) Hållbara intermodala transporter av dagligvaror,
TFK 2011:5, TFK, Stockholm.
Jonsson, C. (2012) Datemarking of food products - in benefit for producers, traders and
consumers, Master thesis in food science, 2012:4, Swedish University of Agricultural Science,
Uppsala.
Lagerberg Fogelberg C., Vågsholm I., Birgersson, A. (2011) From Loss to Gain – How to Reduce
In-Store Food Waste. Report, Department of Biomedical Sciences and Veterinary Public
Health, Swedish University of Agricultural Science, Uppsala.
Laurent A., Bakas I., Clavreul J., Bernstad A., Niero M., Gentil E., Hauschild M.Z., Christensen
T.H. (2013a) Review of LCA applications to solid waste management systems – Part I:
lessons learned and perspectives. Waste Management 34, 573-588.
Laurent A., Clavreul J., Bernstad A., Bakas I., Niero M., Gentil E., Christensen T.H., Hauschild
M.Z. (2013b) Review of LCA applications to solid waste management systems – Part II:
Methodological guidance for a better practice. Waste Management 34, 589-606.
Lindbom, I., Gustavsson, J., Sundström, B. (2014) Minskat svinn i livsmedelskedjan – ett
helhetsgrepp, Slutrapport, SR 866, SIK – the Swedish Institute for Food and Biotechnology,
Gothenburg.
Lindberg, U., Axell, M., Fahlén, P. & Fransson, N., 2010, Vertical display cabinets without and
with doors - a comparison of measurements in a laboratory and in a supermarket. Proc. of
Sustainability and the Cold Chain. Cambridge.
Lebersorger S., Schneider F. (2014) Food loss rates at the food retail, influencing factors and
reasons as a basis for waste prevention measures. Waste Management 34, 1911-1919.
Lee, P. Willis, P. (2010) Waste arising in the supply of food and drink to households in the UK,
WRAP, Branbury.
79
Lee S.H., Choi K.I., Osako M., Dong J.I. (2007) Evaluation of environmental burdens caused by
changes of food waste management systems in Seoul, Korea, Science of the Total
Environment 387, 42-53.
Lundquist, J., de Fraiture, C., Molden, D. (2008) Saving Water: From Field to Fork – Curbing
Losses and Wastage in the Food Chain, SIWI policy brief, Stockholm Inyernational Water
Institute, Stockholm.
Katajajuuri, J-M., Silvennoinen, K., Hartikainen, H., Heikkilä, L., Reinikainen, A. (2014) Food
waste in the Finnish food chain, Journal of Cleaner Production, 73, 322-329.
Mattsson, L., Williams, H. (2015) Waste reduction potential of fresh fruit and vegetables at retail
stores: A case study of three supermarkets in Sweden. 2nd International Conference on Global
Food Security, October 11-14, 2015, Ithaca, USA.
Marklinder, I., Eriksson, M. (2012) Best-before date mass experiment: food storage temperatures
registered by Swedish school pupils, European Symposium on Food Safety, Warszawa.
Marklinder, I., Eriksson, M. (2015) Best-before date – food storage temperatures recorded by
Swedish students, British Food Journal, 117, 1764-1776.
Marklinder, I., Eriksson, M., Thomasson, L. (2012) Slutrapport från Forskarfredags bäst före
försök, Vetenskap & Allmänhet, Stockholm.
Martinsson, H. (2014) Food legislations effect on food waste in supermarkets, Bachelor thesis in
food science, Nr. 398, Swedish University of Agricultural Science, Uppsala.
Menikpura S.N.M., Sang-Arun J., Bengtsson M. (2013) Integrated Solid Waste management: an
approach for enhancing climate co-benefits through resource recovery, Journal of Cleaner
Production 58, 34-42.
Miljøstyrelsen (2014) Kortlægning af madaffald i servicesektoren, Undgå affald stop spild nr. 05,
Miljøstyrelsen, Copenhagen.
Ministry of the Environment and Energy (2001) Förordning (2001:512) om deponering av avfall,
Ministry of the Environment and Energy, Stockholm.
Muñoz, P., Antón, A., Nuñez, M., Paranjpe, A., Ariño, J., Castells, X., Montero, J.I., Rieradevall,
J. (2007) Comparing the Environmental Impacts of Greenhouse Versus Open-field Tomato
Production in the Mediterranean Region, In International Symposium on High Technology
for Greenhouse System Management (Greensys2007), 801, 1591–1596.
NFA (2015) Har du tivoli i kylskåpet?, http://stoppamatsvinnet.nu/, Accessed 2015-08-04.
Nature (2010) How to feed a hungry world, Editorial, Nature 266, 531-532.
Nellemann, C., MacDevette, M., Manders, T., Eickhout, B., Svihus, B. and Prins, A.G. (2009).
The environmental food crisis – the environment’s role in averting future food crises. United
Nations Environment Programme (UNEP), Norway.
Nilsson H. (2012) Integrating Sustainability in the Food Supply Chain - Two Measures to Reduce
the Food Wastage in a Swedish Retail Store, Master thesis in Sustainable Development Nr.
94, Uppsala University, Uppsala.
Papargyropoulou E., Lozano R., Steinberger J.K., Wright N., bin Ujang Z. (2014) The food waste
hierarchy as a framework for the management of food surplus and food waste, Journal of
Cleaner Production 76, 106-115.
80
Persson, M. (2015) Wastage of eggs in grocery stores - a case study of wastage and the effect of
an extended sales period, Bachelor thesis in environmental science, Thesis 2015:04,
Department of Energy and Technology, Swedish University of Agricultural Science, Uppsala.
Roy, P., Nei, D., Orikasa, T. (2009) A review of life cycle assessment (LCA) on some food
products, Journal of Food Engineering, 90, 1–10.
Rutten, M., Nowicki, P., Bogaardt, M.-J., & Aramyan, L. (2013) Reducing food waste by
households and in retail in the EU: A prioritisation using economic, land use and food
security impacts, LEI report 2013-035, LEI, part of Wageningen UR, The Hague.
Röös, E. (2012) Mat-klimat-listan version 1.0., Report 040, Department of Energy and
Technology, Swedish University of Agricultural Sciences, Uppsala.
Röös, E. (2013) Analysing the Carbon Footprint of Food: Insights for Consumer Communication,
Doctoral thesis 56, Acta Universitatis agriculturae Sueciae, Swedish University of
Agricultural Sciences, Uppsala.
Röös, E., Sundberg, C., Tidåker, P., Strid, S., Hansson, P-A. (2013) Can carbon footprint serve as
an indicator of the environmental impact of meat production?, Ecological Indicators 24, 573–
581.
Salhofer S., Obersteiner G., Schneider F., Lebersorger S. (2008) Potentials for the prevention of
municipal solid waste. Waste Management 28, 245-259.
Scherhaufer S., Schneider F., 2011, Prevention, recycling and disposal of waste bread in Austria,
Thirteenth International landfill Symposium, Cagliari, Sardinia
Schneider F. (2013a) The evolution of food donation with respect to waste prevention, Waste
Management 33, 755–763.
Schneider F. (2013b). Review of food waste prevention on an international level. Water and
Resource Management 166, 187-203.
Scholz, K. (2013) Carbon footprint of food waste, Master thesis, Department of Energy and
Technology, Swedish University of Agricultural Sciences, Uppsala.
SEPA. (2011). Nyttan med att minska livsmedelssvinnet i hela kedjan. Report 6454, Swedish
Environmental Protection Agency, Stockholm.
SEPA. (2012). Nyttan med att minska livsmedelssvinnet. Report 6527, Swedish Environmental
Protection Agency, Stockholm.
SEPA (2013) Food waste volumes in Sweden, Swedish Environmental Protection Agency,
Stockholm.
Solomon, S., Qin, D., Manning, M., Alley, R.B., Berntsen, T., Bindoff, N.L. (2007). Technical
summary. In: Solomon S, Qin D, Manning M, Chen M, Marquis M, Averyt KB, Tig-nor M,
Miller HL, editors. In Climate Change: 2007 The Physical Science Basis. Contribution of
Working Group I to the Fourth Assessment Report of the Inter-governmental Panel on
Climate Change. Cambridge University Press, Cambridge and New York.
Stenmarck, Å., Hanssen, O. J., Silvennoinen, K., Katajauuri, J-M. & Werge, M. (2011). Initiatives
on prevention of food waste in the retail and wholesale trades. Nordic Council of Ministers,
Copenhagen.
Stoessel, F., Juraske, R., Pfister, S., & Hellweg S. (2012) Life Cycle Inventory and Carbon and
Water Food Print of Fruits and Vegetables: Application to a Swiss Retailer, Environmental
Science & Technology, 46, 3253–3262.
81
Strid, I. (2012). Prioritizing retail food waste prevention – Potatoes, Tomatoes or Carambolas?
The 8th International Conference on LCA in the Agri-Food Sector, 2-4 October 2012, Rennes.
Strid, I., Eriksson, M. (2013) Valorization of meat waste from retail stores, The 6th International
Conference on Life Cycle Management in Gothenburg 2013.
Strid, I., Eriksson, M., Andersson, S., Olsson, M. (2014) Svinn av isbergssallat i
primärproduktionen och grossistledet i Sverige, Report 2014:06, Swedish Board of
Agriculture.
Strid, I., Eriksson, M. (2014) Losses in the supply chain of Swedish lettuce – wasted amounts and
their carbon footprint at primary production, whole sale and retail, The 9th International
Conference on LCA in the Agri-Food Sector, San Francisco.
Stuart, T. (2009). Waste: uncovering the global food scandal. Penguin Books, London.
Stare, M., Johansson, M., Dunsö, O., Stenmarck, Å., Sörme, L., Jensen, C. (2013) Förbättrade
matavfallsfaktorer för verksamheter, Report 117, Sveriges Meteorologiska och Hydrologiska
Institut, Norrköping.
Steinfeldt, H., Gerber, P., Wassenaar, T., Castel, V., Rosales, M. and de Haan, C. (2006).
Livestock’s long shadows: environmental issues and options. Food and Agricultural
Organization of the United Nations, Rome.
Stensgård, A., Hanssen, O.J., 2015, Food Waste in Norway 2014: Status and Trends 2009-14,
ForMat, Østfoldsforskning.
Tapper, J., Nordell, M., Torstensson, A. (2013) Svinn av frukter och grönsaker - Fallet ICA
Grossist, Master thesis in Technology Management Nr 246/2013, Department of Design
Sciences, Lund University, Lund.
Taylor, D. (2006). Demand management in agri-food supply chains: An analysis of the
characteristics and problems and a framework for improvement, The International Journal of
Logistics Management 17, 163-186.
Tesco (2014) Tesco and Society, Tesco, Cheshunt, Hertfordshire.
USEPA (2015) http://www.epa.gov/foodrecoverychallenge/, Accessed 2015-07-29.
Van Ewijk, S., Stagemann, J.A. (2015) Limitations of the waste hierarchy for achiving absolute
reduction in material throughput, Journal of Cleaner Production, (article in press).
Vander Stichele, M., van der Wal, S. & Oldenziel, J. (2006). Who reaps the fruit? Critical issues
in the fresh fruit. SOMO, Centre for Research on Multinational Corporations, Amsterdam.
Vandermeersch T., Alvarenga R.A.F., Ragaert P., Dewulf J. (2014) Environmental sustainability
assessment of food waste valorization options, Resources, Conservation and Recycling 87, 57-
64.
Ventour, L. (2008). The food we waste. WRAP, Banbury, UK.
Williams, A.G., Pell, E., Webb, J., Tribe, E., Evans, D., Moorhouse, E., Watkiss, P. (2008)
Comparative Life Cycle Assessment of Food Commodities Procured for UK Consumption
through a Diversity of Supply Chains, Project FO0103, DEFRA, London.
Williams, H., Wikström, F., Löfgren, M., 2008, A life cycle perspective on environmental effects
of customer focused packaging development, Journal of Cleaner Production, 16, 853-859.
Williams, H., Wikström, F., 2011, Environmental impact of packaging and food losses in a life
cycle perspective: a comparative analysis of five food items, Journal of Cleaner Production,
19, 43-48.
82
Willy:s. (2010). Concept hand book. Willy:s AB, Gothenburg.
WRAP (2011) Food waste in schools, WRAP, Banbury, UK.
WRAP (2015) Love Food Hate Waste, http://www.lovefoodhatewaste.com/, Accessed 2015-08-
04.
Zorpas A., Lasaridi K. (2013) Measuring waste prevention, Waste Management 33, 1047-1056.
Åhnberg, A., Strid, I. (2010). When food turns into waste – a study on practices and handling of
losses of fruit and vegetables and meat in Willys Södertälje Weda. Report 025, Department of
Energy and Technology, Swedish University of Agricultural Sciences, Uppsala.
Östergren, K., Gustavsson, J., Bos-Brouwers, H., Timmermans, T., Hansen, O-J., Møller, H.,
Anderson, G., O’Connor, C., Soethoudt, H., Quested, T., Easteal, S., Politano,A., Bellettato,
C., Canali, M., Falasconi, L., Gaiani, S., Vittuari, M., Schneider, F., Moates, G., Waldron, K.,
Redlingshöfer, B. (2014) FUSIONS Definitional Framework for Food Waste, EU FUSIONS.
83
Acknowledgements
First of all, I would like to express my gratitude to my supervisors and main
co-authors, Per-Anders Hansson and Ingrid Strid, this thesis would not exist
without your guidance and support. I would also like to thank Katharina Scholz
for co-authoring one of the papers.
During the work with this thesis, a number of people were involved in
helping me with data collection and giving me ideas and perspectives on the
work. Therefore I would like to thank Enikö Walter, Glenny Sernström, Mats
Johansson and Joel Forsman at Willy:s; Daniel Månsson, Hans Holmstedt and
Åsa Domeij at Axfood; Per-Olof Hedman at DAGAB; Mikael Hernant at the
Stockholm School of Economics; Åsa Stenmarck at IVL; Lisa Mattsson at
Karlstad University; Ingela Marklinder at Uppsala University; the members of
the SaMMa network and of course my colleagues at the Department of Energy
and Technology.
I would also like to thank Christine Jonsson, Herman Nilsson, Emilia
Sjöberg, Katharina Scholz, Jesper Björkman, Fanny Eriksson, Frida Fellenius,
Rebecca Norman, Helena Martinsson, Katalin Simon, Sanna Lindberg, Emmy
Mattisson, Magnus Persson and Sebastian Lahner for writing their Bachelor’s
or Master’s thesis about food waste and thereby providing me with important
insights into food waste-related problems.
This thesis was funded by the Swedish Retail and Wholesale Development
Council (HUR) and the Swedish Research Council for Environment,
Agricultural Sciences and Spatial Planning (FORMAS). Data were provided by
the company Axfood and its subsidiary companies Willy:s and DABAB. I
would like to thank all these organisations for providing the means to make this
thesis possible.
Last, but definitely not least, I would like to thank my friends and family for
all your help and support, but also for your patience with my sometimes
overwhelming interest in old food.
84
Appendix I. Store department level results.
This appendix contains three tables showing the sum of sold and wasted food
in terms of mass, carbon footprint and money on supermarket department level.
Waste includes recorded in-store waste and pre-store waste, and relative waste
is calculated in relation to the sum of sold and waste (Equation 2).
Table AI.1. Summarised values of mass of sold and wasted perishable food from different
departments in all six supermarkets studied during five years 2010-2014.
Department
Sold
(ton)
Waste
(ton)
Relative waste
(%)
Cheese
5 000
28
0.55
Dairy
52 000
180
0.34
Deli
5 800
91
1.5
Fruit and vegetables
42 000
2 000
4.7
Meat
6 800
88
1.3
Total
110 000
2 400
2.1
Table AI.2. Summarised values of CF associated with the sold and wasted perishable food from
different departments in all six supermarkets studied during five years 2010-2014.
Department
Sold
(ton CO
2
e)
Waste
(ton CO
2
e)
Relative waste
(%)
Cheese
44 000
240
0.54
Dairy
75 000
250
0.33
Deli
34 000
530
1.5
Fruit and vegetables
35 000
1 700
4.7
Meat
110 000
1 200
1.1
Total
300 000
3 900
1.3
Table AI.3. Summarised economic value of sold and wasted perishable food from different
departments in all six stores studied during five years 2010-2014. The value given is the cost
without value-added tax for the store to buy the food and pre-store waste is valued to this price.
Department
Sold
(MSEK)
Waste
(MSEK)
Relative waste
(%)
Cheese
319
1.6
0.50
Dairy
701
2.3
0.32
Deli
296
4.0
1.3
Fruit and vegetables
657
32
4.7
Meat
355
5.3
1.5
Total
2 330
45
1.9
85
Appendix II. Food category level results.
This appendix contains five tables, one for each department, showing the most
dominant categories in terms of wasted mass. All values represent the sum for
all six stores investigated during five years. The figures include results of sold
mass; wasted mass both in-store and pre-store; the number of articles included
in each category; average waste per article in each category; and wasted mass
in relation the sum of wasted mass and sold mass (Equation 2).
Table AII.1. All eleven categories studied, ranked in terms of recorded wasted mass, in the cheese
department
Category
Sold
(ton)
In-store
waste
(ton)
Pre-store
waste
(ton)
Number of
articles sold
(n)
Waste per
article
(kg/art)
Relative
waste
(%)
Dessert cheese
360
5.0
0.22
204
26
1.5
Hard cheese
mild/medium
1 200
4.7
0.45
117
44
0.42
Hard cheese mature
540
4.6
0.30
121
41
0.91
Hard cheese mild
970
4.0
0.46
74
60
0.46
Cheese in food
670
2.4
0.046
186
13
0.36
Sliced cheese
270
1.3
0.12
74
19
0.52
Hard cheese medium
310
1.1
0.12
57
21
0.39
Bagged cheese
220
1.2
0.021
84
14
0.53
Cheese spread
170
0.66
0.086
95
7.9
0.43
Grated cheese
230
0.65
0.008
41
16
0.28
Whey cheese
62
0.24
0.042
13
22
0.45
86
Table AII.2. The top 14 categories (out of 22) in terms of recorded waste in the dairy department
Category
Sold
(ton)
In-store
waste
(ton)
Pre-store
waste
(ton)
Number of
articles sold
(n)
Waste per
article
(kg/art)
Relative
waste
(%)
Milk
20 000
48
0.44
85
570
0.24
Chilled juice
5 700
25
0.38
162
160
0.45
Flavoured yoghurt
4 400
18
0.24
280
66
0.42
Sour milk
3 500
17
0.29
97
180
0.50
Cream products
3 400
14
0.29
160
92
0.43
Eggs
3 200
10
1.1
46
240
0.35
Juice
1 700
6.4
0.16
79
83
0.38
Non-dairy alternative
products
2 000
5.1
0.38
89
62
0.27
Food fat for spreading
1 800
4.7
0.048
62
76
0.26
Low calorie drinks
1 300
4.4
0.060
71
63
0.34
Cottage cheese
700
4.2
0.076
62
69
0.61
Yoghurt
1 700
3.2
0.13
23
150
0.20
Food fat for baking
1 500
3.2
0.043
36
89
0.22
Chilled desserts
370
2.4
0.086
67
37
0.68
Table AII.3. The top 14 categories (out of 16) in terms of recorded waste in the deli department
Category
Sold
(ton)
In-store
waste
(ton)
Pre-store
waste
(ton)
Number of
articles sold
(n)
Waste per
article
(kg/art)
Relative
waste
(%)
Barbecue sausages
1 100
17
4.4
201
100
1.9
Cold cuts
980
14
4.7
506
37
1.9
Wiener sausages
470
10
1.9
78
150
2.5
Salted or smoked deli
200
6.3
3.4
59
170
4.6
Thick sausages
960
4.2
1.4
61
91
0.6
Meatballs
610
4.6
0.46
28
180
0.8
Pâtés
290
3.6
1.1
79
61
1.6
Smoked pork loin
270
2.9
0.43
15
220
1.2
Luncheon meat
110
2.3
0.54
64
44
2.5
Blood pudding
250
1.5
1.3
18
160
1.1
Bacon
460
2.2
0.33
47
54
0.5
Head-cheese
30
1.0
0.24
22
58
4.0
Deli – no cold storage
40
0.32
0.15
49
10
1.1
Pickled food
5.8
0.38
0.34
28
15
6.7
87
Table AII.4. All nine categories studied, ranked in terms of recorded wasted mass, in the fresh
fruit and vegetables department
Category
Sold
(ton)
In-store
waste
(ton)
Pre-store
waste
(ton)
Number of
articles sold
(n)
Waste per
article
(kg/art)
Relative
waste
(%)
Everyday vegetables
13 000
150
550
228
3.1
5.2
Everyday fruits
11 000
71
500
111
5.2
4.8
Luxury fruits
4 100
60
230
98
2.9
6.6
Luxury vegetables
1 500
34
90
91
1.3
7.4
Herbs
4 000
42
73
137
0.8
2.8
Potatoes
7 300
57
53
33
3.3
1.5
Exotic fruits
680
24
15
46
0.86
5.5
Berries
190
7.7
18
21
1.2
12
Pre-cut lettuce
4.8
0.07
0
29
0.0023
1.4
Table AII.5. The top 14 categories (out of 30) in terms of recorded waste in the meat department
Category
Sold
(ton)
In-store
waste
(ton)
Pre-store
waste
(ton)
Number of
articles sold
(n)
Waste per
article
(kg/art)
Relative
waste
(%)
Swedish pork
804
20
0.69
177
120
2.5
Swedish poultry
1 200
17
0.18
133
130
1.4
Swedish beef
390
12
0.057
113
110
3.0
Swedish minced meat
1 800
8.6
0.051
62
140
0.47
European beef
310
5.1
0.040
32
160
1.6
Imported pork (case
ready packed)
220
4.4
0.088
24
190
2.0
Imported pork
600
3.6
0.018
29
130
0.60
Imported minced meat
650
2.7
0.002
10
260
0.42
Raw sausages
11
1.9
0.004
11
170
15
Imported lamb
87
1.2
0.16
62
22
1.6
South American beef
89
1.2
0.011
36
34
1.4
Swedish veal
14
1.2
0.012
21
56
7.7
Chitterlings
7.5
1.0
0.003
7.0
140
12
Christmas ham
240
0.77
0.18
38
25
0.40
88
Appendix III. Food product level results
This appendix contains five tables, one for each department, showing the most
dominant products in terms of wasted mass. The product level is a created level
of aggregation that in some cases can equal a single article and sometimes a
whole food category. The level is created to display results of products such as
bananas, apples and meatballs, in order to make the results comparable with
those in other studies of supermarkets with other articles and food categories.
The figures include results on the number of articles included in each
product; the wasted mass and the carbon footprint of this waste for all six
stores during five years; the share of waste (in terms of mass and CF
respectively) in each department; and the wasted mass in relation the sum of
wasted mass and sold mass (Equation 2).
Table AIII.1. The top 20 products in terms of wasted mass in the cheese department
Product
Number
of
articles
(n)
Wasted
mass
(ton)
Waste CF
(ton CO
2
)
Share of
department
wasted mass
(%)
Share of
department
wasted CF
(%)
Relative
wasted
mass
(%)
Gouda
46
2.5
24
9.0
10
0.30
Herrgård cheese
32
2.3
22
8.3
9.2
0.77
Hushålls cheese
53
2.3
21
8.1
8.8
0.32
Präst cheese
44
2.0
19
7.3
8.1
0.45
Brie
32
1.9
16
6.9
6.6
1.0
Soft cream cheese
83
1.2
6.6
4.3
2.8
0.52
Mozzarella
28
1.1
7.9
4.1
3.3
0.75
Cheddar
33
1.1
10
3.9
4.3
0.94
Edam cheese
40
1.0
9.1
3.6
3.8
0.50
Grevé cheese
24
1.0
9.4
3.6
4.0
0.40
Salad cheese/Feta
77
0.86
6.1
3.1
2.6
0.28
Cheese spread
94
0.74
2.9
2.7
1.2
0.44
Gorgonzola
12
0.47
3.8
1.7
1.6
2.0
Blue cheese
24
0.40
3.3
1.4
1.4
0.80
Billinge cheese
8.0
0.35
3.4
1.3
1.4
1.0
Jarlsberg cheese
10
0.34
3.3
1.2
1.4
1.1
Svecia cheese
7.0
0.32
3.1
1.1
1.3
2.3
Whey cheese
13
0.28
0.45
1.0
0.19
0.45
Port Salut cheese
10
0.27
2.5
1.0
1.1
0.40
Emmental
9.0
0.15
1.4
0.53
0.60
0.61
89
Table AIII.2. The top 20 products in terms of wasted mass in the dairy department
Product
Number
of
articles
(n)
Wasted
mass
(ton)
Waste CF
(ton CO
2
)
Share of
department
wasted mass
(%)
Share of
department
wasted CF
(%)
Relative
wasted
mass
(%)
Semi-skimmed milk
37
23
22
13
9.0
0.18
Flavoured yoghurt
309
20
24
11
10
0.42
Skimmed milk
20
15
14
8.7
5.8
0.47
Orange juice
84
13
8.0
7.3
3.3
0.33
Whole milk
26
12
13
6.6
5.3
0.21
Sour milk
60
11
12
6.4
4.8
0.37
Eggs
44
11
16
6.3
6.6
0.35
Apple juice
55
9.3
5.7
5.3
2.3
0.59
Cream
79
8.2
36
4.6
15
0.45
Butter blends
109
8.1
32
4.6
13
0.24
Flavoured sour milk
43
7.5
7.9
4.2
3.2
1.2
Cottage cheese
73
4.5
14
2.3
5.6
0.62
Plain yoghurt
24
3.7
4.5
2.1
1.9
0.21
Crème fraiche
69
3.2
11
1.8
4.6
0.44
Tropical juice
18
2.4
1.5
1.3
0.60
0.38
Strained yoghurt
22
2.2
5.5
1.3
2.3
0.31
Sour cream
15
2.1
5.8
1.2
2.4
0.50
Cranberry juice
8
1.1
0.37
0.63
0.15
0.78
Smoothies
42
1.0
0.68
0.58
0.28
1.8
Rice milk
21
0.90
1.1
0.51
0.45
0.43
90
Table AIII.3. The top 20 products in terms of wasted mass in the deli department
Product
Number
of
articles
(n)
Wasted
mass
(ton)
Waste CF
(ton CO
2
)
Share of
department
wasted mass
(%)
Share of
department
wasted CF
(%)
Relative
wasted
mass
(%)
Barbecue
sausage
67
13
59
14
11
1.8
Wiener sausage
52
8.8
46
10
8.7
2.7
Smoked ham
90
5.0
28
5.5
5.3
1.3
Meatballs
19
4.5
46
4.9
8.8
0.75
Salted pork
20
4.2
24
4.7
4.6
4.0
Liver pâté
63
4.0
3.8
4.5
0.72
1.4
Cooked ham
40
4.0
22
4.4
4.3
3.6
Falun sausage
22
3.7
19
4.1
3.7
0.50
Smoked pork
loin
15
3.3
18
3.7
3.5
1.2
Blood pudding
18
2.8
1.0
3.1
0.19
1.1
Chorizo
45
2.7
18
3.0
3.4
1.1
Prins sausage
22
2.6
13
2.9
2.5
2.0
Bacon
49
2.5
14
2.8
2.8
0.54
Smoked pork
shoulder
10
2.3
13
2.5
2.4
8.0
Salami
121
2.2
24
2.5
4.5
1.4
Medwurst
37
1.5
7.9
1.7
1.5
2.1
Head cheese
22
1.3
7.7
1.4
1.5
4.0
Cured ham
54
1.1
8.4
1.3
1.6
1.8
Bratwurst
14
1.0
8.8
1.1
1.7
2.6
Mortadella
20
0.69
4.2
0.76
0.80
1.1
91
Table AIII.4. The top 20 products in terms of wasted mass in the fresh fruit and vegetables
department
Product
Number
of
articles
(n)
Wasted
mass
(ton)
Waste CF
(ton CO
2
)
Share of
department
wasted mass
(%)
Share of
department
wasted CF
(%)
Relative
wasted
mass
(%)
Tomatoes
33
215
330
10
18
6.8
Bananas
9
210
231
9.8
13
5.7
Lettuce
57
183
99
8.5
5.4
9.7
Oranges
6
137
85
6.3
4.6
5.6
Peppers
19
134
310
6.2
17
9.4
Apples
59
120
47
5.6
2.5
3.6
Clementines
5
117
78
5.4
4.3
7.3
Potatoes
33
115
14
5.4
0.75
1.6
Melons
34
97
90
4.5
4.9
5.7
Cucumbers
15
69
65
3.2
3.6
3.9
Grapes
14
66
38
3.1
2.1
9.2
Nectarines
5
63
37
2.9
2.0
8.8
Pears
31
57
23
2.7
1.3
5.4
Mushrooms
17
55
22
2.5
1.2
12
Onions
33
48
14
2.2
0.78
1.8
Avocadoes
5
43
24
2.0
1.3
5.7
Carrots
14
42
5.5
2.0
0.30
2.1
Herbs in pots
39
36
36
1.7
2.0
12
Lemons
4
31
21
1.4
1.1
3.3
92
Table AIII.5. The top 20 products in terms of wasted mass in the meat department
Product
Number
of
articles
(n)
Wasted
mass
(ton)
Waste CF
(ton CO
2
)
Share of
department
wasted mass
(%)
Share of
department
wasted CF
(%)
Relative
wasted
mass
(%)
Pork chops
57
7.8
47
8.9
4.0
1.3
Minced beef
32
7.1
200
8.1
17
0.33
Pork leg
42
5.8
33
6.6
2.8
1.7
Spareribs
41
4.7
17
5.4
1.5
3.5
Grilled chicken
12
4.3
9.5
4.9
0.81
6.0
Chuck steak
19
4.3
120
4.9
10
2.3
Chicken leg
43
4.2
9.8
4.7
0.83
1.8
Beef steak
48
4.1
130
4.7
11
3.9
Ham
26
4.1
25
4.7
2.1
3.5
Chicken breast
42
4.1
11
4.6
0.97
1.4
Minute beef
13
3.0
88
3.5
7.5
1.7
Roast beef
29
2.9
86
3.3
7.3
2.3
Chicken whole
13
2.9
6.2
3.3
0.53
0.53
Mixed minced meat
25
2.7
46
3.0
4.0
0.81
Entrecôte
37
2.2
67
2.5
5.7
2.1
Pork shoulder
14
2.0
13
2.3
1.1
6.2
Raw pork sausage
4
1.6
13
1.8
1.1
16
Pork tenderloin
33
1.5
9.3
1.7
0.80
0.35
Minced pork
15
1.0
5.9
1.1
0.50
2.4
Lamb steak
22
0.90
17
1.0
1.4
1.0
93
Appendix IV. Article level results
This appendix contains eight tables, two for each department except FFV,
displaying the most dominant products in terms of wasted mass and relative
waste, respectively. All values represent the average waste per year and store
and include both recorded in-store waste and pre-store waste. Relative waste is
calculated using wasted mass in relation the sum of wasted mass and sold mass
(Equation 2).
Fresh fruit and vegetables are excluded from this appendix, since data on
pre-store waste were not available on article level. This is due to the separate
article number systems used by the stores and the supplier.
The first four tables show results for wasted mass and wasted carbon
footprint per store and year, relative wasted mass and the aggregated share of
the departments wasted CF. The latter four tables show the sum for all six
supermarkets during five years of wasted mass, sold mass and relative wasted
mass.
Table AIV.1. The 10 articles making the highest contribution to the CF associated with waste in
the cheese department
Article (% fat content)
Wasted
mass
(kg/store/yr)
Relative
wasted mass
(%)
Wasted CF
(kg CO
2
e/store/yr)
Aggregated share of
department wasted CF
(%)
Präst 35%
18
0.6
170
2.1
Brie 32%
20
1.5
160
4.2
Herrgård 28% Mild
14
1.1
130
5.9
Gouda 28%
12
0.3
120
7.3
Herrgård 28%
11
2.8
110
8.7
Gouda slices 27%
10
0.5
95
9.9
Cheddar whiskey 32%
10
5.7
94
11.1
Gouda 28%
10
1.1
94
12.3
Hushålls cheese 17%
10
1.9
86
13.4
Hushålls cheese 17%
10
1.6
85
14.4
94
Table AIV.2. The 10 articles making the highest contribution to the CF associated with waste in
the dairy department
Article
Wasted
mass
(kg/store/yr)
Relative
wasted mass
(%)
Wasted CF
(kg CO
2
e/store/yr)
Aggregated share of
department wasted CF
(%)
Butter 75%
33
0.21
220
2.6
Cream 5dl 40%
31
0.75
160
4.6
Cream
25
0.60
130
6.2
Eggs 24-pack
86
0.32
130
7.7
Skim milk 0.5%
140
0.42
130
9.3
Medium fat milk 1.5%
120
0.20
110
10.6
Eggs 24-pack
69
0.42
100
11.8
Cream 40%
19
0.67
100
13.1
Medium fat milk 1.5%
99
0.11
95
14.2
Medium fat milk 1.5%
99
0.18
95
15.4
Table AIV.3. The 10 articles making the highest contribution to the CF associated with waste in
the deli department
Article
Wasted
mass
(kg/store/yr)
Relative
wasted
mass
(%)
Wasted CF
(kg CO
2
e/store/yr)
Aggregated share of
department wasted CF
(%)
Mamas meatballs
45
0.6
560
3.2
Family hotdogs
92
5.2
480
6.0
Hotdogs with skin
63
2.1
330
7.9
Salted rump steak
11
25
280
9.5
Meatballs
19
0.5
230
10.8
Meatballs
19
0.4
230
12.1
Prins sausage
43
4.6
220
13.4
Hot barbecue sausage
69
1.3
220
14.6
Barbecue sausage thick
37
5.5
200
15.8
Barbecue sausage thin
36
7.6
190
16.8
95
Table AIV.4. The 10 articles making the highest contribution to the CF associated with waste in
the meat department
Article
Wasted
mass
(kg/store/yr)
Relative
wasted
mass
(%)
Wasted CF
(kg CO
2
e/store/yr)
Aggregated share of
department wasted CF
(%)
Irish minced beef 20%
52
0.34
1 500
3.9
Imported minute steak
50
1.5
1 500
7.6
Minced beef 20%
34
0.36
1 000
10.1
Stew beef
33
2.6
950
12.5
Chuck steak rib
29
3.7
830
14.7
Minced beef 10-12%
21
0.22
610
16.2
Imported stew beef
21
5.5
600
17.8
Minute steak 9-12 slices
21
1.6
590
19.3
Sirloin steak
14
4.2
580
20.8
Minced veal 17%
20
4.5
570
22.2
Table AIV.5. The 10 articles in the cheese department with the highest relative waste in terms of
mass, given for six stores during five years
Article
Sold mass (kg)
Wasted mass (kg)
Relative wasted mass (%)
Almkäse 28%
10
24
72
Bacon-flavoured cheeseballs
17
8
31
Raclette cheese
62
23
27
Jarlsberg 28%
220
78
26
Sörgård cheese 31%
24
8
24
Grated blue cheese 30%
139
38
22
Sour cream-flavoured cheeseballs
19
5
21
Goats cheese 23%
57
15
21
Garlic cheese 29%
86
22
21
Emmental
45
12
20
96
Table AIV.6. The 10 articles in the dairy department with the highest relative waste in terms of
mass, given for six stores during five years
Article
Sold mass
(kg)
Wasted mass
(kg)
Relative wasted mass
(%)
Raw eggs (deshelled)
21
24
54
Light crème fraiche 15%
10
6
38
Custard
73
43
37
Lactose-free milk 2%
28
14
33
Gourmet cottage cheese 10%
30
14
32
Milk drink ‘Bone Health’ 1%
600
260
30
Organic raspberry-flavoured sour
milk
92
34
27
Yoghurt with cottage cheese
160
47
23
Milk drink ‘Immune’ 1%
890
250
22
Blueberry juice
570
160
22
Table AIV.7. The 10 articles in the deli department with the highest relative waste in terms of
mass, given for six stores during five years
Article
Sold mass
(kg)
Wasted mass
(kg)
Relative wasted mass
(%)
Head cheese
16
26
62
Everyday cold cuts
14
19
57
Friday luxury cold cuts
13
16
55
Sliced roast beef
15
12
46
Barbecue sausage
260
210
45
Barbecue sausage with cheese and
bacon
210
170
45
Smoked lamb leg
250
200
45
Boiled ham with mustard
18
14
44
Liver pate
130
94
42
Crumbed ham
15
10
39
97
Table AIV.8. The 10 articles in the meat department with the highest relative waste in terms of
mass, given for six stores during five years
Article
Sold mass (kg)
Wasted mass (kg)
Relative wasted mass (%)
Minced beef
15
33
69
Pork
90
100
53
Pork tenderloin with bacon
120
130
52
Pork skewers
39
36
48
Spicy chicken fillet
130
110
47
Pork liver
52
35
41
Sliced Angus beef
24
15
38
Grilled warm ribs
130
79
38
Thai spiced pork
190
110
37
Mini beef fillets for grilling
17
9.1
35
