Journal of Cleaner Production

Manuscript Draft
Manuscript Number: JCLEPRO-D-14-02124
Title: Climate impact of food waste management options in the waste hierarchy
Article Type: Original Research Paper
Abstract: Food waste is a problem that has economic, environmental and social implications, making it
both important and complex. In order to prioritise food waste management in supermarkets, the waste
hierarchy and supporting food waste hierarchies should provide enough information to choose the
waste management method. An investigation of the climate impact of different waste management
options offers a limited perspective, but is still important for validating generations from the waste
hierarchy in a local context. This study aimed to compare the effect on greenhouse gas emission of
different food waste management scenarios representing different levels in the waste hierarchy. A life
cycle assessment was therefore performed for six waste management scenarios (landfill, incineration,
composting, anaerobic digestion, animal feed and donations) using five food products (bananas,
chicken, lettuce, beef and bread) as examples when treated as individual waste streams. All five waste
streams confirmed that the established waste hierarchy is a useful but approximate tool in prioritising
the various options, since landfill proved to be the worst option and donation, anaerobic digestion and
incineration with energy recovery were the best options. The greatest potential for reducing
greenhouse gas emissions was found to be the bread waste stream, since bread is an energy-rich
product with a comparatively low carbon footprint, increasing the possibilities for replacing fossilenergy carriers. Lettuce with its high water content was found to have the least potential to reduce
greenhouse gas emissions when the waste management method was changed.

Manuscript
Click here to view linked References

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

1. Introduction
Food wastage is a problem of increasing concern with a rising awareness of the extent of the
problem in recent years (Stuart, 2009; Nellemann, 2009; FAO, 2011; FAO, 2012). This concern has
been raised in Sweden as well, and the Swedish government has therefore set goals to reduce the
amount of waste and increase the organic treatment of food waste (SEPA, 2013). These efforts fit
well with the European Waste Framework Directive (WFD) whose waste hierarchy (EC, 2008) gives an
order of priority in waste prevention and management. The WFD also obliges member states to
encourage options that deliver the best overall environmental outcome taking a lifecycle approach,
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 the climate impact alone
offers a very limited version of the overall environmental outcome, but is no more or less
appropriate than any other environmental impact categories.
In the case of food waste, the environmental choice of waste management system from a life-cycle
perspective follows the hierarchy closely in many cases (Laurent et al., 2013a), However, since each
waste management system is dependent on a local context, the waste hierarchy must still be seen as
a rough generalisation. An actual investigation of each local context is necessary to fulfil the
obligation in the WFD.
Early versions of the waste hierarchy have been part of European policy since the 1970s (EC, 1975)
and even though it has been developed (EC, 2008), it is still a very general guideline for all waste.
More food waste-specific guidelines have therefore been issued. Examples of such systems are the
Moerman ladder in Holland (Dutch Ministry of Economic Affairs, Agriculture and Innovation, 2014),
the Food Recovery Hierarchy in the United States (USEPA, 2014), and the Food Waste Pyramid in the
United Kingdom (Feeding the 5000, 2014). All these systems prioritise prevention since the options of
waste management 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., 2013), which might be due to the methodical difficulties of measuring
something that is not there (Zorpas & Lasaridi, 2013). Priority is also given to donations to people in
need, even though this is limited by the fact, pointed out by Papargyropoulou et al. (2014), that food
waste can only be donated to charity if it is still considered to be a food surplus fit for human
consumption. Since the requirements of food hygiene or biosecurity increase the higher the level in
the waste hierarchy, there will be less chance that the whole waste flow will be sufficient for the
same type of waste management. This creates a need for more complex systems where a food waste
flow is developed and used for higher-priority waste treatments and the rest is treated with a lower
priority, but general method (Vandermeersch et al., 2014). To analyse the potential of separating the
food waste stream, instead of looking at it in its homogeneity, an approach can be taken such as the
one used by Vandermeersch et al. (2014) with individual waste streams.
Of the investigated waste management methods regarding food waste or organic waste including
food waste, the majority of the studies describe and sometimes compare landfill, incineration,
composting and anaerobic digestion (Bernstad & la Cour Jansen, 2012; Laurent et al., 2013a; 2013b)
which are all found in the less prioritised part of the waste hierarchy. Some studies also include
animal feed in the comparison (e.g. Lee et al., 2007; Vandermeersch et al., 2014), but no studies have
been found that include comparisons with the highest levels in the food waste hierarchies, such as
1/15

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

donations and prevention. One study, however, describes the environmental benefits of preventing
food waste (Gentil, 2011) but this study just assumes that there could be a 20 % reduction in a food
waste stream, but does not specify how this reduction could be achieved or the cost of doing so.
Another prevention study, Salhofer et al. (2008), uses the term prevention as being equal to
donation, but does not quantify the actual potential in this measure. However, Schneider (2013)
values donated food by its emissions during production, instead of by the produce that could be
replaced. The lack of studies quantifying higher levels of the waste hierarchy with a methodology
comparable to the lower levels makes it hard to evaluate the actual environmental benefits of
donations and prevention in relation to other waste management options. Without this extended
study, the life-cycle perspective described in the WFD will not actually be considered when selecting
waste management options.
In the large number of articles and reports reviewed by Laurent et al. (2013a; 2013b), there is a
pattern among studies comparing different waste management alternatives. The least favourable
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) conclude that
the local infrastructure is essential for the result, which makes it harder to generalise results.
In the local context of Uppsala in Sweden, supermarkets are not included in the municipal waste
monopoly. This means that they are free to use any contractor to take care of their waste. According
to Eriksson (2012) and Nilsson (2012), this means composting at the local composting facility.
Swedish supermarket food waste treatment favours the existing local infrastructure, which often
includes an incineration plant since this is the most common waste treatment method in Sweden
(SEPA, 2012a; 2012b). This is despite the environmental goal of 50 % of wasted food from
supermarkets, households, canteens and restaurants being biologically treated with the recovery of
plant nutrients by 2018 (SEPA, 2012a). The environmental goals also state that food waste in the
whole Swedish food supply chain should be reduced by 20 % between 2010 and 2020 (SEPA, 2013).
For supermarkets to achieve these environmental goals, there is a need for a change in waste
management methods. The question arises, however, of what to choose and what environmental
benefits could be achieved by different waste management alternatives. The aim of this study was
therefore to compare the outcome, with regard to greenhouse gas emissions, of different food waste
management scenarios available to supermarkets in Uppsala. The objective was to provide more
detailed knowledge about the quantity of emissions avoided when applying a more prioritised step in
the waste hierarchy in the management of food waste.

2. Materials and methods

This study used life cycle assessment methodology to calculate the global warming potential
connected to different waste management scenarios applied to five different types of food. The
functional unit is the removal of 1 kg of food waste (including packaging) from the supermarket.

2.1. Study area

This study attempted to use specific data for different waste management scenarios that are likely to
be used to handle food waste from supermarkets in the municipality of Uppsala in Sweden. Uppsala
was selected partly because detailed waste data that is normally difficult to collect was already
known from previous projects, and partly because the infrastructure for all scenarios (landfill,
2/15

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

composting, incineration, anaerobic digestion, animal feed and donations) already existed and could
be used to collect site specific data. Several of these waste management facilities have also been
investigated in previous studies (i.e. Bjrklund, 1998; Nilsson, 2012; Fors, 2013) for which there were
site-specific data.
The fundamental assumption in each scenario was that a retail company uses a waste management
method for all four of its branches located in one city. Two of these outlets were investigated by
Eriksson (2012) as regards routines, wasted mass and the composition of the waste. One of these
stores was larger and on the outskirts of the city in a shopping centre with other stores in the same
building, while the other store was in a more residential area. The remaining two supermarkets were
assumed to correspond to the investigated stores: the outlying store corresponding to the outlying
store investigated and the residential store to the residential store investigated. Calculations of
transport-related emissions were based on the distances to each of the four supermarkets, and more
waste was located in the two stores further from the city centre.

2.2. Properties of the investigated food products

Since food waste consists of a mix of food with a wide variety of properties and characteristics, it is
difficult to consider the waste as an homogenous whole. Therefore five different products were
chosen as examples due to their large contribution to total supermarket waste, either in terms of
mass (bread, bananas, lettuce and grilled chicken) or greenhouse gas emissions (beef) (Eriksson,
2012; Scholz, 2013). The selected food products also represented a different energy and water
content and different environmental burdens in the life cycle from field to retail. The products
selected were bananas, iceberg lettuce, grilled chicken, stewing beef and wheat bread, all produced
in Sweden except for the bananas which were grown in Costa Rica. The water, energy, ash and
protein content of each food product were taken from the Swedish national food database (SNFA,
2013). Greenhouse gas emissions for the life cycle of each product from field to retail were assumed
to correspond to the studies listed in Table 1, which also gives the specific waste of each product.

2.3. Description of scenarios

Six waste management scenarios were selected to represent different levels in the waste hierarchy,
but also what would be possible for supermarkets to use in Uppsala with the infrastructure already in
place. Since variations exist in the waste hierarchy that are more food-waste specific, Table 2
illustrates how the scenarios represent different levels in the food waste hierarchies from the USA,
UK and Holland.
Landfill is used for waste management in the first scenario, but as a theoretical example since the
dumping of organic waste is illegal in Sweden. Even though this is not a legal scenario for the
investigated stores, there is an infrastructure in place and landfill is commonly used for food waste
management in a global context. Therefore it provides an important perspective for the other
scenarios. In this scenario food waste is transported in a container to the waste management facility
in Hovgrden outside Uppsala, where it is dumped, compacted by machines and covered when full.
When organic waste is dumped, microbes digest carbohydrates, fats, proteins and other available
carbon sources. If there is a lack of oxygen, the digestion will be anaerobic and create methane. Since
methane is a potent greenhouse gas, some landfills collect and burn the methane, but in this
scenario no methane is collected.
3/15

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

Incineration with energy recovery is the second scenario investigated. The incineration plant located
in Uppsala is one of the largest of its kind in Sweden and incinerates waste, peat, wood and oil as fuel
to produce mainly district heating but also electricity, steam and district cooling (Vattenfall, 2013).
Since the plant normally incinerates about 340,000 tonnes of waste per year, it is assumed that the
plant will be sufficient to handle approximately 250 tonnes of food waste per year extra without any
adjustments. The marginal fuel in this plant, fuel oil, is used during cold winter periods, but since oil
is used in another boiler and then waste and peat, it is assumed that waste only replaces peat in a
system expansion.
Composting is used for waste management in the third scenario. It is based on the present situation
of waste management, as described in Nilsson (2012) for one of the four supermarkets. In this
scenario the sorted food is discarded in a garbage container at the supermarket and driven in the
container by truck to the Hovgrden compost facility outside Uppsala. The food waste is composted
in windrow composts and the compost outcome used internally as restoring soil in landfills. This
restoring soil is assumed not to replace any other product or service.
Anaerobic digestion for the production of biogas and fertiliser is used for waste management in the
fourth scenario in which the local biogas plant on the outskirts of Uppsala was used as a model. Since
the plant has a lack of substrate, there is no marginal effect of increasing production with waste from
the four stores, which corresponds to approximately 1 % of current production. There is also a
shortage of biogas, which allows an assumption that increased production would mean more buses
could be driven on methane rather than diesel. The stumbling block in the process is finding farmers
prepared to use the digestate as a diluted fertiliser. Therefore this by-product will need to be
transported to farms even further away than it is today.
Feeding animals with food waste is the fifth waste management scenario. Feeding animal byproducts such as food waste to animals is highly restricted in Sweden, as it is in the rest of EU (EC,
2009) Therefore, a store that wants to use this alternative for waste management will need to be a
registered feed producer, separate all animal products from those with a vegetative origin and
probably unpack food items that are in packages. This extra work makes this scenario less likely, but
it is assumed that this is all done and that the extra work does not produce greenhouse gas
emissions. Due to the risk of feeding production animals with animal waste, this scenario is illegal for
grilled chicken and beef, but these products were included anyway as a theoretical calculation. Since
there is an existing return system for bread from specific producers, where the producer uses the
delivery trucks to bring back unsold bread and then uses this bread to feed pigs, it is assumed that
this return system could also be used for other food items. Even though the truck would otherwise
be driven back to the supplier empty, the full emissions were still allocated to the transport of
wasted food. When fed to pigs the food waste could replace other feed. In this scenario all vegetable
food (banana, salad and bread) was assumed to replace oat (based on energy content) in the system
expansion and the animal products (chicken and beef) were assumed to replace soy (based on
protein content).
Donation of food waste to charity is used as the waste management option in the sixth scenario.
Some, but not all food is discarded because supermarket staff consider it to be unsellable even
though it is still edible. It could be donated to charity, but there are also food items that are
discarded due to a broken cold chain, passed use-by date, broken packaging or other reasons that
4/15

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

make it impossible for the store to guarantee food safety. Instead it was assumed that the discarded
items were handled according to the description in the compost scenario. The food was sorted in
store and placed in re-used cardboard boxes (secondary packages used for incoming deliveries,
which would otherwise be thrown away in store) in a cooling room awaiting transportation from the
four stores to the charity. 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 variation in food items that could be replaced by donated products, but here the same
assumption is made as by Eriksson & Strid (2013) that all donated food replaces bread based on
energy content. The reason why bread is 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
bread does not require preparation like other cheap and energy-rich products such as pasta and
potatoes. This is an uncertain assumption, however, and in the sensitivity analysis the extent to
which the results depended on a few assumptions was tested by making the same calculation
between the system expansion and the other assumptions. One sensitivity analysis is based on the
assumption that people in need would otherwise eat the same food as the kinds of food donated,
which corresponds to a fairly average Swedish diet. The last comparison is made on the assumption
that people in need have nothing other than the donated food to eat and therefore the food does
not substitute any other product. The last variation in which people will go hungry if they have no
donated food corresponds best with the description from the charity in the current situation.
Replacement with bread is used as the main scenario because it is assumed that it represents reality
for a larger group of people in need then those in the greatest need. Since a person can starve for
only a limited time, it is assumed that even starving people will eventually eat something.

2.3.1. Landfill
In this scenario the food waste was collected in a container for organic waste, but without packaging
being removed, so that the fraction was not completely biodegradable. It was assumed that the
container was collected every second week and for the calculations in NTMcalc (2014) the
transportation vehicle was assumed to be a small diesel truck (<7.5 tonne) with a euro 3 engine and a
maximum loading capacity of 5 tonnes. The emissions per transported tonne were set to 0.27 kg
CO2/tkm on urban roads with a traffic situation categorised as stop-go. The small truck was assumed
to drive with an empty 1.5-tonne container to the store and then back with another container loaded
with two weeks mean food waste. The mean waste collection was assumed to be 2050 kg for the
two smaller stores and 2736 kg per store for the two larger supermarkets in outlying locations (based
on Eriksson (2012)). Emissions per transported kg of food waste were calculated using the sum of
tkm divided by the total amount of transported food. The average distance from the stores to the
landfill was 25.4 km. The transport distance was set using a road description for the fastest route
(not necessarily the shortest) in Google maps (2014).
At the landfill the machinery cost for maintaining and compacting the landfill were assumed to be
14g CO2e/kg food waste, half of that reported by Nilsson (2012) as being used by the same facility to
turn the composting windrows. Methane production was calculated in accordance with equation 2 as
described in section 2.2.4 for the anaerobic digestion scenario. Since all food products contain easily
available nutrients, it was assumed that half of the carbon would be converted to methane and the
rest oxidised into carbon dioxide (Bjrklund, 1998). Of the methane produced, it was assumed that

5/15

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

15 % would be captured in landfill and oxidised into carbon dioxide in accordance with Bjrklund
(1998)

2.3.2. Composting
Nilsson (2012) calculated emissions from turning composting windrows to be 28 g CO2/kg composted
waste, which was adopted for this study. The transport of food waste is the same as in the landfill
scenario since the stores and the waste facility are the same.

2.3.3. Incineration with energy recovery

In this scenario the food waste was collected in a container for organic waste and transported in a
small truck, as in the landfill scenario. The difference is that the average distance from the stores to
the incineration plant was 4.4 km.
In the plant the food waste was assumed to substitute fossil peat with a specific emission factor of
386.3 g CO2e/kWh (Gode et al., 2011 cited in Fors, 2013). The amount of peat substituted was
assumed to be equal to the energy content of the food waste, including packaging. The lower heating
value (LHV) was used for both peat (11 MJ/kg (Strmberg & Herstad Svrd, 2012)) and food waste.
LHV for food waste was calculated by using equation 1 (Alvarez, 2010) where the higher heating
value (HHW) and water content (W) were taken from Table 1. Since the incineration plant uses flue
gas condensation to extract heat from moisture in the flue gases, the energy recovered from the
burned food should actually be higher than the LHV, but no higher than the HHV (Spngberg, 2014).
To address this, both HHV and LHV are presented in a sensitivity analysis:
Equation 1: LHV  1    2.447 W

2.3.4. Anaerobic digestion

To calculate the potential production of methane from each type of food, equation 2 was used,
where the theoretical potential of methane production is determined by the percentage dry matter
(DM) percentage volatile solids of dry matter (VS) and specific production factor (Nm3 CH4/ton VS).
DM and VS were calculated from the values in Table 1 and the specific production factor was taken
from literature (Carlsson & Uldal, 2009). The literature value for fruit and vegetable waste (0.666
Nm3 CH4/kg VS) was used for bananas and salad, fish waste (0.93 Nm3 CH4/kg VS) was assumed to
correspond to beef and chicken, and the category bread (0.35 Nm3 CH4/kg VS) was used for bread.
Equation 2 gives the theoretical potential of methane and in reality this can be difficult to achieve,
but since all investigated food products contain easily available nutrients, it was assumed that the
entire theoretical yield was actually produced. The only loss of yield was assumed to be the 1 % of
food that got sorted out with the packaging and the loss of 3.5 % dry matter that ended up in the
digestate (Uppsala Vatten, 2013).
Equation 2:     !" #

 
$%& #

To produce biogas, the plant requires heat and electricity, and 10 kWh of electricity per tonne of
substrate was used with the marginal energy emission factor of 654 g CO2e/kWh (Gode et al., 2011).
Since biogas was used internally to produce heat, this 10 % was considered simply to reduce the yield
of biogas. This yield was also reduced by occasional flaring during production problems,
6/15

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

corresponding to 2 % of total production. In order to use the gas as a road fuel, it needs to be
upgraded and pressurised. This requires electricity corresponding to 5 % of the energy content of the
upgraded gas (Brjesson, 2003).
The outcome of the anaerobic digestion was biogas, which replaced diesel as a fuel for city buses,
and the digestate was assumed to replace production of nitrogen fertiliser and phosphorus fertiliser.
The upgraded biogas with an energy content of 9.7 kwh/Nm3 was assumed to replace diesel with an
energy content of 9.87 kWh/l (Fors, 2013). The diesel emissions replaced were assumed to be 82.3 g
CO2/MJ, which corresponded to the diesel used in Sweden (Eriksson & Ahlgren, 2013). Digestate
produced in the biogas plant replaced production of nitrogen fertiliser. The substituted production of
fertiliser was assumed to use natural gas as an energy source and, according to Ahlgren et al. (2010),
production of mineral fertiliser emits 2.41 kg CO2e/kg N. The amount of replaced fertiliser was based
on the nitrogen content of each food product, which corresponded to the protein content divided by
6.25 as a general factor and by 5.7 for bread (SNFA, 2013). Mineral phosphorus emitting 3.6 kg
CO2e/kg P (Linderholm et al., 2012) was also replaced by the digestate with an average P-content of
12.7 g P/kg DM (Uppsala Vatten, 2013). The marginal transportation distance for digestate to reach
the farmers was assumed to equal the present maximum distance: 30 km from the biogas plant. This
was performed by a heavy truck (<26 ton) with a euro 3 engine (NTMcalc, 2014).

2.3.5. Animal feed

In this scenario it was assumed that 100 % of the waste food products was used for animal feed. The
existing return system for bread was assumed to use a medium truck (<18 tonne) with a euro 3
engine and a maximum loading capacity of 7 tonnes. The emissions per transported tonne were set
to 0.19 kg CO2/tkm on rural roads in a traffic situation categorised as saturated (NTMcalc, 2014). The
truck was assumed to drive from a bakery in central Uppsala to all stores, deliver bread, pick up food
waste and return to the bakery (22.2 km). Thereafter the same type of truck was used for
transportation of food waste to pig farmers at an average distance of 20 km from the bakery.
When fed to pigs, the food waste could replace other feed. In this scenario all vegetable food
(banana, salad and bread) was assumed to replace oat in the system expansion. The substitution was
based on energy content to replace oat (Lindberg, 2009) with a carbon footprint of 0.42 kg CO2e/kg
(SIK foder, 2014). The animal products (chicken and beef) were assumed to replace soy in the system
expansion. The substitution was based on protein content to replace soy (Lindberg, 2009) with a
carbon footprint of 0.66 kg CO2e/kg (SIK foder, 2014).

2.3.5. Donation
According to the store interviewed, all the food in the investigated categories was wasted because it
was unsellable, but not inedible, which means that all of it could be donated to charity. Here the
more moderate assumption is used that 90 % of each product could be donated to charity and the
remaining waste composted in accordance with the composting scenario.
Distribution of donated food requires a significantly larger organisation than Uppsala can offer today.
It was therefore assumed that this kind of organisation can be developed to handle approximately
700 kg of food per day, that the number of people in need will not decrease and that they will use
the service from the charity so that none of the donated food is wasted in redistribution.

7/15

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

It was assumed that the food sorted in the store was placed in re-used cardboard boxes (the
secondary packaging used for deliveries that would otherwise be thrown away in store) in a cooling
room awaiting transportation. The collection of food was assumed to be carried out 300 days per
year with a pick-up (<2.5 tonne) diesel truck with a euro 3 engine. The emissions per transported
tonne were set to 2.1 kg CO2/tkm on urban roads with a traffic situation categorised as stop-go. The
collection round (21.2 km) was assumed to be circular, starting at the charity, stopping at each
supermarket to pick up 90 % of the daily mean waste, and then returning to the start location. At the
charity it was assumed that none of the food would be wasted as a result of a fluctuation in the
donated quality and type of food. The packaging (including banana peel) was incinerated in
accordance with the incineration scenario.
The food items donated to charity were assumed to replace bread, and in the system expansion the
production of bread (Angerwall & Sonesson, 2011) was substituted based on the energy content of
each food product and the corresponding amount of bread (assumed to have the same properties as
the bread described in Table 1).

3. Results
3.1. Emissions from the waste management scenarios
Food waste can be managed in many different ways, but from the scenarios investigated there is a
trend of a reducing climate impact with higher-priority stages in the food waste hierarchy. For all five
investigated food products, landfill was the option with the highest greenhouse gas emissions (0.21
to 3.1 kg CO2e/kg food waste). At the other end of the scale, donation and anaerobic digestion were
the alternatives with the fewest greenhouse gas emissions from the five food products (-0.67 to 0.047 kg CO2e/kg food waste in the donation scenario and -0.61 to -0.013 kg CO2e/kg food waste in
the anaerobic digestion scenario). Donation was the alternative with the fewest emissions for
chicken and bread, but for banana, lettuce and beef anaerobic digestion generated the fewest
emissions.
The other scenarios did not fully agree with the waste hierarchy. Incineration (-0.67 to 0.25 kg
CO2e/kg food waste) was a good option for bread and grilled chicken, but a poor option for lettuce
and banana where composting (0.043 kg CO2e/kg food waste) provided a better alternative.
Similarly, anaerobic digestion was a better alternative than animal feed (-0.13 to 0.0055 kg CO2e/kg
food waste) and for some products it was better than donation. The priority order for these scenarios
applied to bananas, grilled chicken, iceberg lettuce, beef and bread should therefore be anaerobic
digestion, donation, animal feed, incineration, composting and the least favourable alternative
landfill when solely considering greenhouse gases emissions and viewed as general options. If seen
as specific options for specific food products, incineration is a favourable alternative for dry
foodstuff.
For each type of food there is a least favourable and most favourable waste management option and
the difference between the two extreme alternatives indicates the potential for reducing greenhouse
gas emissions by changing the waste management system. When including all scenarios investigated,
the difference between landfill and any of donation, anaerobic digestion or incineration was greatest
for chicken (3.5 kg CO2e/kg food waste), followed by beef (2.8 kg CO2e/kg food waste), bread (2.6 kg
CO2e/kg food waste), bananas (1.8 kg CO2e/kg food waste) and lettuce (0.30 kg CO2e/kg food waste).
8/15

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

Since some of the scenarios are forbidden in Sweden, the same analysis could also be conducted on
all legal scenarios, which exclude landfill and animal feed from animal products. Then the difference
between composting and any of donation, anaerobic digestion or incineration was highest for bread
and beef (0.71 kg CO2e/kg food waste), followed by bananas (0.49 kg CO2e/kg food waste), chicken
(0.40 kg CO2e/kg food waste) and lettuce (0.30 kg CO2e/kg food waste).
Emissions originating from the waste management of supermarket food waste were also put into the
context of life cycle emissions of each food product from field to retail (Figure 2). Here the results for
bread stand out from the other products due to the high relative waste (10 %) and low carbon
footprint (0.7 kg CO2/kg bread), which gives the waste management a greater influence over the life
cycle emissions of this product. Beef is an example of the opposite with low relative waste (3.8 %)
and a high carbon footprint (29 kg CO2/kg beef), and therefore the waste treatment from the
supermarket waste has a low influence on the life cycle of beef.

3.2. Sensitivity analysis

In all scenarios the greatest influence on the result was from the substituted product in the system
expansion, except composting and landfill where there were no substituted products. Instead the
landfill scenario was influenced mainly by the methane produced in the landfill and the compost
scenario had no sub-process with a similar dominant influence. Since all system expansions are based
on assumptions, there are potential variations in the results. The scenario with the most uncertain
assumption is the donation scenario and two different assumptions were used alongside the original
scenario to test it in a sensitivity analysis. It is clear that this scenario is highly dependent on the
assumption made, and for the most extreme food product (beef) the result varies from 0.053 kg
CO2e/kg food waste, when the donated food replaces nothing, to -26 kg CO2e/kg food waste when
the beef replaces beef.
Other assumptions with a potential influence over the system expansion that need to be determined,
however, are whether peat is a renewable energy source or a fossil energy source and whether lower
heating values or higher heating values are used. A choice is not made here, but instead the
combination of both alternatives is compared. For all food products the fossil alternative gives a
more extreme result, whether negative or positive. Also if the food produces energy in line with HHV,
the amount of peat that can be replaced is higher for all food products then for LHV. The true value
for incinerating wet food products is likely to be found between LHV and HHV, so these values can be
considered to present the extreme values.

4. Discussion
When food waste is managed in a way that is more prioritised in the waste hierarchy, it normally
generates fewer greenhouse gas emissions compared to less prioritised waste management options.
This pattern corresponds well with previous studies (Laurent et al., 2013a), but since the scenarios
are applied in a local context with specific data, general conclusions based on the results should be
made with caution. On the other hand, Uppsalas local infrastructure is likely to correspond well with
several other Swedish cities. The results even corresponded well with the completely different
Belgian context described by Vandermeersch et al. (2014), which indicates that the results from the
present study could be useful for supermarkets outside Uppsala. The reason for making the results
more general would be the unique number of scenarios evaluated, where more prioritised levels in
9/15

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

the waste hierarchy are also included in the investigation. This gives an insight into what higher levels
in the waste hierarchy actually have to offer in the context of selection of waste management
options for supermarket food waste, but also into the methodical difficulties faced when
investigating waste management options such as donation.
For each food product different scenarios applied different problems and possibilities that influenced
the result. In the landfill scenario the major part of the emissions came from methane production in
the landfill. Since methane potential was dependent on the easy available nutrients in the volatile
solids fraction, dry and energy-rich products such as chicken, beef, bread and bananas generated the
highest level of greenhouse gas emissions. Since lettuce contains mostly water, it was less
detrimental to put it in landfill in comparison with other food items. On the other hand landfill was
the least favoured of the investigated alternatives. The greatest potential for reducing the impact of
supermarket food waste treatment on the climate was to stop landfilling the food waste. A ban on
landfilling organic waste, as is in place in Sweden already, therefore seems to be an efficient measure
for reducing the climate impact of food waste management.
Composting gave rise to small amounts of greenhouse gases and can be categorised as disposal
rather than recovery in Table 2, since there is no recovery of the nutrients in the waste. If there were
nutrient recovery that could replace the production of mineral fertilizer in this process, it might
improve this waste management scenario, but without it is definitely among the less favourable
options of what to do with food waste.
For incineration the results showed an opposite pattern to the landfill scenario, since high water
content here is less favourable. The incineration scenario showed the highest levels of greenhouse
gas emissions for the products with the highest water content (lettuce and bananas), where the
lower heating value was even negative due to the large amount of water that had to be evaporated.
A negative LVH means that fuel needs to be added to the system expansion, and that no fuel is
replaced. Since flue gas condensation was used in the incineration plant, a large proportion of the
energy required for burning the wet food could be recovered and the results presented are therefore
likely to describe incineration as being less favourable then it is. On the other hand, there is little
need for heat during the summer months, which makes extra fuel less necessary. There is therefore
uncertainty about whether there is any peat to replace, thus having a negative impact on the results
of this scenario. In general it is not a good idea to burn really wet food, and since supermarket food
waste often contains a high proportion of fruit and vegetables (Eriksson, 2012) it is normally too wet
to be suitable for incineration. However, bread is dry enough to burn in order to recover energy.
Anaerobic digestion is a middle range alternative in the waste hierarchy, but in this study it was still
one of the most favourable options. One reason for this is that all food products provide easily
available nutrients for the anaerobic bacteria producing methane in the biogas reactor. Another
reason is that the main product, methane, can be used to substitute diesel, with its comparably large
carbon footprint, as a fuel for city buses. For one product, namely chicken, the result was surprisingly
low, but this is because a third of the chicken is bones, which cannot be degraded to methane during
the short retention time unless already sorted out before they reach the reactor. If the product were
bone-free meat, like beef, the outcome would be even fewer emissions for grilled chicken in the
anaerobic digestion scenario.

10/15

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

Using food waste from supermarkets for animal feed is restricted (EC, 2009), but even though the
legislation only limits the possibility of using animal products for animal feed, it is sometimes seen as
an obstacle to using food waste for animal feed. The exception is bread where large-scale return
systems already exist in which sorted-out bread is used to feed pigs. Even though this scenario is
already in use for some bread products, it did not show lower levels of emissions than the lower
prioritised alternative with anaerobic digestion. This fits well with a Belgian study (Vandermeersch et
al., 2014) that concluded that animal feed instead would be better than anaerobic digestion when
considering environmental impact categories regarding land use rather than categories focusing on
fossil fuel. This is due to the low emissions of greenhouse gas per energy unit and protein content
from the substituted product, which in this case was assumed to be oat for the vegetable products
and soy for the animal products. There is also a possible limitation on how much and of which quality
the food waste has to be in order to be suitable for animal feed. One way to increase the availability
of this waste management method is to use biological barriers in order to stop possible diseases
(Rumpold & Schlter, 2013). It might also be possible to feed yeast or fly larvae in order to convert
food waste unsuitable for direct use as pig feed into protein-rich fodder. It is unlikely that these more
complex systems would make the animal feed scenario more efficient as regards greenhouse gas
emissions, even though they may provide possibilities other than the very basic scenario investigated
in this study.
Donation of food waste was the highest prioritised scenario investigated and the climate impact from
this scenario was highly dependent on what this food replaces in the system expansion. When
replacing nothing (i.e. people in need would otherwise go hungry or starve), this scenario gave rise to
a low level of emissions due to transportation and cold storage. This sensitivity analysis represented
an extreme assumption that had a considerable influence on the results. The other end of the scale
of possible assumptions was that the donated food replaced the same products as those donated
(i.e. people in need would otherwise eat banana, chicken, lettuce, beef and bread in the same
amounts as donated) which corresponded with the method used in Schneider (2013). The
assumption best describing reality was probably found between these two extreme scenarios, and
therefore bread was selected as the substituted product as in Eriksson & Strid (2013). However, this
was an assumption with a large possible variation and the results must therefore be interpreted with
caution.
The only way in which a waste management option can reduce its environmental impact is by
eliminating other material that had otherwise been used for the purpose. Therefore it was important
to analyse the competing effect of the waste management options. However, use of food waste by
people who do not otherwise buy food can have strong social benefits even if it has a minimal
climate impact. Selling food waste as in Strid & Eriksson (2013) is therefore of greater environmental
as well as of economic benefit in comparison with the lesser environmental, but stronger social
benefit of donation.
The method used in this study presented an opportunity to analyse not only a waste management
scenario for a mixed fraction of food waste, but also which types of food are suitable for which type
of waste management method. Since only five types of food were included, the analysis did not give
the full picture, but some interpretations can still be made. A first remark is that there was a
comparatively large difference between low and high prioritised scenarios for bread (2.6 kg CO2e/kg
food waste when including all scenarios and 0.71 kg CO2e/kg food waste excluding landfill), but also
11/15

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

that supermarket food waste management had a relatively large influence over the whole life cycle
of bread (Figure 2). Therefore there was greater potential in changing the waste management system
from the present use of composting (or animal feed) to donation, anaerobic digestion or incineration.
The opposite to bread is lettuce where the difference between low and high prioritised scenarios was
comparatively small (0.3 kg CO2e/kg food waste). Furthermore the effect on the life cycle of lettuce
from field to retail was not greatly affected by the choice of waste methodology, even though lettuce
waste was high (9.7 %) and the carbon foot print was low (0.45 kg CO2e/kg food), just like bread.
Another way to exemplify this was to compare the energy content per greenhouse gas emission ratio
presented in Table 1. Of the five investigated food products, only bread had a value (15 MJ/kg CO2e)
that was comparable with fossil energy carriers such as diesel (9.3 MJ/kg CO2e) and peat (12 MJ/kg
CO2e). Food waste is a way of making the market pay for bio-based products, and then letting the
energy sector use them for free instead of using fossil resources. From an environmental perspective
this is only beneficial if the bio-based products emit less than the corresponding fossil energy carrier
per MJ, provided the energy sector uses fossil energy.
The comparison of bread and lettuce showed that the potential in moving to higher prioritised levels
in the food waste hierarchy is dependent on the type of food considered and what that type of food
can replace in the system expansions, rather than on the amount of waste or the carbon footprint.
High energy content and low water content are therefore important factors when considering waste
management methods. This corresponded well with the conclusions made by Vandermeersch et al.
(2014). The exception is expected to be found when also including the prevention of food waste as a
scenario, as in Gentil et al. (2011) where the products with a large carbon footprint, such as beef, are
likely to be the most favourable to work with to reduce the climate impact of food waste. The most
efficient way to work with food waste in supermarkets could therefore be to reduce further the
already low waste of meat products, donate products that are easy to handle, such as bread and
bananas, to people in need and send the rest for anaerobic digestion if wet and incineration if dry,
when considering the climate impact alone. When also considering the social value of feeding hungry
people, the donation scenario might be the most efficient for all food products. However, from a
strict climate perspective, donations are not necessarily more efficient then anaerobic digestion or
incineration, which highlights the need for several indicator values as a base for any measures taken.

5. Conclusions
The main conclusion of this study was that the prioritising levels in the waste hierarchy corresponded
to some extent with the studied waste management scenarios with food waste from supermarkets in
Uppsala. The greatest potential for reducing emissions of greenhouse gases was found in anaerobic
digestion and donations. Bread was identified as the product with the greatest potential for reducing
greenhouse gas emissions by managing supermarket food waste in a more prioritised way, and
lettuce had the lowest potential of the food investigated.
The result in all scenarios was highly dependent on the substituted product in the system expansion.
It is therefore most favourable to choose the scenario in which the most resource-intensive products
or services can be replaced by the waste food. What can be substituted is dependent on the type of
food, since a high energy content and high dry matter content offers more potential for replacing
other products.

12/15

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

6. Acknowledgments
The study was funded by the Swedish Retail and Wholesale Development Council and was carried
out in cooperation with Axfood AB and its subsidiary Willy:s AB. The Swedish Retail and Wholesale
Development Council took no part in the execution of the study. Willy:s AB provided study objects
and recorded data, but took no part in interpreting the data. The authors would like to thank the
staff of all the companies involved for their help and cooperation.

7. References
Ahlgren, S., Bernesson, S., Nordberg, ., Hansson, P.-A., 2013, Nitrogen fertiliser production based on biogas
Energy input, environmental impact and land use, Bioresource Technology 101, 71817184.
Alvarez, H, 2010, Energiteknik, Studentlitteratur, Lund.
Angervall, T. & Sonesson, U., 2011, Frenklad metod fr klimat-/GWP berkningar av livsmedel, SIK the
Swedish Institute for Food and Biotechnology, Gothenburg.
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.
Bjrklund, A., 1998, Environmental systems analysis of waste management with emphasis on substance flows
and environmental impact, Licentiate thesis, Department of Chemical Engineering and
Technology/Industrial Ecology, Royal Institute of Technology, Stockholm.
Brjesson, P., 2003, Energianalys av drivmedel frn spannml och vall. Report 54, Department of Technology
and Society, Lund University, Lund.
Carlsson, M. & Uldal, M., 2009, Substrathandbok fr biogasproduktion. Rapport SGC 200, Svenskt Gastekniskt
Center, Malm.
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, SR 828, SIK the Swedish Institute for Food and Biotechnology, Gothenburg.
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, 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 Union, Strasbourg.
EC, 2009, Regulation 1069/2009 of the European Parliament and of the Council of 21 October 2009 laying down
health rules as regards animal by-products and derived products not intended for human consumption and
repealing Regulation (EC) 1774/2002 (Animal by-products Regulation), Official Journal of the European
Union, Strasbourg.
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, Sweden.
Eriksson, M. & Ahlgren, S., 2013, LCAs for petrol and diesel - a literature review. Report 058, Department of
Energy and Technology, Swedish University of Agricultural Science, Uppsala.
Eriksson, M. & Strid, I., 2013, Svinnreducerande tgrder i butik - Effekter p kvantitet, ekonomi och
klimatpverkan, 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.

13/15

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

Fors, E., 2013, Ground coffee waste as substrate for biogas or as fuel in a heating plant a study of effects on
greenhouse gas emissions and economical costs, Master thesis 2013:11, Department of Energy and
Technology, Swedish university of Agricultural Science, Uppsala.
Gentil, E., Gallo, D., Christensen, T.H., 2011, Environmental evaluation of municipal waste prevention, Waste
management, 31, 2371-2379.
Gode, J., Martinsson, F., Hagberg, L., man, A., Hglund, J., Palm, D., 2011, Miljfaktaboken 2011 - Estimated
emission factors for fuels, electricity, heat and transportation in Sweden, Vrmeforsk Service AB,
Stockholm, Sweden.
Google maps, 2014, https://www.google.se/maps/, Accessed 2014-06-12.
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.
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.
Lindberg, J.E., 2009, Principer fr skattning av energiinnehll i foder till grisar, Report to The Swedish meat
farmers research foundation, Department of Animal Nutrition and Management, Swedish university of
Agricultural Science, Uppsala.
Linderholm, K., Tillman, A-M., Mattsson, J.E., 2012, Life cycle assessment of phosphorus alternatives for
Swedish agriculture, Resources, Conservation and Recycling, 66, 2739.
Nellemann C, MacDevette M, Manders T, Eickhout B, Svihus B, Prins AG, et al. 2009, The environmental food
crisis the environments 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
and the Swedish University of Agricultural Science, Uppsala.
NTMcalc, 2014, Metod fr berkning av gods & logistik. http://www.ntmcalc.se/index.html, Accessed 2014-0520.
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, 106115.
Rumpold, B.A. & Schlter, O.K., 2013, Potential and challenges of insects as an innovative source for food and
feed production, Innovative Food Science & Emerging Technologies, 17, 1-11.
Salhofer, S., Obersteiner, G., Schneider, F., Lebersorger, S., 2008, Potentials for the prevention of municipal
solid waste. Waste Management, 28, 245-259.
Schneider, F., 2013, The evolution of food donation with respect to waste prevention, Waste Management, 33,
755763.
Scholz, K., 2013, Carbon footprint of retail food wastage a case study of six Swedish retail stores, Independent
thesis 2013:05, Department of Energy and Technology, Swedish University of Agricultural Science,
Uppsala.
SEPA, 2012a, Frn avfallsantering till resurshushllning, Sveriges avfallsplan 2012-2017, Report 6502, Swedish
Environmental Protection Agency, Stockholm.
SEPA, 2012b, Avfall i Sverige 2010, Report 6520, Swedish Environmental Protection Agency, Stockholm.
SEPA, 2013, Frslag till etappml fr minskad mngd matavfall, NV-00336-13, Swedish Environmental
Protection Agency, Stockholm.
SIK foder, 2014, LCA Foder Database, http://www.sikfoder.se/En/Sidor/default.aspx, Accessed 2014-05-26.

14/15

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

SNFA, 2014, The Swedish national food database, http://www.slv.se/sv/grupp1/Mat-och-naring/Vad-innehallermaten/Livsmedelsdatabasen-/, Accessed 2014-06-12.
Spngberg, J., 2014, Recycling plant nutrients from waste and by-products, Doctoral thesis 2014:20, Swedish
University of Agricultural Science, Uppsala.
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 primrproduktionen och
grossistledet i Sverige, Report 2014:06, Swedish Board of Agriculture.
Strmberg, B. & Herstad Svrd, S., 2012, The Fuel Handbook 2012, V Vrmeforsk Service AB, Stockholm,
Sweden.
Stuart, T., 2009, Waste: uncovering the global food scandal. Penguin Books, London, UK.
Uppsala Vatten, 2013, Miljrapport 2013 Biogasanlggningen vid Kungsngens grd, Uppsala Vatten, Uppsala.
USEPA, 2014, http://www.epa.gov/foodrecoverychallenge/, Accessed 2014-06-13.
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.
Vattenfall, 2013. Safety, Health and Environment 2013, Vattenfall Heat, Uppsala
Zorpas, A., Lasaridi, K., 2013, Measuring waste prevention, Waste Management, 33, 1047-1056.

Table 1. Description of the investigated food products regarding content, package waste and carbon
footprint
Table 2. Four waste-prioritising systems are listed to illustrate how the levels in the systems
correspond to one other and to the investigated waste management scenarios
Figure 1. A schematic description of the waste management scenarios, their outcome and what each
scenario replaces
Figure 2. Greenhouse gas emissions from each waste management scenario and food product
Figure 3. Greenhouse gas emissions from each waste management scenario and food product in
relation to life cycle emissions of each product from field to retail (Table 1)
Figure 4. Sensitivity analysis for the donation scenario where the replaced food is assumed to be
nothing, bread or the same as the food donated in the system expansions. The figure is reduced and
the value for the original beef bar is -26 kg CO2e/kg food waste
Figure 5. Sensitivity analysis for the incineration scenario where peat is assumed separately to be
either a renewable energy source or a fossil energy source in the system expansions

15/15

Highlights (for review)

Highlights:
Five individual waste streams of bananas, grilled chicken, lettuce, stewing beef and toast bread
from supermarkets in Uppsala in Sweden were analysed.
Six scenarios (landfill, incineration, composting, anaerobic digestion, animal feed and donations)
representing different levels in the waste hierarchy were investigated.
Emissions of greenhouse gases were calculated using the life cycle assessment methodology.
The results confirmed waste hierarchy to be a useful but approximate tool in prioritising waste
management options.

Figure1

Waste managemant scenarios

Sold food

Surplus
food

Donation

Food

Bread

Animal feed

Feed

Oat,
soy
Substitute

Anaerobic
digestion

Biogas,
fertilizer

Incineration
Landfill

Diesel,
product
mineral fert.

Substituted
system

Composting

Produced food

Substituted production

Dist. heating,
electricity

Peat

Figure2

Landfill

Incineration

Composting

Anaerobic digestion

Animal feed

Donation

GWP for WM (kg CO2/kg FW)

3.5
3.0
2.5
2.0
1.5

1.0
0.5
0.0
-0.5
-1.0
banana

chicken

lettuce

beef

Waste managemanet scenario and type of food wasted

bread

Figure3

Figure3
Click here to download high resolution image

Figure4

Figure4
Click here to download high resolution image

Figure5

Figure5
Click here to download high resolution image

Table1

Product

Country

Energy
content1
(kJ/kg prod)

Water
content1
(%)

Package
material

Package
weight
(g/kg prod)

Banana
Costa Rica
3340
74
Banana peel
300
Grilled
Sweden
9980
57
Polypropylene
10
chicken
Iceberg
Sweden
610
96
High-density
7
lettuce
polythene
Stewing beef
Sweden
6000
72
Polypropylene
40
Toast bread
Sweden
10350
31
Polypropylene
10
1. Data from SNFA (2014) corresponding to the higher heating value (HHV)
2. In relation to sold mass and based on data from two of the investigated stores during 2010-2012
3. Life cycle emissions from field to retail
4. Based on Nilsson (2012)
5. Based on Davies et al. (2011) modified by Strid et al. (2014)
6. Based on Cederberg et al. (2009) modified by Scholz (2013)
7. Based on Angerwall & Sonesson (2011)
8. Energy content refers to the higher heating values

Wasted mass2
(annual
ton/store)

Relative
waste2
(%)

3.7
0.26

3.3
7.0

Carbon
footprint3
(kg CO2e/kg
prod)
1.14
2.14

Energy per climate impact8

(MJ/kg CO2e)

3.8

9.7

0.455

1.4

0.054
4

3.8
10

296
0.77

0.21
15

3.1
4.8

Table2

Priority according to the

waste hierarchy

Waste hierarchy
EU

Food waste pyramid

UK/London
Feeding the 5000

Food recovery hierarchy

USA
US EPA

Moerman ladder
Holland

Highest

Prevention

Reduce

Source reduction

Prevention

Re-use and preparation for

re-use

Feed people in need

Feed hungry people

Use for human food

Investigated scenarios
Sweden/Uppsala
Present study

Donation

Conversion to human food

Recycle

Recovery

Disposal

Lowest

Feed livestock

Compost & 100 % renewable

energy

Disposal

Feed animals

Use in animal feed

Industrial use

Raw materials for industry

Composting

Processing to make fertiliser

for co-fermentation

Biogas

Processing to make fertiliser

through composting

Composting

Use for sustainable energy

Incineration

Incineration or Landfill

Animal feed

Burning as waste

Dumping

Landfill