Science of the Total Environment 710 (2020) 135580

Contents lists available at ScienceDirect

Science of the Total Environment

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

Assessing the economic and environmental sustainability of household

food waste management in the UK: Current situation and
future scenarios
Peter C. Slorach, Harish K. Jeswani, Rosa Cuéllar-Franca, Adisa Azapagic ⁎
Department of Chemical Engineering and Analytical Science, The University of Manchester, Manchester, M13 9PL, UK

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• Sustainability of five 2030 scenarios for

household food waste treatment is
assessed.
• Scenarios with the highest share of an-
aerobic digestion are the most sustain-
able.
• Life cycle costs are minimised by
prioritising the upgrading of biogas.
• The best scenarios could save £338 mil-
lion and 480,000 t CO2 eq. per year.
• Preventing just 2% of waste matches the
GHG savings of the best treatment
scenario.

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

Article history: The value embedded in food waste is increasingly being recognised, with the UN targeting a 50% reduction in
Received 1 October 2019 consumer food waste and the EU recycling of 60% of all household waste, both by 2030. Aiming to provide guid-
Received in revised form 15 November 2019 ance on the most sustainable food waste utilisation routes, this study evaluates the life cycle environmental and
Accepted 15 November 2019
economic sustainability of five plausible scenarios for the year 2030. Focusing on the UK for context, these are
Available online 18 November 2019
compared to the current treatment of food waste as well as to its potential future prevention. The scenarios con-
Editor: Damia Barcelo sider a differing share of four widely-used treatment methods: anaerobic digestion, in-vessel composting, incin-
eration and landfilling. The scenario with the highest anaerobic digestion share that recovers both heat and
Keywords: electricity is the best option for seven out of 19 environmental impacts and the second best for life cycle costs.
Anaerobic digestion Upgrading anaerobic digestion biogas to biomethane achieves the lowest global warming potential and life
Food waste cycle costs. Net-negative global warming potential (savings) can be achieved if the heat from anaerobic digestion
Incineration and incineration or biomethane are utilised to displace natural gas. Displacing a future electricity mix does not
In-vessel composting lead to significant global warming potential savings due to the expected grid decarbonisation. However, savings
Life cycle assessment
are still achieved for metal depletion and human and terrestrial toxicities as they are higher for decarbonised grid
Life cycle costing
electricity due to the increased share of renewables. A greater share of in-vessel composting leads to higher im-
pacts because of the high electricity consumption. Landfill reduction has an economic advantage for all the sce-
narios, except for the business-as-usual, with life cycle costs 11–75% lower than for the current situation. While

⁎ Corresponding author.
E-mail address: [email protected] (A. Azapagic).

https://doi.org/10.1016/j.scitotenv.2019.135580
0048-9697/© 2020 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
2 P.C. Slorach et al. / Science of the Total Environment 710 (2020) 135580

future scenarios improve the overall sustainability compared to the current situation, halving food waste by 2030
can save 15 times more greenhouse gas emissions than the best treatment scenario without waste reduction.
Therefore, any commitments to improve the sustainability of food waste treatment must be accompanied by
an effective waste prevention strategy. The outcomes of this work can help waste treatment operators and policy
makers towards more sustainable food waste management. Although the focus is on UK situation, the overall
conclusions and recommendations are applicable to other regions.
© 2020 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).

1. Introduction mandatory separate FW collections will be legislated in England by

2023 (DEFRA, 2018a). The devolved parliaments for Scotland, Wales
The global food system is estimated to generate 19–29% of the total and Northern Ireland have all already mandated requirements for FW
anthropogenic greenhouse gas emissions (Vermeulen et al., 2012). collections by local authorities. This is partly driven by the national tar-
With approximately a third of food wasted each year, a significant frac- get for recycling 50% of household waste by 2020 (DEFRA, 2018b). In ad-
tion of these emissions is generated unnecessarily (Gustavsson et al., dition, the EU Circular Economy Package, to which the UK is committed
2011). The production value of this waste is estimated at $1 trillion, in- (DEFRA, 2018a), targets an increase to 60% recycling of municipal waste
creasing to $2.6 trillion when adding the environmental and social costs by 2030 (EC, 2018b; EC, 2018a) and a limit on municipal waste
(FAO, 2014). With a growing global population and the purchasing landfilled of b10% by 2035 (EC, 2018a). Meeting these targets will re-
power of an expanding middle class (Kharas, 2017), the future demand quire significant changes to the current FW management. However, be-
for food will increase. A sustainable way to feed a world, where 821 mil- fore any changes are implemented, it is essential to evaluate the
lion people still lack sufficient nourishment (FAO, 2018), is to reduce environmental and economic sustainability of future FW management
waste and ensure the food can be consumed by those who need it. options.
This is reflected in the United Nations' Sustainable Development Goal Several studies have addressed the environmental impacts and, to a
(SDG) 12.3 which aims to halve food waste (FW) at the retail and con- lesser extent, life cycle costs of the current FW treatment routes over the
sumer levels by 2030 (UN, 2015). Waste prevention also has the highest past decade. These studies are summarised in Table 1. For a review of
priority in waste management hierarchies (EC, 2008; Feeding the 5000, FW life cycle assessment (LCA) studies dating back before 2010, see
2016) as it can reduce the environmental and economic impacts associ- Bernstad and la Cour Jansen (2012). The majority of the studies that in-
ated with land use, livestock rearing, crop cultivation, processing and lo- cluded AD as a waste treatment option concluded that it had the lowest
gistics (Papargyropoulou et al., 2014; Oldfield et al., 2016; Eriksson and overall impacts. With energy systems forecast to continue to change in
Spangberg, 2017). However, some FW remains an inevitable product of the coming years, it is significant that the benefits of AD are highly de-
our lifestyle and sustainable use of this waste stream can reclaim other- pendent on the system credits for displacing recovered fertilisers and
wise lost resources. energy (Evangelisti et al., 2014). Furthermore, while AD has the lowest
Within the EU, the total annual FW is estimated at 88 ± 14 Mt, of global warming potential (GWP), the application of digestate to land
which the majority (47 ± 4 Mt) is generated by households leads to higher eutrophication and acidification (Bernstad and la Cour
(Stenmarck et al., 2016). A UK waste reduction initiative (WRAP, Jansen, 2011; Slorach et al., 2019b) than incineration due to nutrient
2008) resulted in a 14% fall in household FW between 2007 and 2010; run-off and volatilisation of ammonia. If digestate is gasified instead to
however, very little progress has been made in recent years, as shown recover energy, the GWP is reduced; however, using digestate on land
in Fig. 1. Although recycling rates for the metal, plastic, paper and still has lower impacts in all other categories (Tong et al., 2018).
glass, in which food is often packaged, range from 69 to 82% in the UK Studies also reported that incineration had lower overall impacts
(DEFRA, 2018b), only an estimated 10% of the FW is collected for than composting (Bernstad and la Cour Jansen, 2011; Oldfield et al.,
targeted treatment via anaerobic digestion (AD) or in-vessel 2016; Slorach et al., 2019a). This is attributable to the electricity
composting (IVC) (Quested and Parry, 2017). The remaining collected displaced by incineration and the high acidification and eutrophication
FW is embedded in the municipal solid waste (MSW) and is either in- caused by composting. Oldfield et al. (2016) found landfilling to be the
cinerated or landfilled. worst option, primarily due to the methane emissions and eutrophica-
In the UK, separate FW collections are offered to 99% of Welsh tion linked to leachate. Another study (Slorach et al., 2019a) also esti-
households but in England, where local authorities control its provision, mated that landfill was the worst option but only for GWP, while IVC
this service is available only to 45% of households (UK Parliament, performed worse for all other impacts due to high electricity consump-
2017). There is currently no legal requirement in UK law stipulating tion. However, Levis and Barlaz (2011) suggested that a modern landfill
the provision of separate food collections but, subject to consultation, with energy recovery had a lower GWP than composting.

Avoidable Possibly avoidable Unavoidable Total

10
Houeshold FW generated

6
(Mt/yr)

0
2007 2010 2012 2014 2015

Fig. 1. Household food waste (FW) generated in the UK from 2007 to 2015 (Quested and Parry, 2017). [Avoidable: food suitable for human consumption; Possibly avoidable: edible food,
depending on consumer preferences (e.g. bread crusts and potato peel); Unavoidable: inedible parts of food (e.g. egg shells and animal bones)].
 P.C. Slorach et al. / Science of the Total Environment 710 (2020) 135580 3

Table 1
Summary of life cycle assessment and life cycle costing studies on the management of mixed food waste.

Study Type of Country Waste Waste treatments LCA Indicatorsa

assessment software

Bernstad and la LCA Sweden FW AD/composting/incineration EASEWASTE AP, GWP, NE, ODP, POF
Cour Jansen (household)
(2011)
Edwards et al. LCA Australia FW AD/Food waste disposer to sewer/composting SimaPro 8.0 ADP, AP, EP, FD, GWP, HT, OD, POF
(2018) (household) (home and centralised)/MBT
Eriksson et al. LCA Sweden FW Donation/animal None GWP
(2015) (supermarket) feed/AD/composting/incineration/landfill
Evangelisti et al. LCA UK OFMSW AD/incineration/landfill GaBi 5 AP, GWP, NE, POF
(2014)
Kim and Kim LCA South FW Composting/landfill/animal feed TOTAL 3.0 GWP
(2010) Korea (household)
Levis and Barlaz LCA United FW AD/composting/landfill None GWP, NOX, SO2
(2011) States (commercial)
Oldfield et al. LCA Ireland FW AD/composting/incineration/landfill GaBi 6 AP, EP, GWP
(2016) (household)
Slorach et al. LCA UK FW AD/incineration/landfill GaBi 8 PED, 18 ReCiPeb impacts
(2019b) (household)
Tong et al. (2018) LCA Singapore FW AD (with composting, incineration and GaBi ADP, AP, EP, FD, FET, GWP, HT, MET,
(restaurant) gasification)/Incineration OD, POF, TET
Vandermeersch LCA/exergetic Belgium FW AD/Animal feed None 18 ReCiPeb impacts, cumulative exergy
et al. (2014) LCA (supermarket) from the natural environment
Carlsson Reich LCA/LCC Sweden All household AD/composting/incineration/landfill ORWARE LCC with economic weighting for
(2005) waste inc. FW environmental effects
Kim et al. (2011) LCC South FW AD/composting/animal feed/composting/food TOTAL 3.0 GWP, LCC, carbon costing
Korea (household) waste disposer to sewer/incineration/landfill
Martinez-Sanchez LCA/LCC Denmark FW AD/incineration/animal feed EASETECH GWP, POF, LCC
et al. (2016) (household)
a
AD: anaerobic digestion; ADP: abiotic depletion potential; AP: acidification potential; EP: eutrophication potential; EU: energy use; FD: fossil depletion; FET: freshwater ecotoxicity;
FW: food waste (household unless otherwise stated); GWP: global warming potential; HT: human toxicity; LCA: life cycle assessment; LCC: life cycle costing; MBT: mechanical biological
treatment; MET: marine ecotoxicity; NE: nutrient enrichment; OD: ozone depletion; OFMSW: organic fraction of municipal solid waste; PED: primary energy demand; POF: photochem-
ical oxidant formation; TET: terrestrial ecotoxicity.
b
ReCiPe life cycle impact assessment methodology (Goedkoop et al., 2013).

Eriksson et al. (2015) and Vandermeersch et al. (2014) considered comparative impacts of prevention must also be considered alongside
specific FW streams, including waste bread at retailers. The authors future treatment options.
showed that producing animal feed from waste bread had low impacts Therefore, this paper looks at potential future scenarios for the man-
but was not sustainable for feed from lower energy-content foods, such agement of household FW with the aim of recommending the most sus-
as green vegetables. Eriksson et al. (2015) also suggested that high tainable option(s) to waste treatment operators, local authorities and
water-content foods could have a detrimental impact on incineration. national policy makers. Life cycle assessment and life cycle costing
However, it is not realistic to expect household FW to be segregated have been used for these purposes. Focusing on the UK for context,
by food type so any waste management options have to consider co- five future scenarios have been developed for the year 2030 based on
mingled FW. the targets for waste management and prevention, current treatment
There are far fewer studies on the life cycle costs of FW treatment. trends and expected technological advancements. The treatment sce-
The authors of this paper (Slorach et al., 2019a) found previously that narios are also compared to the implications of FW prevention. As far
incineration had the lowest life cycle cost compared to AD, composting as the authors are aware, this is the first study internationally to assess
and landfill. However, AD could have the lowest cost to local authorities the environmental and economic impacts of future scenarios for
if sufficient food was collected and, without landfill tax, landfilling treating household FW.
would have the lowest cost. Kim et al. (2011) found that in South
Korea landfilling had the lowest cost. However, when the value of by- 2. Methods
products and the cost savings related to avoided greenhouse gas
(GHG) emissions were accounted for, landfilling was the costliest op- The LCA of the current and future FW treatment scenarios has been
tion. The best alternative was wet animal feed, followed by dry feed, in- carried out in accordance with the ISO 14040/14044 methodology (ISO,
cineration, composting and AD. Martinez-Sanchez et al. (2016) 2006a; ISO, 2006b). GaBi V8 software (Thinkstep, 2019) has been used
highlighted the additional costs associated with separate food collec- to model the treatment technologies and scenarios. The impacts have
tions for animal feed or AD compared to incineration. The authors also been estimated following the ReCiPe 1.08 life cycle impact assessment
emphasised the cost saving potential from waste prevention, which methodology (Goedkoop et al., 2013), considering 18 midpoint impacts
may also be beneficial environmentally as long as the saving is spent included in this method. In addition, primary energy demand has also
on low-impact activities, to avoid a rebound effect. Oldfield et al. been determined, based on the GaBi methodology (Thinkstep, 2019).
(2016) also discussed the significant environmental and cost savings re- Therefore, a total of 19 life cycle impacts are considered. The life cycle
lated to FW prevention compared to other treatment scenarios. costs (LCC) have been estimated following the method proposed by
All existing literature on FW management is focused on the current Hunkeler et al. (2008) and Swarr et al. (2011).
situation with no studies considering how this may develop in the fu-
ture. With the EU and UK targeting increased recycling and separate 2.1. Goal and scope
FW collections in the near future, it is vital to identify the most sustain-
able options for managing this waste stream. These targets coincide The goal of this study is to evaluate the environmental and economic
with national and international goals to reduce FW and, therefore, the sustainability of possible 2030 scenarios for management of household
4 P.C. Slorach et al. / Science of the Total Environment 710 (2020) 135580

FW in the UK in comparison with the current situation. The following incinerated, 34% landfilled, 10% composted and 6% anaerobically
FW treatment options are used currently for FW treatment: AD, IVC, in- digested (Slorach et al., 2019a).
cineration and landfilling. It is envisaged that these options will still be
used in 2030 and hence they are also considered in the scenario analy- 2.2.2. Future scenarios
sis, but assuming various technological improvements, as detailed later Five 2030 scenarios have been defined by considering the current
on. trends and targets for FW recycling; these are detailed below. Waste
The scope of the study is from cradle to grave. As shown in Fig. 2, the prevention is not considered in the scenarios but is evaluated separately
following life cycle stages are considered: collection of FW from house- to compare its sustainability implications with FW treatment scenarios;
holds, treatment and the recovery of valuable resources, and associated for details on waste prevention, see Section 2.4.
environmental credits and revenue. The construction of waste treat-
ment facilities is included but their decommissioning is excluded due 2.2.2.1. Business as usual (BAU). The BAU scenario assumes the same
to a lack of data. FW is assumed to be burden-free. treatment mix as the current situation but includes technology im-
The functional unit is defined as “treatment of 1 t of household FW”. provements, future electricity mix and any forecast changes to costs.
The impacts and costs are also considered at the UK level to determine These are detailed in Section 2.3.
potential implications of managing FW in the future. The implications
of waste prevention are also evaluated. 2.2.2.2. Scenario 1-A/B: no change in FW collection and incineration
For the LCC, the waste collection costs and gate fees are paid by the prioritised. Scenario 1 (S1) represents a situation where targeted FW
local authority. The waste facility operators bear the costs of plant con- collections remain stable, but there is a reduction in landfilling and con-
struction and operation but also receive the revenue from the gate fees sequent increase in incineration. The EU Circular Economy Package tar-
and from the sale of any recovered products, including any related in- gets that less than 10% of municipal waste be landfilled by 2035 (EC,
centives. In alignment with the LCA methodology, labour costs are 2018a). Between 2013 and 2016, the UK household MSW sent to landfill
excluded. fell from 36% to 22% (DEFRA, 2017; SEPA, 2017; StatsWales, 2017;
Current trends and forecasts have been used to predict future gate DAERA-NI, 2017). Therefore, it is feasible that the proposed 10% target
fees and electricity costs. However, current costs have been used for could be met by 2030 which is assumed in this scenario. The UK incin-
the collection services, consumables and capital costs as they are likely eration capacity is expected to continue to increase until at least 2021
to be stable over the next decade. Furthermore, the lack of data does (Eunomia, 2017). However, the amount of residual waste is forecast to
not allow any confident forecasts. Inflation is not considered as it reduce due to increased recycling and, as a results, there may be excess
would apply to all costs and would not differentiate the scenarios. capacity of incineration facilities from 2020 onwards (Eunomia, 2017).
The next section provides an overview of the current situation and Thus, there will be sufficient capacity for the increased incineration de-
future scenarios for FW treatment considered in the study. mand assumed in this scenario.
While there are plans to legislate the provision of FW collections in
2.2. Current situation and future scenarios England, which would likely see an increase in separate FW collections
(DEFRA, 2018a), these are yet to be approved. The current situation in
2.2.1. Current situation Wales (see Fig. 3) shows that it is possible to increase significantly the
UK households currently generate 7.3 Mt of FW each year (Quested FW collected separately through legislation. However, one council in
and Parry, 2017) which consists of: England has terminated their separate FW collections citing high costs
and poor uptake by residents (Letsrecycle, 2018). For these reasons, it
• 4.4 Mt of avoidable waste (food intended for human consumption); is assumed that the government is unsuccessful in increasing separate
• 1.3 Mt of possibly avoidable waste (food consumed on a preferential FW collections but continues to phase out landfill.
basis, such as potato skin and bread crust); and As indicated in Table 2, the share of FW sent to AD (6%) and IVC
• 1.6 Mt of unavoidable waste (inedible part of food, such as egg shells (10%) remains at current levels in this scenario, while incineration in-
and bones). creases to 74% and landfilling falls to 10%. Two sub-scenarios are also
considered. The first (S1-A) assumes no heat is utilised from AD or in-
cineration, while the second (S1-B) utilises the heat from AD and incin-
Local authorities collect 4.9 Mt out of total 7.3 Mt of FW generated, eration for district heating or local industry, where it displaces heat from
either as targeted collections (separately or co-mingled with garden natural gas. Although the use of natural gas for electricity generation is
waste) or mixed within MSW. In 2016, 50% of the collected FW was projected to fall by 2030, the projections for natural gas use, including

Construction materials/costs
Collection
Household
service/ Gate fees
food waste Operating materials/costs
costs

Treatment

Anaerobic In-vessel
Incineration Landfill
digestion composting

Incentives Compost/digestate Heat/biogas export Electricity export

credits/revenue credits/revenue credits/revenue

Fig. 2. System boundaries for the food waste treatment scenarios life cycle assessment (LCA) and life cycle costing (LCC). [The dashed lines refer to costs only].
 P.C. Slorach et al. / Science of the Total Environment 710 (2020) 135580 5

Collected separately Collected within garden waste Mixed with MSW

100%

80%
55%
FW collected (%)

60% 76% 76%

86% 83%

40% 4%

20% 8% 41%
5% 20% 6%
16% 11%
9% 4%
0%
England Scotland Wales Northern Ireland UK

Fig. 3. Collection options for food waste (FW) managed by local authorities in the UK (WasteDataFlow, 2018; Office for National Statistics, 2017).

by industry and for residential heating, show no significant decline by The assumed treatment mix in this scenario is 40% AD, 10% IVC, 45%
2030 (BEIS, 2018b). incineration and 5% landfilling (see Table 2). Like S1, this scenario is also
divided into two sub-scenarios: S2-A assumes no heat utilisation from
2.2.2.3. Scenario 2-A/B: 50% in targeted FW collections and AD prioritised. AD or incineration and S2-B considers the heat to be used for district
In this scenario (S2), 50% of FW is managed in targeted collection, heating or by local industry, displacing natural gas.
with the majority collected separately and treated via AD. This corre-
sponds to the current situation in Wales, where 99% of households are 2.2.2.4. Scenario 3: 50% of FW in targeted collections and IVC prioritised.
provided with separate FW collections to prioritise AD (Welsh Scenario 3 (S3) mirrors S2, with the exception that IVC provides the
Government, 2011). The existing capacity for IVC is maintained, as the new capacity for targeted FW treatment. The total amount of waste
majority of current plants will still be in operation. sent to all forms of composting by local authorities increased moder-
To meet recycling targets and make use of FW in a circular economy, ately from 3 Mt in 2007 to 4.3 Mt in 2014; approximately two thirds
the waste must be collected separately. Currently, the separate FW col- of this was green waste treated at open-air (windrow composting) facil-
lection rates differ across the UK (Fig. 3). For example, Wales has the ities (Burns and Hopwood, 2017). Woody green waste is not suited to
highest separate collection rates of 33.5 kg FW per person, which repre- treatment by AD and therefore mixed garden and FW typically requires
sents approximately 45% of the total FW collected by local authorities IVC.
per capita (74 kg/person), compared to the average of 17% for the Despite the collections of mixed garden and FW remaining constant
whole of the UK. With the government in England planning to legislate between 2014 and 2016 (WasteDataFlow, 2018), it is possible that En-
for the provision of separate FW collections for all households by 2023 glish councils could provide targeted FW collections through mixed col-
(DEFRA, 2018a), collecting separately at least 50% of FW across the UK lections. With reduced incentives for electricity from AD (Ofgem,
is a realistic 2030 target. 2018a), this option could become less attractive and more expensive
The current feedstock capacity for waste-fed AD plants is 6.5 Mt per than IVC. Thus, the share of AD in S3 remains stable at 6%, while IVC
year (NNFCC, 2019a), of which approximately 0.3 Mt is household FW goes up to 44%. The remaining waste is mainly incinerated (45%) and
(Burns et al., 2017). Eunomia (2014) forecasts a surplus AD capacity to- only 5% is landfilled (Table 2).
wards 2024, which will necessitate an increase in separate household
FW collections to utilise the available capacity. 2.2.2.5. Scenario 4-A/B/C: 95% of FW in targeted collections and AD
Since landfilling of organic waste has been banned in Sweden (EEA, prioritised. Scenario 4 (S4) assumes that almost all (92%) household
2013b) and Germany prohibits the landfilling of untreated organic FW is collected separately, with a small fraction (3%) mixed with garden
waste (EEA, 2013a), it is assumed in S2 that the landfilling of FW waste and treated via IVC; this matches the current treatment ratio of
would be prohibited in the UK by 2030. Total elimination of food from targeted food collections in Wales (WasteDataFlow, 2018). A small frac-
landfills is not practical for at least two reasons: some food residues re- tion (5%) of FW is assumed to remain in the general waste, due to prac-
main in packaging and full compliance will be difficult to achieve. There- tical difficulties in full separation from packaging and lack of consumer
fore, 5% of FW may still end up in landfill. As discussed for S1, there will compliance. The ratio of incineration to landfilling matches S2 and S3.
be continued growth in incineration for MSW in S2. Therefore, the treatment mix for S4 is 92% AD, 3% IVC, 4% incineration

Table 2
Current situation (2016) and future scenarios (2030) for household food waste treatment.

Scenario Provision of targeted collections (% of UK households) Treatment mix

a
Separate food waste Co-mingled with garden waste Anaerobic digestion In-vessel compostingb Incinerationc Landfillc

Current situation 29% 20% 6% 10% 50% 34%

BAU 29% 20% 6% 10% 50% 34%
S1-A/Bd 29% 20% 6% 10% 74% 10%
S2-A/Bd 80% 20% 40% 10% 45% 5%
S3 29% 71% 6% 44% 45% 5%
S4-A/B/Cd 100% 5% 92% 3% 4% 1%
a
All food waste treated via AD is assumed to be collected separately.
b
All food waste treated via IVC is assumed to be collected co-mingled with garden waste.
c
All food waste incinerated or landfilled is assumed to be collected co-mingled with municipal solid waste.
d
Sub-scenarios “A” assume that recovered heat is not utilised, while “B” consider heat utilisation. Sub-scenario “C” refers to upgrading of biogas for injection into the gas network.
6 P.C. Slorach et al. / Science of the Total Environment 710 (2020) 135580

and 1% landfill (Table 2). As AD has the lowest LCA impacts among the Table 3
FW treatment methods (Bernstad and la Cour Jansen, 2011; Oldfield Process improvements for anaerobic digestion of food waste (FW) in the 2030 scenarios
compared to the current situationa.
et al., 2016; Slorach et al., 2019b), this scenario represents environmen-
tally the best possible situation. Parameter Unit Current Current 2030 Reference
Three sub-scenarios are considered: S4-A assumes no use of heat situation situation scenarios
(average) (range)
from AD and incineration; S4-B utilises the heat from both for district
heating or for local industry; and S4-C considers biogas upgrading for Biogas Nm3/t FW 137 85–187 162 Bernstad and la
production Cour Jansen (2011);
injection into the national gas network to displace natural gas.
Banks et al. (2011);
Monson et al.
2.3. Inventory data and assumptions (2007)
Electricity kWh/t 277 172–378 328 Ecoinvent (2016)
This section provides an overview of the four waste treatment sys- generated FW
Electricity kWh/t 254 128–368 311 ~II~
tems considered in the scenarios, including the inventory data and as- exported FW
sumptions made based on the changes expected by 2030. The full Heat kWh/t 476 295–649 563 ~II~
inventory tables for the LCA (Tables S2, S5, S7 and S9) and LCC generated FW
(Tables S1, S4, S6, S8 & S10) can be found in the Supplementary infor- Heat kWh/t 0 252 ~II~
exported FW
mation (SI). The life cycle models for the individual treatment tech-
Parasitic kWh/t 23 10–44 17 Bernstad and la
niques considered in the future scenarios have been adapted and electricity FW Cour Jansen (2011);
developed further based on a previous work by the authors (Slorach Banks et al. (2011);
et al., 2019a), which considered the sustainability of these options for Monson et al.
the present situation only. (2007)
Parasitic kWh/t 82 36–113 59 ~II~
heat FW
2.3.1. Waste collection Methane % of 2 1–3 1 Bernstad and la
Currently, kerbside waste is collected by diesel vehicles. The UK gov- emissions produced Cour Jansen (2011);
ernment will ban the sale of diesel and petrol cars and vans by 2040, CH4 Evangelisti et al.
(2014); Naroznova
aiming for almost all these vehicles to have zero emissions by 2050
et al. (2016)
(Department for Transport, 2018). However, zero-emission technolo-
a
gies for heavy goods vehicles (HGVs) are not widely available or devel- See Table S2 in the Supplementary information for full inventory data.

oped so it is expected that they will continue to use diesel. The

government set out a voluntary target to reduce GHG emissions from (Table 3). This sees the biogas yield increase by 18% and exported elec-
HGVs by 15% by 2025 (Department for Transport, 2018). However, as tricity by 22%. These assumptions are similar to the learning curves ap-
there is no clear pathway for achieving this goal, this consideration is plied by the European Commission for forecasting biogas production
not included in the model. from AD in 2030 (EC, 2016), with a 20% increase in the yield and a
Therefore, in all scenarios, FW is assumed to be collected from 20% reduction in the parasitic energy use. The fugitive methane emis-
households by a 21 t diesel waste-collection lorry, travelling an average sions are also assumed to reduce from 2% to 1% of the produced biogas.
distance of 20 km to the treatment facility. The Ecoinvent (2016) trans- This could be achieved by the use of flares, higher quality seals and cov-
portation model has been used to estimate the environmental impacts. ered storage of digestate (Edelmann et al., 2005). The assumptions for
The costs of collection have been determined via a costing tool used by digestate remain the same as at present since FW composition is not ex-
local authorities (WRAP, 2018b). Based on this tool, the average annual pected to change significantly and the fertiliser displacement is based
cost per household for collection of general waste, excluding labour, is on nutrient content. The sub-scenarios which include heat utilisation
£13.81 and for separate FW £7.13. The cost per tonne is dependent on (S1-B, S2-B and S4-B) assume that 50% of the excess heat is used for dis-
the volume of FW collected; these are detailed in Table S1 in the SI. trict heating or industry, displacing natural gas. The infrastructure for
As shown in Table 2, BAU and S1 assume that 29% of UK households heat distribution is not considered.
have separate FW collections and 20% co-mingled with garden waste, as For the upgrading of biogas to biomethane in S4-C, a high-pressure
is the case at present (WRAP, 2016b). S2, S3 and S4 all include the pro- water scrubbing process with compression for grid injection has been
vision of targeted FW collections to all households, with an increase in modelled, as described by Budzianowski et al. (2017). The latter pro-
separate collections when AD is prioritised (S2-A/B and S4-A/B/C) and vides inventory values for upgrading 1000 and 250 Nm3/h of biogas –
an increase in collections of FW co-mingled with garden waste when linear scaling has been used to attain vessel sizes and performance pa-
IVC is prioritised (S3). rameters for the 419 Nm3/h considered in this study. Material require-
ments for the equipment are based on the pressure vessel design by
2.3.2. Anaerobic digestion Towler and Sinnott (2013) using AISI 304 stainless steel. Ecoinvent
Currently, 0.31 Mt of separately collected FW in the UK is treated via (2016) data have been used for pumps and compressors. The plant elec-
AD (Burns et al., 2017). The representative AD plant used in the study is tricity is provided by the CHP, using 17% of the produced biogas. An
a commercial facility with a high-solids 2500 m3 continuous mesophilic overview of the biogas upgrading process is given in Fig. S1 and the in-
digester, treating 25 kt of feedstock per year. The produced biogas is ventory data in Table S3.
used in a combined heat and power (CHP) engine to generate heat The feed-in tariff (FIT) scheme for exported electricity closed to new
and electricity. Some of the heat and electricity is used to operate the applicants in April 2019 (Ofgem, 2018a). Therefore, no electricity incen-
plant. At present, only electricity is exported to the grid and excess tives are included in the 2030 scenarios. However, the Renewable Heat
heat is wasted; the potential utilisation of the latter is considered in Incentive (RHI) now offers 5.60 p/kWh for the first 40 GWh of
some scenarios. The digestate is used as a fertiliser. The system has biomethane injected into the grid (Ofgem, 2018d), this is expected to
been credited for displacing grid electricity, heat from natural gas increase the share of facilities upgrading biogas (NNFCC, 2019b). The
(where considered) and mineral fertilisers. For further details, see RHI also incentivises heat exports, but currently in the UK it is rarely
Slorach et al. (2019b). possible to export the heat directly (Scholes and Areikin, 2014). Future
For the 2030 scenarios, biogas production rate and heat and electric- heat utilisation could be improved through a growth in district heating
ity use are assumed to improve with new plants coming online, assum- schemes and better pairing with industrial users. The domestic RHI is
ing a value half way between the current averages and upper ranges set to end in 2021 but there are no plans to close the non-domestic
 P.C. Slorach et al. / Science of the Total Environment 710 (2020) 135580 7

users (Ofgem, 2018c). The sub-scenario that include heat use (S1-B, S2- 2.3.4. Incineration with energy recovery
B and S4-B/C) assume that RHI is still applicable at the current rate of A moving-grate furnace with an annual capacity of 300 kt of waste is
1.36 p/kWh of heat exported (Ofgem, 2018e) and 5.60 p/kWh considered for incineration of FW. Some of the recovered electricity is
biomethane injected into the grid. used on site and the rest is exported, crediting the system for displacing
The current average gate fee paid by local authorities for AD facilities grid electricity. Similar to AD, utilisation of heat from incineration is cur-
is £29/t (WRAP, 2018a), but this has decreased from £40/t in 2016 rently limited, with only three out of 31 plants operating in the UK
(WRAP, 2016a). The current pricing is inflated by historical local author- exporting any significant amount of steam or hot water (Environment
ity contracts. By comparison, the present median spot price is £15/t Agency, 2016). The flue gas is treated to remove particulate matter,
(WRAP, 2018a). The future scenarios assume that the average fees will acids as well as sulphur and nitrogen oxides. For a detailed process de-
continue to fall, reaching £15/t by 2030. The revenue changes related scription, see Slorach et al. (2019a).
to incentives and sales are summarised in Table 4. The changes to the With waste-to-energy incineration forecasted to grow and the UK
electricity export tariff are discussed in Section 2.3.6. The plant capital government targeting improvements in its efficiency (DEFRA, 2018a),
and operating costs have been modelled using current values as these the 2030 scenarios assume a 16% increase in electricity produced per
are not expected to change over the next ten years. tonne of FW (from 255 to 297 kWh) and a 20% fall in parasitic load
(from 81 to 65 kWh). These values are half way between the current av-
erage and favourable range limit (see Table 5).
2.3.3. In-vessel composting Utilisation of heat is also assumed in some scenarios. This is based on
The representative IVC plant in this study has a capacity to treat 50 the design values for the Sheffield Energy Recovery Facility, one of the
kt/yr. The FW is shredded before entering the rotating drum UK's few operating plants that is connected to a district heating network
composters, where it is retained for four days. The compost is subse- (Environment Agency, 2016). The plant has an annual capacity of 245 kt
quently steam-pasteurised at 70 °C to ensure compliance with legisla- of waste, the availability of 91.9% and the design value for heat export is
tion (UK Government, 2014). The compost is then matured in open air 60 MW. In 2016, it exported 113 GWh of heat to the district heating sys-
windrows before being used as a fertiliser in agriculture; the system tem, equivalent to 25% of the design value (Environment Agency, 2016).
has been credited for displacing an equivalent amount of mineral For the 2030 scenario, it is assumed that 50% of the design value is
fertilisers. Wet scrubbers and a biofilter are used in the IVC facility to re- achieved, resulting in 986 kWh/t of heat from all the waste treated.
move pollutants and odours, particularly ammonia and hydrogen sul- When scaled for the lower heating value of FW (3.8 MJ/kg) compared
phide. IVC is a high user of electricity due to the air-handling system to the MSW (8.9 MJ/kg), the resulting heat export value is 421 kWh/t
for maintaining the controlled atmosphere; for further details, see FW.
Slorach et al. (2019a). The average incineration gate fee has not changed much over the last
In the future scenarios, the direct emissions from composting are not three years, fluctuating between £83/t and £86/t (WRAP, 2017a; WRAP,
assumed to change but the electricity consumption could be reduced. 2018a). However, as landfill capacity reduces, gate fee for incineration is
This could be achieved through a wider adoption of membrane- expected to increase (WRAP, 2018a). A moderate increase to £95/t is as-
covered piles rather than the energy-intensive air-handling system sumed in 2030. The capital and operating costs for incineration are un-
(Net Zero Waste, 2018). Like the AD system, future electricity consump- changed in the future scenario; the only change is the revenue from
tion is chosen half way between the current average and the lowest electricity and heat exported, which have the same export prices as
range, which sees it decrease by 33% (from 93 to 62 kWh/t FW). AD of £0.055/kWhe and £0.03/kWhth, respectively (see Table 4). The
The average gate fee for IVC has remained stable over recent years, full LCA and LCC inventories can be found in Tables S7 and S8 in the SI.
increasing from £43/t in 2011 (WRAP, 2011) to £46/t in 2017 (WRAP,
2018a). No significant increase is expected in the future and hence a 2.3.5. Landfill with landfill gas utilisation
gate fee of £50/t is assumed for 2030. The cost of electricity increases The remaining MSW that is not incinerated is landfilled. Under UK
in the future scenarios by 19% (0.0973 to 0.116 £/kWhe (BEIS, 2019). regulations, the landfill gas (LFG) must be collected for utilisation or
There are no other changes on the current situation. For the full LCA flaring and leachate must be treated (Environment Agency, 2009). The
and LCC inventory data, see Tables S5 and S6 in the SI, respectively. landfill modelled here covers an area of 90,000 m2 and is 20 m deep,
with captured LFG used to generate electricity in a gas engine; the sys-
tem has been credited for displacing grid electricity. Leachate is con-
veyed to a wastewater treatment plant via sewer. A full description of
Table 4
Changes to revenue and incentives for the life cycle costing of anaerobic digestion in the
the landfill model can be found in Slorach et al. (2019a).
future scenarios compared to the current situationa. In the future scenarios, the current average LFG capture rate of 68%
(55–85%) increases to 76% as older landfill sites close and newer sites
Parameter Unit Current Future Reference
take on a greater share of landfilling. Consequently, the exported elec-
situation scenarios
tricity increases by 12% (from 68 to 76 kWh/t FW) and the export
Gate fee £/t FW 29 15 WRAP
sales price by 12% (from £0.050/kWhe to £0.056/kWhe; see Table 4). Al-
(2017a);
WRAP (2018a) though stable at about £20/t, the landfill gate fee is expected to increase
Feed-in tariff (0.5–5 MW £/kWhe 0.016 0 Ofgem with reduced capacity (WRAP, 2018a) so the future scenarios assume a
facilities) (discontinued) (2018b) value of £30/t. Landfill tax is assumed to increase from £84.40/t at pres-
Export tariff £/kWhe 0.050 0.056 Ofgem ent to £90/t by 2030. The current increases in landfill tax are in line with
(2018b); BEIS
the retail price index, but the longer-term trend is an increase in real
(2019)
Renewable heat incentive: £/kWhth No heat 0.0136 Ofgem terms and therefore a modest increase is assumed for 2030 (HMRC,
exported heat exported (2018e) 2018).
Renewable heat incentive: £/kWhth No gas 0.056 Ofgem There are no other changes to the landfilling option in the future sce-
exported biomethane to exported (2018d)
narios compared to the current situation. Full inventory data are pro-
grid
Heat export price £/kWhth No heat 0.03 Based on CMA vided in Tables S9 and S10 in the SI.
exported (2018)b
a
See Table S4 in the SI for full LCC data.
2.3.6. Future grid electricity
b
Limited current market for heat sales – value based on 50% of the £0.06/kWhth paid by As mentioned earlier, AD, incineration and landfilling are all credited
current domestic consumers. for supplying electricity to the grid and IVC is a significant user of
8 P.C. Slorach et al. / Science of the Total Environment 710 (2020) 135580

Table 5 Table 7
Process improvements for the incineration of food waste in the future scenarios compared Global warming potential (GWP) of UK household food waste taking into account the full
to the current situation.a life cycle of food production (Quested and Parry, 2017; Quested et al., 2013).

Parameter Unit Current Current Future Reference GWP and Agriculture Processing Logistics Packaging Use Total
situation situation scenarios contribution of
(average) (range) life cycle stages

Gross kWh/t 255 172–338 297 Environment Agency GWP (t CO₂

2.00 0.36 0.86 0.18 0.35 3.74
electricity FW (2016) eq./t)
produced Contribution
53% 10% 23% 5% 9% 100%
Electricity kWh/t 81 48–179 65 Environment Agency (%)
consumed FW (2016)
on site
Heat kWh/t 0b 421 Based on the capacity that, if an amount of waste can be prevented, then the ultimate conse-
exported FW of Sheffield incinerator
(Environment Agency,
quence would be to avoid producing an equal amount of food.
2016) The analysis considers two potential cases for FW prevention by
a 2030: reduction by 20% and by 50%. The former is based on the UK
See Table S7 in the Supplementary information for full LCA inventory data.
b
The actual average is 11 kWh/t FW but as only three plants out of 31 export heat, this Courtauld Commitment for reducing FW across the supply chain by
is not considered for the current situation. 2025 (WRAP, 2017b) and the latter on the UN target for reduction of re-
tail and consumer FW (UN, 2015). In addition, an extreme case is con-
sidered with the prevention of all avoidable or possibly avoidable
electricity. Therefore, the make-up of the UK grid mix can have a large waste to determine the maximum savings in GHG emissions if it was
influence on the environmental impacts of the treatment options. possible to achieve this goal. The minimum waste reduction required
Decarbonisation of the grid will lead to lower credits and hence higher to match the GWP of the best waste treatment scenarios is also
impacts from AD, incineration and landfilling. Conversely, IVC will ben- identified.
efit from a decarbonised electricity mix.
The UK grid mix has changed extensively over the last two decades. 3. Results and discussion
For example, the percentage of electricity supplied by coal was 30.9% in
2000 and peaked at 37.1% in 2006, but in 2017, this reduced to 6.5%. In This section discusses the environmental and economic sustainabil-
the same period, the share of wind, solar and bioenergy increased from ity of different scenarios for FW treatment in comparison with the cur-
1.1% to 26.8% (BEIS, 2018a). A simplified 2017 UK electricity mix is rent situation. The results are presented first for the functional unit of 1 t
shown in Table 6, along with two future scenarios proposed by the of FW treated, followed in Section 3.2 by the sustainability implications
National Grid (2018). Their “Steady Progression” scenario is considered at the UK level. The latter also includes a comparison of FW treatment to
here in the base case, where a large growth in offshore wind power is its prevention.
expected, while maintaining significant use of natural gas. The “Two De-
grees” scenario is evaluated in the sensitivity analysis. This scenario 3.1. Sustainability implications per tonne of food waste treated
achieves the highest decarbonisation by maintaining a higher nuclear
fraction and reducing the reliance on natural gas, while expanding sig- 3.1.1. Environmental sustainability
nificantly the share of renewables. The environmental impacts of the future scenarios are compared to
The electricity mixes have been modelled based on Ecoinvent the current situation in Fig. 4; for further details on these as well as on
(2016) data; any future changes to electricity technologies, such as effi- the individual treatment options, see Tables S11–S13 in the SI.
ciencies and resource use, have not been considered due to high levels As it can be observed in Fig. 4, S4-B and S4-C each have the lowest
of uncertainty in projections. The 2030 electricity costs projected by environmental impacts in seven of the 19 categories. For S4-B, this in-
the UK government have been used to estimate the LCC. Specifically, cludes primary energy demand and metal depletion, and for S4-C,
the electricity cost to the operators of IVC is assumed at £0.116/kWh GWP and fossil depletion. S4-B displaces the highest amount of grid
and revenue for the AD, incineration and landfill operators for selling electricity and both scenarios benefit from displacing natural gas. The
electricity to the grid at £0.056/kWh (BEIS, 2019). credits for the latter reduce significantly the impacts for these as well
as the other scenarios exporting heat (S1-B and S2-B), especially for
2.4. Waste prevention GWP and depletion of fossil resources.
The current situation has the lowest particulate matter formation,
To explore the impacts of waste prevention, it is necessary to assign marine ecotoxicity and ionising radiation, S1-B has the smallest terres-
upstream food impacts to the FW, which includes the agriculture, pro- trial acidification and S3 is the best option for marine eutrophication.
cessing, supply and consumption stages. Owing to the lack of data, the Scenarios S3 is the worst in six categories (primary energy demand, fos-
analysis is limited to global warming potential (GWP). The total up- sil depletion, freshwater and marine ecotoxicities, ozone depletion and
stream GWP of 1 t of avoidable or possibly avoidable FW is estimated natural land transformation) and second worst in further two impacts
at 3.74 t CO2 eq. (Table 7). The impact is allocated entirely to the edible (metal depletion and ionising radiation). BAU has the highest GWP,
fractions so the unavoidable fraction of FW is impact-free. It is assumed human toxicity and marine ecotoxicity and is the second worst option
for eight other impacts.
Table 6 If there is no utilisation of recovered heat or upgrading of biogas to
Current UK electricity mix (BEIS, 2018a) and the National Grid 2030 scenarios “Steady Pro- replace natural gas (i.e. only S1-A, S2-A, S3 and S4-A are considered),
gression” and “Two Degrees” (National Grid, 2018) considered in the food waste treat- S4-A performs the best in 13 of the 19 indicators. However, the current
ment scenarios.
situation is still best in five categories. When heat can be utilised (“B”
Electricity mix Coal Oil Gas Nuclear Renewables Storage Other and “C” sub-scenarios), this decreases to three impacts. The next sub-
fuels sections give an overview of the results for each impact in turn.
Current (2017) 6.7% 0.7% 39.7% 20.9% 35.7% 0.9% 1.7%
“Steady 0.0% 0.0% 24.5% 7.3% 64.7% 0.0% 3.4% 3.1.1.1. Primary energy demand (PED). All scenarios are net producers of
Progression” electricity, with the exported amount ranging from 0.5 (S3) to 1.1 GJ/t
“Two Degrees” 0.0% 0.0% 11.4% 16.9% 65.1% 0.0% 4.0%
FW (S4-A/B). Where applicable, the heat exported ranges from 0.9
 P.C. Slorach et al. / Science of the Total Environment 710 (2020) 135580 9

(S4-B) to 1.2 GJ/t FW (S1-B). The AD system generates the most electric- being 6% higher than for the current situation. As shown in Fig. 4, the
ity (1.12 GJ/t FW) and when the heat from the CHP engine can be scenarios displacing natural gas with either heat or biomethane (S1-B,
utilised, it has the greatest PED credits of all the treatment options. S2-B and S4-B/C) result in net savings of −0.2 to −29.1 kg CO2 eq./t
With 92% of the FW treated via AD in S4-B, this scenario has the lowest FW. This is because the electricity generated from waste has a higher
PED (−2.2 GJ). S2-B exports both heat and electricity from incineration GWP than the projected 2030 electricity mix; therefore, net savings
and AD and so has the second lowest PED (−1.7 GJ). are only observed when heat can be utilised to displace natural gas.
If heat cannot be exported, S4-A has the lowest PED at −1.61 GJ/t The BAU scenario has the highest GWP as it retains the current levels
FW, followed by S4-C (−1.55 GJ/t FW). IVC is the only treatment system of landfilling and reduced electricity credits for the low-carbon 2030
that is a net user of electricity and, since it makes up 44% of S3, this sce- grid mix. All other scenarios have a lower GWP as they include a reduc-
nario has the highest PED (0.11 GJ/t FW). It is also the only scenario with tion in landfilling, thus reducing the associated emissions of methane.
a greater PED than the current situation (−0.39 GJ/t FW). The PED trend
is also reflected in many other impacts, as discussed below. 3.1.1.3. Fossil depletion (FD). Similar to GWP, the only net savings in FD
are found for S4-A (−4.6 kg oil eq./t FW) and all sub-scenarios that dis-
3.1.1.2. Global warming potential (GWP). The estimated GWP ranges place fossil fuels through heat utilisation: S4-B/C (−16.7 to −32 kg oil
from −29.1 (S4-C) to 73.8 kg CO2 eq./t FW (BAU), with the latter eq./t FW), S1-B (−11 kg oil eq./t FW) and S2-B (−12.7 kg oil eq./t

Current situation BAU S1-A S1-B S2-A S2-B S3 S4-A S4-B S4-C

10.30
6.83

7.56
4.85
7.37
6.91

15

1.33
4.38
4.17

4.93
4.22
3.72
2.80

0.96

1.08

0.10
10
5
0
-0.020

-5
-0.93
-1.13

-2.24
-2.91
-3.94

-4.58
-3.89

-4.91
-10
-6.54

-10.33
-8.59
-9.91

-10.96
-15

-12.74
-14.01

-15.53

-20
-16.54

-16.66
-16.10

-25

-19.46
-21.77

-22.61
-30
-35

-31.99
PED x 0.1 GWP x 10 FD MD x 0.1
[GJ] [kg CO₂ eq.] [kg oil eq.] [kg Fe eq.]

6.10
10

5.09
4.73
2.49

2.07
1.76
1.60

1.22

0.55
0.50

5
0.33

0
-0.05

-0.09
-0.27
-0.34
-1.15
-1.25

-1.67

-1.84
-1.81

-2.22
-2.52

-2.53
-2.34
-3.49
-3.07

-5
-3.03
-3.50
-3.73
-3.99

-4.13

-4.25
-4.36
-4.74
-5.05

-4.16

-5.29
-5.48

-5.59
-5.83

-6.03

-5.58
-6.97

-10
-8.11
-8.94
-10.39
-10.89

-10.33
-13.00
-12.54

-15
FET x 0.1 MET x 0.1 TET HT FE
[kg 1,4-DB eq.] [kg 1,4-DB eq.] [g 1,4-DB eq.] [kg 1,4-DB eq.] [g P eq.]

15
9.87
9.61
9.19

7.77
7.54
6.41

7.13
7.27
7.19

7.08
6.51
6.72
7.01

6.12

5.83
5.27

10
3.63
5.12
4.27
4.17

3.42

3.85
3.31
3.28

3.37
2.65
2.84
2.35

2.31
2.65
2.65

1.89
1.78
1.84

1.55
1.79
0.92
0.92
0.75
0.74

0.73
0.73

0.81
0.64
0.73

5
0
-1.71

-5
-1.76
-2.76

-6.05
-4.81

-10
ME TA PMF x 0.1 POF x 0.1 OD
[kg N eq.] [kg SO₂ eq.] [kg PM10 eq.] [kg NMVOC eq.] [mg CFC-11 eq.]
1.22
0.94

0.43

0.02
0.05

0.07

2
0
-0.33
-0.02

-0.10
-0.05

-0.15
-0.21
-0.42

-0.54

-0.49
-0.37

-0.69
-0.74

-0.69

-2
-0.40

-0.95
-1.32
-1.06

-1.49

-0.67
-1.40

-0.91
-1.70
-1.72

-1.98

-1.06
-0.95
-1.16
-1.19

-4
-2.80
-3.54
-3.54

-1.59
-1.64
-1.65

-2.04
-4.20
-4.21

-2.34

-6
-2.58
-2.85
-5.56
-5.57

-8
-3.61
-3.83

-10
ALO x 10 ULO NLT x 0.01 IR x 10 WD x 100
[m²yr] [m²yr] [m²] [kg U235 eq.] [m³]

Fig. 4. Environmental impacts of the 2030 food-waste treatment scenarios in comparison with the current situation. [All impacts are expressed per tonne of food waste treated. The
scenarios are based on the “Steady Progression” 2030 electricity mix. Where shown, the ranges represent the impacts based on the Two Degrees scenario (see Table 6). Some impacts
have been scaled to fit and should be multiplied by the factor shown on the x-axis to obtain the original values. ALO: agricultural land occupation; FD: fossil depletion; FE: freshwater
eutrophication; FET: freshwater ecotoxicity; GWP: global warming potential; HT: human toxicity; IR: ionising radiation; MD: metal depletion; ME: marine eutrophication; MET:
marine ecotoxicity; NLT: natural land transformation; OD: ozone depletion; PED: primary energy demand; PMF: particulate matter formation; POF: photochemical oxidants formation;
TA: terrestrial acidification; TET: terrestrial ecotoxicity; WD: water depletion; ULO: urban land occupation; DB: dichlorobenzene; NMVOC: non-methane volatile organic compounds].
10 P.C. Slorach et al. / Science of the Total Environment 710 (2020) 135580

FW). Of the scenarios that do not recover heat, only S4-A produces the amount of energy used by and credited to the system. Consequently,
enough electricity to achieve a net FD saving. BAU, S1-A, S2-A and S3 S4-A/B/C, S1-B and S2-B exhibit net savings, ranging from −1.7 to
all deplete more fossil resources than the current situation. The latter −6.1 mg CFC-11 eq./t FW. The overall ranking of the scenarios is the
benefits from the credits associated with displacing the current electric- same as for PED.
ity grid which is dominated by fossil fuels.

3.1.1.4. Metal depletion (MD). The estimated MD ranges from −2.3 (S4- 3.1.1.9. Land use and transformation, water depletion and ionising radia-
B) to 0.5 kg Fe eq./t FW (S3), with all scenarios outperforming the cur- tion. All scenarios exhibit net savings in agricultural land occupation
rent situation. This is due to the greater credits for displacing grid elec- (−0.2 to −56 m2yr) and water depletion (−4.9 to −38.3 m3), which
tricity, as the future electricity mix requires more resources for the correlates with the amount of electricity used by and credited to the
increased share of wind turbines and solar photovoltaics (PV). However, treatment options. S4-B achieves the greatest savings in both impacts
in the case of IVC, this results in a greater impact from electricity use, due to its displacing the highest amount of electricity and utilising
causing S3 to have the highest MD. S1-A has the second highest MD heat. However, the credits for displacing natural gas have a limited con-
due to the high share of incineration, as the plant is resource intensive tribution to these impacts and on average they are only 3% lower than
to build and operate. Therefore, the scenarios with higher contribution for S4-A. All scenarios represent a better option than the current situa-
of incineration (S1-A/B) and IVC (S3) lead to the greatest depletion of tion for these two impacts, with the exception of S4-C, which exports
metals. no electricity.
Urban land occupation (ULO) ranges from −0.8 (S4-B) to 0.9 m2yr/t
3.1.1.5. Ecotoxicity and human toxicity. All scenarios have net savings in FW (BAU). Landfilling has the biggest ULO, leading to the highest land
freshwater ecotoxicity (FET), ranging from −0.1 to −1.3 kg 1,4-dichlo- requirement in BAU (and the current situation) due to the high share
robenzene (DB) eq./t FW. A similar trend is found for marine ecotoxicity of landfilling (34%). This is in contrast to natural land transformation
(MET) with a range of −0.1 to −0.9 kg 1,4-DB eq./t FW. However, the (NLT), where landfilling is credited for the land regenerated once the
only saving compared to the current situation is observed for S4-C landfill is closed (DEFRA, 2004). Displacing the future electricity mix in-
where natural gas is displaced by biomethane. The future grid electricity curs greater NLT credits and, therefore, BAU has lower impacts than the
has lower FET and MET than the current mix due to the reduction in fos- current situation. All other scenarios have decreased landfilling and it is
sil fuels, especially coal, and increase in wind power. Conversely, the only when heat is utilised (S1-B, S2-B and S4-B/C) that the NLT is lower
higher share of solar PV and wind increases the TET of the future grid than in the current situation.
mix due to the use of heavy metals. Therefore, all scenarios have a All the scenarios and the current situation have a net-negative ionis-
lower TET than the current situation because of the higher credits for ing radiation (IR) due to the credits for electricity. However, the lower
displacing future electricity mix, ranging from −0.05 (BAU) to −5.6 g share of nuclear power in the future electricity mix results in all scenar-
1,4-DB eq./t FW (S4-B). ios receiving 4–75% lower IR credits than currently. As a result, the cur-
The reduction in coal and increase in wind also reduce the human rent situation is the best option for this impact.
toxicity (HT) of the future electricity mix, which leads to reduced credits
for exported electricity. This results in BAU having a higher HT than the 3.1.1.10. Sensitivity analysis. As mentioned earlier, the above results
current situation, but the increase in exported electricity and/or heat is (”base case”) correspond to the 2030 electricity mix “Steady Progres-
sufficient to generate net savings in all other scenarios. sion” as defined by the National Grid (2018). This section considers
the effect on the impacts of assuming the “Two degrees” mix instead
3.1.1.6. Freshwater and marine eutrophication (FE and ME). S1-A has the (see Table 6). The results in Fig. 4 and Table S12 show that the increased
highest FE (6.1 g P eq./t FW), which is 145% greater than the impact share of renewables and nuclear power has little effect on the impact
for the current situation. Incineration, which makes up 74% of the treat- trends and the ranking of the scenarios. Compared to using the “Steady
ment mix in this scenario, is the main contributor, with an impact more progression” electricity mix, there are greater PED savings, ranging from
than double that of landfill. On the other hand, AD and IVC both achieve 8% (S1-B, S2-B and S4-B) to 56% (S-3). However, the lower GWP of the
net savings in FE (−4.9 and −4 g P eq./t FW, respectively). Nevertheless, “Two Degrees” results in lower credits for exported electricity and thus
only S4-A/B/C provide net savings due the large share of AD. higher GWP for all scenarios than for the case of “Steady Progression”.
ME is primarily caused by landfill leachate, causing the current situ- The credits for MD, HT, FET, MET and WD all improve as renewable
ation and BAU to have the highest impact. Even though all scenarios are technologies, especially solar PV, require more material resources than
net contributors to ME, notable reductions of 65% (S1-A/B) to 76% (S3) natural gas. There is negligible change to POF, ME, TA and PMF as they
can be achieved compared to the current situation. are predominantly driven by direct emissions from the treatment plants
rather than the electricity mix. S4-C is not affected as it neither uses nor
3.1.1.7. Terrestrial acidification (TA) and particulate matter formation exports electricity.
(PMF). These two impacts are driven by ammonia emissions from AD
and IVC and are hence not affected by energy credits as observed for
most other impacts. Owing to the large share of AD (93%), S4-A has 3.1.2. Economic sustainability
the highest TA (7.3 kg SO2 eq./t FW) and PMF (1 kg PM10 eq./t FW). All costs discussed in this section are based on the value of the Pound
Compared to S4-A/B/C, the current situation, BAU and S1-A/B all have Sterling (£) in 2016. For context, the average annual exchange rate for
TA four times lower and PMF almost three times smaller because of the Euro was £1 = €1.22 and for the US Dollar £1 = $1.35 (Office for
the low fraction of IVC and AD. National Statistics, 2019).
As indicated in Fig. 5, S4-C has the lowest LCC (£23/t FW) due to the
3.1.1.8. Photochemical oxidant formation (POF) and ozone depletion (OD). incentives for the injection of biomethane to the gas grid. BAU and the
S1-A has the highest POF, estimated at 0.78 kg non-methane volatile or- current situation have the highest costs (£92/t FW), related primarily
ganic compounds (NMVOC) eq./t FW; this is 19% higher than for the to the landfill tax. The increase in separate collections of FW pushes
current situation. Incineration is the biggest contributor to this impact up the collection costs in scenarios S2 and S3. However, the reduced
because of flue gas emissions of NMVOCs. gate fees related to IVC and AD compared to incineration and landfilling
Transportation is the main cause of OD, contributing between 69% mean that the total local-authority costs in BAU and S1–3 remain similar
(S3) to 94% (S4-A/B/C) of the total impact. As the transport does not to the current situation. Local authorities only make a significant saving
vary significantly between scenarios, the OD is determined mainly by (47%) when separate food collections are greatly increased, as in S4.
 P.C. Slorach et al. / Science of the Total Environment 710 (2020) 135580 11

Disposal operator costs Local authority costs

Disposal revenue - gate fees Disposal revenue - Disposal revenue - Disposal - operating
sales incentives
Disposal - capital Gate fees Collection costs
£67
£180 £92 £77 (£67)
£92 £69
£79 (£71) £75
£34 £34 £34
£34
£130 £44
£53 £53
£23
£37
(£60)
£88 £88 £48 (£40)
£80 £94
LCC (£/t)

£84
£72 £43
£60 £60 £43 £43

£30 £23 £23 £20

£19 £19 £17 £17 £16 £20 £20
£9
£16 £16 £21 £21 £17 £17 £18 £8 £8
£10 £10 £16
-£19 -£19 -£19
-£20 -£8
-£56 -£55 -£55 -£14 -£21
-£63 -£67
-£79 -£79 -£3
-£37
-£6 -£12
-£70 -£0.28 -£9 -£20 -£7
-£1
-£11 -£20
-£0.22
-£120
Current BAU S1-A S1-B S2-A S2-B S3 S4-A S4-B S4-C
situation

Fig. 5. Life cycle costs (LCC) of the 2030 food-waste treatment scenarios in comparison with the current situation. [The LCC of the scenarios without incentives are shown in italics in
brackets below the total values].

Without government incentives, which may no longer be active in In summary, local authorities can only make significant savings
2030, S4-B becomes the option with the lowest LCC (£37/t FW). BAU (£56/t FW) if large quantities of FW are collected separately, as found
is unaffected by incentives and still has the highest LCC (Fig. 5). in S4-A/B/C. Operators make the most profits from incineration, which
means that S1-B is the most profitable option for the operator. Combin-
3.1.2.1. Costs to local authorities. The collection costs are the lowest when ing the costs to both stakeholders, S4-C has the lowest LCC but, without
FW can be collected in bulk. Thus, the current situation, BAU and S1-A/B incentives, S4-B is the best option.
have the lowest costs (£34/t FW; see Fig. 5) as the majority of FW is col-
lected together with the rest of the MSW. However, S4-A/B/C are the 3.1.2.4. Sensitivity analysis. To evaluate the effect of some costs and as-
second best options (£43/t FW) as the costs of collecting FW separately sumptions on the LCC, the following parameters that could influence
reduce from the current £108 to £42/t when 95% of all household FW is the results are considered in the sensitivity analysis:
collected separately.
The high costs of separate FW collections are counteracted by lower i) waste collection costs;
gate fee for AD (£15/t FW), which leads to the lowest overall gate fees ii) gate fees; and
for S4-A/B/C (£20/t FW) owing to 93% of FW being processed via AD. iii) electricity prices.
This results in the overall lowest local-authority cost of £63/t FW for
these scenarios. The highest total costs are found for BAU because of
the dominance of landfilling (gate fee of £120/t FW) and incineration
They are each varied arbitrarily by ±25% and the results are
(£95/t FW).
discussed in the next sections.

3.1.2.2. Costs to operators. S1-A/B have both the highest capital (£23/t 3.1.2.5. Waste collection costs. The effect of varying the cost of each
FW) and operating (£21/t FW) costs, giving a total cost to operators of collection option (separate FW, co-mingled with garden waste and
£44/t. This is due to high capital and operating costs of incineration mixed with MSW,) is shown in Fig. S2 in the SI. It can be seen that
(£54/t FW), which treats 74% of FW in these scenarios. IVC requires there is no change in the ranking of the scenarios with changes to
the lowest expenditure (£14/t FW) which benefits S3 but is MSW collection costs. However, if the cost of separate FW collec-
counteracted by the incineration costs, leading to a total of £33/t FW. tions is reduced by 25%, the ranking of S2-A and S2-B is improved
S4-A/B are the least costly for the operator (£18/t FW) due to the low due to their 40% AD share, which requires separate collections. S2-
AD costs (£16/t FW) and its high fraction in these scenario (92%); they B increases its ranking from fifth to fourth, becoming more
also benefit from the smallest share of incineration (4%). S4-C has 40% favourable than S1-B. S2-A becomes the sixth best options, ranked
higher total operating costs than S4-A/B due to the additional infrastruc- above S3 and S1-A. The only consequence of changing the collection
ture and processes needed to upgrade the biogas to biomethane. costs of FW co-mingled with garden waste is that S3 falls from the
sixth to seventh place if the cost is increased by 25%. Nevertheless,
3.1.2.3. Revenue for operators. The majority of the revenue derives from there is no change in the top ranking, with S4-C remaining the best
the gate fees in all the scenarios, with the exception of S4-B and S4-C. As- option.
suming current incentives for the injection of biomethane to the grid and
the renewable heat incentive, S4-C provides £45/t FW of revenue, in addi- 3.1.2.6. Gate fees. As the gate fees are a cost to the local authority and rev-
tion to the £19/t FW from gate fees. The combined sale of electricity and enue for the waste facility operators, they cancel out in the overall LCC.
heat in S1-B from both incineration and AD provides the highest overall Hence, the sensitivity analysis focuses on the cost to the local authority
revenue of £21/t FW, alongside £79/t FW from gate fees. by varying the gate fees for each treatment route by ±25%. The only
12 P.C. Slorach et al. / Science of the Total Environment 710 (2020) 135580

case where the rankings of the scenarios change is for a reduction in the impacts (after S3). The difference in the scores between the current sit-
IVC gate fee by 25%, which results in S3 becoming preferable to S2-A/B. uation and BAU is small (16.5 vs 16.8).
In all other cases the results in Fig. S3 show that the gate fees are suffi-
ciently differentiated between the treatment options to cause a change 3.2. Environmental and economic sustainability at the UK level
in scenario rankings. The highest local-authority costs (£140/t FW)
occur for S1-A/B and BAU when the incineration gate fee is 25% higher. The results for the functional unit of 1 t of FW treated have been
Overall, the significantly lower gate fee for AD means that S4-A/B/C re- scaled up to determine the annual sustainability implications of manag-
main the lowest-cost options for local authorities. ing FW at the UK level. To enable comparison of the future scenarios
with the current situation, the analysis is based on the amount of
3.1.2.7. Electricity prices. Increasing the purchase price of electricity for waste generated at present (4.9 Mt/yr). It is acknowledged that the
industrial users only impacts IVC as all other treatment options are amount of FW will change in the future but it is not possible to speculate
self-sufficient for electricity. S3 has the biggest share of IVC and a ± by how much. This section also considers the sustainability implications
25% change in the electricity price results in a 1% change in total LCC of the prevention of FW compared to its treatment.
(see Table S14). The effect on the other scenarios is negligible.
The price at which electricity can be sold by the operator to the grid 3.2.1. Environmental sustainability
has a greater effect on the LCC (Table S15), which vary from ±0.6% (S4- To streamline the discussion, the focus here is on the GWP and PED;
C) to ±11.2% (S4-B). However, the overall ranking of the scenarios does for the other impacts, see Table S16 in the SI.
not change. The current treatment of household FW in the UK generates an esti-
mated 0.34 Mt CO2 eq./yr. If the same treatment mix was used in 2030
3.1.3. Overall environmental and economic sustainability (BAU), the GWP would increase by 6% to 0.36 Mt CO2 eq./yr due to
The environmental impacts and economic costs of the current situa- lower credits for displacing the future low-carbon electricity mix. S4-C
tion and future scenarios are integrated in Fig. 6 to rank them according achieves the highest annual reduction of −0.14 Mt CO2 eq.; this repre-
to their overall sustainability. A simple ranking method has been used to sents a saving of 0.5 Mt CO2 eq./yr relative to the worst option (BAU).
obtain a total score for each option. The score of an option has been es- Against the current situation, S4-C saves 0.48 Mt CO2 eq./yr, equivalent
timated by summing up its rankings in each environmental impact and to 342,000 cars taken off the road each year (Department for Transport,
the life cycle costs (LCC). To avoid bias, the total score for the environ- 2015; Department for Transport, 2019).
mental impacts has been divided by the number of impacts before The current situation is a net generator of energy with a net-negative
adding it to the LCC score. A lower score represents a better option. PED (−1.9 PJ/yr). S3 (−0.6 PJ/yr) is the only scenario to perform worse
As can be seen in Fig. 6, S4-B is the most sustainable option overall as than the current situation due to the increase in IVC, which is a net con-
it has the lowest average environmental score (2.5) and is ranked sec- sumer of electricity. The recovery of energy is maximised in S4-B
ond best for the LCC. The second best option is S4-C, with the lowest (−10.7 PJ/yr), which utilises the most electricity and heat from its
LCC and the third best environmental score (3.9). Even though S1-A large AD share. This results in a saving of 8.8 PJ/yr compared to BAU.
and S3 have worse environmental rankings compared to the current sit-
uation, they are still better options when taking into account the LCC. 3.2.2. Economic sustainability
BAU is the least sustainable overall as it is the second most costly option As shown in Fig. S4 in the SI, the total LCC for the current situation is
(after the current situation) and second worst for the environmental estimated at £452 million per year (m/yr). The slight improvement in

Current
Impacts and costs situation
BAU S1-A S1-B S2-A S2-B S3 S4-A S4-B S4-C

Primary energy demand 8 9 7 5 6 2 10 3 1 4

Global warming potential 9 10 7 4 6 3 8 5 2 1
Fossil depletion 6 9 8 4 7 3 10 5 2 1
Metal depletion 10 7 8 6 5 4 9 2 1 3
Freshwater ecotoxicity 2 9 8 6 7 5 10 4 3 1
Marine ecotoxicity 1 9 8 7 6 5 10 4 3 2
Terrestrial ecotoxicity 10 9 7 5 6 4 8 3 1 2
Human toxicity 9 10 8 6 5 4 7 3 2 1
Freshwater eutrophication 7 9 10 8 5 4 6 3 1 2
Marine eutrophication 9 10 8 7 6 5 1 4 2 2
Terrestrial acidification 2 3 4 1 6 5 7 10 9 8
Particulate matter formation 1 3 4 2 6 5 7 10 9 8
Photochemical oxidants formation 6 7 10 9 5 4 8 3 2 1
Ozone depletion 8 9 7 5 6 3 10 4 2 1
Agricultural land occupation 10 7 6 5 4 3 8 2 1 9
Urban land occupation 10 9 7 6 4 3 5 2 1 8
Natural land occupation 6 5 9 3 8 4 10 7 2 1
Ionising radiation 1 8 7 6 5 4 9 3 2 10
Water depletion 9 7 6 5 4 3 8 2 1 10
Avg. environmental score 6.5 7.8 7.3 5.3 5.6 3.8 7.9 4.2 2.5 3.9
Environmental ranking 7 9 8 5 6 2 10 4 1 3
Economic ranking 10 9 7 4 8 5 6 3 2 1
Total score 16.5 16.8 14.3 9.3 13.6 8.8 13.9 7.2 4.5 4.9
Overall ranking 9 10 8 5 6 4 7 3 1 2

Legend 1 2 3 4 5 6 7 8 9 10
Best Worst

Fig. 6. Heat map comparing the overall sustainability of different scenarios with the current situation.
 P.C. Slorach et al. / Science of the Total Environment 710 (2020) 135580 13

process efficiencies leads to a 0.7% decrease in total costs for BAU. S4-C and energetic dependency on incineration has led to a stagnation in
has the lowest LCC with the incentives (£114 m/yr), saving £338 m/yr recycling rates in Sweden (Zero Waste Europe, 2017). Therefore, as
compared to the current situation. However, the LCC of S4-C more the incineration capacity in the UK continues to expand, it is important
than doubles (£298 m/yr) if the renewable heat incentive is that long-term contracts for waste treatment facilities do not detract
discontinued. S4-B has the lowest LCC without incentives (£197 m/yr). from prevention and recycling of FW.
S4-A/B/C also offers the greatest annual savings to local authorities
of £272 m/yr compared to the current situation: while collection of sep- 4. Conclusions
arate FW is higher, the lower gate fee for AD more than balances this
out. BAU and S1-A/B have higher local-authority costs than at present This study has assessed the environmental and economic sustain-
due to increased incineration and landfill gate fees. Conversely, waste ability of five 2030 scenarios for the management of household FW in
operators gain the greatest profits in S1-B (£270 m/yr), which is the UK in comparison with the current situation. Across 16 of the 19 im-
£143 m/yr higher than at present; this is due to the high incineration pacts, at least one scenario outperforms the current situation and in
gate fee and revenue from electricity and heat. The lowest profits are seven categories either the current situation or BAU has the highest im-
found in S4-A (£73 m/yr), £54 m/yr less than today; this is due to the pacts. With the exception of BAU, the reduction in landfilling sees all
low AD gate fee and exporting only electricity. scenarios outperform the current situation for GWP. However, only
when heat can be utilised (S1-B, S2-B and S4-B) or when biomethane
3.2.3. Waste prevention can displace natural gas (S4-C) is a net-negative (saving) GWP
The GWP of different levels of FW prevention is compared in Fig. 7 to achieved. The 2030 electricity mix is sufficiently decarbonised to have
the current situation and two best 2030 scenarios, i.e. S4-A, which has a lower GWP than electricity generation from any of the waste treat-
the lowest GWP and S4-C, which has the lowest overall environmental ment options. However, the higher share of renewables results in
impact (see Fig. 6). For waste prevention, it is assumed that the GWP of greater metal depletion, human toxicity and terrestrial ecotoxicity, for
treating the FW generated is equal to the mean GWP of treating the FW which all future scenarios, except BAU, achieve significant savings com-
in all the 2030 scenarios. Both the prevention and treatment scenarios pared to the current situation.
include the GWP of food production given in Table 7, scaled up to the S4-B has the lowest environmental impacts for seven of the 19 cate-
annual impact. gories and the second lowest in a further eight. S4-C also has the lowest
As can be seen in Fig. 7, when the GWP of food production is in- impacts in seven categories, including GWP, but is ranked third overall
cluded, S4-C has only a 4% lower impact than the current situation, for environmental sustainability. However, both scenarios have the sec-
equal to an annual saving of 0.6 Mt CO2 eq. If the Courtauld Commit- ond highest terrestrial acidification and particulate matter formation
ment is achieved and 20% of avoidable and possibly avoidable FW is due to the high share of AD. When the heat can be utilised from both in-
prevented, 3.2 Mt CO2 eq./yr would be saved compared to the current cineration and AD, S2-B has the second lowest environmental impacts
situation. Reaching the UN target of halving the FW by 2030 would re- and requires only 50% of FW to be collected separately. S3 shows that
duce GWP by 51%, saving 7.42 Mt CO2 eq./yr. Pushing the limits much targeted FW collections do not necessarily lead to lower impacts as it
further and preventing all avoidable and potentially avoidable FW has the highest impacts for six categories and second worst for a further
would decrease the impact by 99.9%, avoiding 14.57 Mt CO2 eq./yr. two, largely due to the IVC.
However, this is an extreme case as it is probably unrealistic to expect S4-C has the lowest LCC of £23/t FW, which increases to £60/t FW
that FW could be reduced to just the unavoidable fraction. Nevertheless, without the incentives for biomethane. S4-B is the second best option
reducing the amount of FW is the best route to decreasing GWP and at £37/t FW, increasing to £40/t FW without incentives. BAU has the
other environmental impacts as well as costs. Even very small reduc- highest LCC due to the projected increase in landfill tax. Increased effi-
tions in FW would still be better than the best treatment options. As in- ciencies and reduction in landfilling result in all future scenarios, includ-
dicated in Fig. 7, prevention of just 2% of FW would save 0.6 Mt CO2 eq. ing BAU, having lower costs than the current situation. In S4-A/B/C, the
per year, equivalent to the savings in GWP in the best treatment sce- increased costs of separate FW collections are more than outweighed by
nario (S4-C). the lower gate fee for AD.
These results provide quantitative evidence in support of a circular The results of this study also demonstrate clearly that FW prevention
economy, demonstrating clearly the value of waste as a resource, partic- provides far greater benefits than its treatment. Taking into account the
ularly before food becomes waste. They also show that the upstream whole life cycle of food wasted, avoiding 20% of FW in the UK could save
impacts of FW must be considered to avoid prioritising treatment over 3.2 Mt CO2 eq. per year compared to the current situation. This is equiv-
prevention and a technological lock-in. The latter is evidenced in alent to 2.3 million cars taken off the road each year. It is also more than
Sweden which has excess incineration capacity and other countries six times the savings achieved in the best FW treatment scenario (S4-C).
now pay to export their waste there to be incinerated. This economic Meeting the UN Sustainable Development Goal for 50% reduction in FW

14.6 14.3 14.1 14.1

15
GWP (Mt CO₂ eq./yr)

11.5
10
7.2

0.019
0
Current S4-A S4-C 2% waste 20% waste 50% waste 100%
situation prevented prevented prevented avoidable and
potentially
avoidable
waste
prevented

Fig. 7. Global warming potential (GWP) of treating food waste compared to waste prevention. [For the GWP of food waste, see Table 7. For waste prevention, it is assumed that the GWP of
treating the waste generated is equal to the mean GWP of treating the waste in all the 2030 scenarios].
14 P.C. Slorach et al. / Science of the Total Environment 710 (2020) 135580

would save 7.5 Mt CO2 eq./yr. A much smaller FW reduction – only 2% – system/uploads/attachment_data/file/683051/UK_Statisticson_Waste_statisti-
cal_notice_Feb_2018_FINAL.pdf, Accessed date: August 2018.
is needed to match the best treatment scenario (S4-C) for GWP. Department for Transport, 2015. New car carbon dioxide emissions. [Online]. Department
Therefore, although there are significant benefits associated with for Transport Available:. www.gov.uk/government/publications/new-car-carbon-di-
treating FW compared to landfilling, prevention is still the best route oxide-emissions [Accessed September 2019].
Department for Transport, 2018. The Road to Zero. https://assets.publishing.service.gov.
to achieving significant environmental and economic savings. The sepa- uk/government/uploads/system/uploads/attachment_data/file/739460/road-to-zero.
rate collection of FW for treatment via AD should be encouraged, espe- pdf: UK Government.
cially if the use of heat or upgrading to biomethane can be widely Department for Transport, 2019. Anonymised MOT tests and results. [Online]. Depart-
ment for Transport Available:. data.gov.uk/dataset/e3939ef8-30c7-4ca8-9c7c-
implemented, but the treatment strategy and waste contracts must be ad9475cc9b2f/anonymised-mot-tests-and-results, Accessed date: 1 September 2019.
compatible with waste prevention targets. EC, 2008. Directive 2008/98/EC of the European Parliament and of the Council of 19 No-
vember 2008 on waste. Off. J. Eur. Union 312 (13), 22.11.
EC, 2016. Optimal use of biogas from waste streams - An assessment of the potential of
Declaration of competing interest biogas from digestion in the EU beyond 2020. https://ec.europa.eu/energy/sites/
ener/files/documents/ce_delft_3g84_biogas_beyond_2020_final_report.pdf
European Commission.
The authors declare no conflict of interest.
EC, 2018a. Directive (EU) 2018/850 of the European Parliament and of the Council of 30
May 2018 amending Directive 1999/31/EC on the landfill of waste. Off. J. Eur. Union
Acknowledgements 61, 100–108 (PE/10/2018/REV/2).
EC, 2018b. Directive (EU) 2018/851 of the European Parliament and of the Council of 30
May 2018 amending Directive 2008/98/EC on waste. Off. J. Eur. Union 61, 109–140
This work was carried out as part of the UK Centre for Sustainable (PE/11/2018/REV/2).
Energy Use in Food Chains (CSEF), funded by the UK Research Councils Ecoinvent, 2016. Ecoinvent Database v3.3 (Zurich and Lausanne, Switzerland).
(EP/K011820/1). The authors acknowledge gratefully this funding. Edelmann, W., Baier, U., Engeli, H., 2005. Environmental aspects of the anaerobic diges-
tion of the organic fraction of municipal solid wastes and of solid agricultural wastes.
Water Sci. Technol. 52 (1–2), 203.
Appendix A. Supplementary data Edwards, J., Othman, M., Crossin, E., Burn, S., 2018. Life cycle assessment to compare the
environmental impact of seven contemporary food waste management systems.
Bioresour. Technol. 248 (Pt A), 156–173.
Supplementary data to this article can be found online at https://doi. EEA, 2013a. Municipal Waste Management in Germany [Online]. European Environment
org/10.1016/j.scitotenv.2019.135580. Agency Available. www.eea.europa.eu/publications/managing-municipal-solid-
waste/germany-municipal-waste-management/view, Accessed date: February 2018.
EEA, 2013b. Municipal waste management in Sweden. [Online]. European Environment
References Agency Available:. www.eea.europa.eu/publications/managing-municipal-solid-
waste/sweden-municipal-waste-management/view, Accessed date: 1 February 2018.
Banks, C.J., Chesshire, M., Heaven, S., Arnold, R., 2011. Anaerobic digestion of source- Environment Agency, 2009. How to comply with your environmental permit Additional
segregated domestic food waste: performance assessment by mass and energy bal- guidance for: Landfill (EPR 5.02). [Online]. Environment Agency Available:. www.
ance. Bioresour. Technol. 102 (2), 612–620. gov.uk/government/uploads/system/uploads/attachment_data/file/296861/
BEIS, 2018a. Energy Trends: Electricity - Annex J: Total Electricity Generation by Source. geho0409bput-e-e.pdf, Accessed date: 1 August 2018.
Department for Business, Energy & Industrial Strategy (BEIS), UK Government. Environment Agency, 2016. Municipal Waste Incinerator Annual Reports for 2016 [On-
BEIS, 2018b. Updated energy and emissions projections: 2017. [Online]. Department for line]. Environment Agency Available. ukwin.org.uk/resources/incinerator-reports/
Business, Energy & Industrial Strategy (BEIS) Available:. http://www.gov.uk/govern- (Accessed July 2018).
ment/publications/updated-energy-and-emissions-projections-2017, Accessed date: Eriksson, M., Spangberg, J., 2017. Carbon footprint and energy use of food waste manage-
August 2018. ment options for fresh fruit and vegetables from supermarkets. Waste Manag. 60,
BEIS, 2019. Updated energy and emissions projections: 2018 - Annex M: Growth assump- 786–799.
tions and prices. [Online]. Department for Business, Energy & Industrial Strategy Eriksson, M., Strid, I., Hansson, P.-A., 2015. Carbon footprint of food waste management
(BEIS), UK Goverment Available:. www.gov.uk/government/publications/updated- options in the waste hierarchy – a Swedish case study. J. Clean. Prod. 93, 115–125.
energy-and-emissions-projections-2018, Accessed date: 1 August 2019. Eunomia, 2014. Anaerobic Digestion Market Update. Eunomia Research & Consulting.
Bernstad, A., la Cour Jansen, J., 2011. A life cycle approach to the management of house- Eunomia, 2017. Residual Waste Infrastructure Review.
hold food waste – a Swedish full-scale case study. Waste Manag. 31 (8), 1879–1896. Evangelisti, S., Lettieri, P., Borello, D., Clift, R., 2014. Life cycle assessment of energy from
Bernstad, A., la Cour Jansen, J., 2012. Review of comparative LCAs of food waste manage- waste via anaerobic digestion: a UK case study. Waste Manag. 34 (1), 226–237.
ment systems – current status and potential improvements. Waste Manag. 32 (12), FAO, 2014. Food wastage footprint - Full-cost accounting - Final Report. http://www.fao.
2439–2455. org/3/a-i3991e.pdf Food and Agriculture Organisation of the United Nations - FAO.
Budzianowski, W.M., Wylock, C.E., Marciniak, P.A., 2017. Power requirements of biogas FAO, 2018. The State of Food Security and Nutrition in the World 2018: Building Resilience
upgrading by water scrubbing and biomethane compression: comparative analysis for Peace and Food Security: The State of Food Security and Nutrition in the World. FAO.
of various plant configurations. Energy Convers. Manag. 141, 2–19. Feeding the 5000, 2016. The Food Waste Pyramid. [Online]. Feedback. Available:. http://
Burns, C., Hopwood, L., 2017. Organics recycling industry status report 2015. http://www. www.feeding5k.org/businesses+casestudies.php, Accessed date: 1 November 2018.
wrap.org.uk/sites/files/wrap/asori%202015.pdf WRAP. Goedkoop, M., Heijungs, R., Huijbregts, M., De Schryver, A., Struijs, J., van Zelm, R., 2013.
Burns, C., Hopwood, L., More, O., Hindle, M., 2017. Organics recycling industry status re- ReCiPe 2008: A Life Cycle Impact Assessment Method Which Comprises Harmonised
port 2015. [Online]. WRAP Available:. www.wrap.org.uk/sites/files/wrap/asori% Category Indicators at the Midpoint and the Endpoint Level. First edition (revised).
202015.pdf, Accessed date: 1 August 2018. PRé Consultants CML, University of Leiden RUN, Radboud University Nijmegen, RIVM.
Carlsson Reich, M., 2005. Economic assessment of municipal waste management systems Gustavsson, J., Cederberg, C., Sonesson, U., Van Otterdijk, R., Meybeck, A., 2011. Global
—case studies using a combination of life cycle assessment (LCA) and life cycle cost- food losses and food waste. www.fao.org/docrep/014/mb060e/mb060e00.pdf The
ing (LCC). J. Clean. Prod. 13 (3), 253–263. Food and Agriculture Organization of the United Nations.
CMA, 2018. Heat Networks market study: final report Appendices. https://assets.publish- HMRC, 2018. Landfill Tax: increase in rates. [Online]. HM Revenue and Customs, UK Gov-
ing.service.gov.uk/media/5b5596e6e5274a3fd4524406/heat_network_annexes.pdf ernment Available:. www.gov.uk/government/publications/increase-of-the-landfill-
Competition & Markets Authority. tax-rates/landfill-tax-increase-in-rates, Accessed date: 1 August 2019.
DAERA-NI, 2017. Annual NILAS Reports [Online]. Department of Agriculture, Environment Hunkeler, D., Lichtenvort, K., Rebitzer, G., 2008. Environmental Life Cycle Costing. Crc press.
and Rural Affairs Available. www.daera-ni.gov.uk/articles/published-waste-data, ISO, 2006a. ISO 14040-Environmental Management—Life Cycle Assessment—Principles
Accessed date: 1 November 2018. and Framework (Geneva).
DEFRA, 2004. Defra Guidance for Successful Reclamation of Mineral and Waste Sites. ISO, 2006b. ISO 14044-Environmental Management—Life Cycle Assessment—Require-
[Online]. Department for Environment, Food & Rural Affairs Available:. ments and Guidelines (Geneva).
webarchive.nationalarchives.gov.uk/20090318074725/http://www.defra.gov. Kharas, H., 2017. The unprecedented expansion of the global middle class: an update.
uk/farm/environment/land-use/reclamation/guidance-full.pdf, Accessed date: Global Economy & Development Working Paper 100. The Brookings Institution,
1 August 2018. Washington, DC.
DEFRA, 2017. ENV18 - Local authority collected waste: annual results tables. [Online]. De- Kim, M.-H., Kim, J.-W., 2010. Comparison through a LCA evaluation analysis of food waste
partment for Environment, Food and Rural Affairs Available:. www.gov.uk/govern- disposal options from the perspective of global warming and resource recovery. Sci.
ment/statistical-data-sets/env18-local-authority-collected-waste-annual-results-ta- Total Environ. 408 (19), 3998–4006.
bles, Accessed date: August 2018. Kim, M.-H., Song, Y.-E., Song, H.-B., Kim, J.-W., Hwang, S.-J., 2011. Evaluation of food waste
DEFRA, 2018a. Our waste, our resources: a strategy for England. [Online]. Department for disposal options by LCC analysis from the perspective of global warming: Jungnang
Environment, Agriculture & Rural Affairs Available:. https://assets.publishing.service.- case, South Korea. Waste Manag. 31 (9–10), 2112–2120.
gov.uk/government/uploads/system/uploads/attachment_data/file/765914/re- Letsrecycle, 2018. Barnet to terminate separate food waste collections [online].
sources-waste-strategy-dec-2018.pdf, Accessed date: 1 December 2018. letsrecycle.com (Accessed March 2019).
DEFRA, 2018b. UK Statistics on Waste [Online]. Department for Environment Food & Levis, J.W., Barlaz, M.A., 2011. What is the most environmentally beneficial way to treat
Rural Affairs Available. assets.publishing.service.gov.uk/government/uploads/ commercial food waste? Environ. Sci. Technol. 45 (17), 7438–7444.
 P.C. Slorach et al. / Science of the Total Environment 710 (2020) 135580 15

Martinez-Sanchez, V., Tonini, D., Møller, F., Astrup, T.F., 2016. Life-cycle costing of food Countryside/Waste-Management/Local-Authority-Municipal-Waste/wastemanaged-
waste management in Denmark: importance of indirect effects. Environ. Sci. Technol. by-management-year, Accessed date: 1 August 2018.
50 (8), 4513–4523. Stenmarck, A., Jensen, C., Quested, T., Moates, G., Buksti, M., Cseh, B., Juul, S., Parry, A.,
Monson, K., Esteves, S., Guwy, A., Dinsdale, R., 2007. Anaerobic Digestion of Biodegradable Politano, A., Redlingshofer, B., 2016. Estimates of European Food Waste Levels. FU-
Municipal Wastes: A Review. University of Glamorgan (ISBN, 978-1). SIONS - European Commission.
Naroznova, I., Møller, J., Scheutz, C., 2016. Global warming potential of material fractions Swarr, T.E., Hunkeler, D., Klöpffer, W., Pesonen, H.-L., Ciroth, A., Brent, A.C., Pagan, R., 2011.
occurring in source-separated organic household waste treated by anaerobic diges- Environmental life-cycle costing: a code of practice. Int. J. Life Cycle Assess. 16 (5),
tion or incineration under different framework conditions. Waste Manag. 58, 389–391.
397–407. Thinkstep, 2019. GaBi Software-System and Database for Life Cycle Engineering
National Grid, 2018. Future Energy Scenarios. [Online]. National Grid Available:. http://fes. (Leinfelden-Echterdingen).
nationalgrid.com/fes-document/, Accessed date: 1 August 2018. Tong, H., Shen, Y., Zhang, J., Wang, C.-H., Ge, T.S., Tong, Y.W., 2018. A comparative life cycle
Net Zero Waste, 2018. Gore Cover System. [Online]. Available:. http://www.netzerowaste. assessment on four waste-to-energy scenarios for food waste generated in eateries.
com/gore-cover/, Accessed date: 1 September 2018. Appl. Energy 225, 1143–1157.
NNFCC, 2019a. Biogas Map. [Online]. Available:. www.biogas-info.co.uk/resources/biogas- Towler, G., Sinnott, R., 2013. Chapter 14 - design of pressure vessels. In: Towler, G.,
map/, Accessed date: 1 June 2019. Sinnott, R. (Eds.), Chemical Engineering Design, Second edition Butterworth-
NNFCC, 2019b. Press Release - NNFCC Publishes 2019 Anaerobic Digestion Deployment in Heinemann, Boston.
the UK Report [Online]. NNFCC Available. https://www.nnfcc.co.uk/press-release-ad- UK Government, 2014. Using animal by-products at compost and biogas sites. [Online].
report-published-2019 (Accessed July 2019). Available:. www.gov.uk/guidance/using-animal-by-products-at-compost-and-bio-
Office for National Statistics, 2017. United Kingdom population mid-year estimate. [On- gas-sites, Accessed date: 1 July 2018.
line]. Available:. www.ons.gov.uk/peoplepopulationandcommunity/ UK Parliament, 2017. In: Affairs, D.F. (Ed.), Commons Select Committee - Food Waste in
populationandmigration/populationestimates/timeseries/ukpop/pop, Accessed date: England https://publications.parliament.uk/pa/cm201617/cmselect/cmenvfru/429/
1 June 2018. 42902.htm.
Office for National Statistics, 2019. Average Sterling exchange rate: Euro. [Online]. Avail- UN, 2015. Resolution adopted by the General Assembly on 25 September 2015.
able:. www.ons.gov.uk/economy/nationalaccounts/balanceofpayments/timeseries/ Transforming our world: the 2030 Agenda for Sustainable Development [online].
thap/mret, Accessed date: 1 November 2019. Available. sustainabledevelopment.un.org/post2015/transformingourworld,
Ofgem, 2018a. About the FIT scheme. [Online]. Available:. www.ofgem.gov.uk/environ- Accessed date: 1 August 2018.
mental-programmes/fit/about-fit-scheme, Accessed date: 1 September 2018. Vandermeersch, T., Alvarenga, R.A.F., Ragaert, P., Dewulf, J., 2014. Environmental sustain-
Ofgem, 2018b. Feed-In Tariff (FIT) rates. [Online]. Office of Gas and Electricity Markets, UK ability assessment of food waste valorization options. Resour. Conserv. Recycl. 87,
Government Available:. www.ofgem.gov.uk/environmental-programmes/fit/fit-tar- 57–64.
iff-rates, Accessed date: 1 August 2018. Vermeulen, S.J., Campbell, B.M., Ingram, J.S.I., 2012. Climate change and food systems.
Ofgem, 2018c. Non-Domestic Renewable Heat Incentive (RHI). [Online]. Ofgem. Avail- Annu. Rev. Environ. Resour. 37 (1), 195–222.
able:. www.ofgem.gov.uk/environmental-programmes/non-domestic-rhi, Accessed WasteDataFlow, 2018. . Available. wastedataflow.org/.
date: 1 September 2018. Welsh Government, 2011. Municipal Sector Plan Part 1. [Online]. Welsh Assembly Gov-
Ofgem, 2018d. Tariffs and payments: Non-Domestic RHI. [Online]. Ofgem. Available:. ernment Available:. https://gov.wales/docs/desh/publications/
www.ofgem.gov.uk/environmental-programmes/non-domestic-rhi/contacts-guid- 110310municipalwasteplan1en.pdf, Accessed date: 1 June 2018.
ance-and-resources/tariffs-and-payments-non-domestic-rhi, Accessed date: 1 June WRAP, 2008. The Food We Waste [Online]. WRAP Available. http://wrap.s3.amazonaws.
2018. com/the-food-we-waste.pdf (Accessed May 2019).
Ofgem, 2018e. Tariffs and payments: Non-Domestic RHI. [Online]. www.ofgem.gov.uk/ WRAP, 2011. Gate fees report 2011. [Online]. Waste & Resources Action Programme
environmental-programmes/non-domestic-rhi/contacts-guidance-and-resources/ (WRAP) Available:. www.wrap.org.uk/sites/files/wrap/Gate%20Fees%20Report%
tariffs-and-payments-non-domestic-rhi Office of Gas and Electricity Markets, UK 202011.pdf, Accessed date: 1 August 2018.
Government [Accessed 14/03/18]. WRAP, 2016a. Gate fees report 2015/16: Comparing the costs of waste treatment options.
Oldfield, T.L., White, E., Holden, N.M., 2016. An environmental analysis of options for http://www.wrap.org.uk/system/files/private/Gate%20Fees%20technical%20report%
utilising wasted food and food residue. J. Environ. Manag. 183 (Part 3), 826–835. 202016.PDF.
Papargyropoulou, E., Lozano, R., Steinberger, J.K., Wright, N., Ujang, Z., 2014. The food WRAP, 2016b. Household food waste collections guide. [Online]. Waste & Resources Ac-
waste hierarchy as a framework for the management of food surplus and food tion Programme (WRAP) Available:. www.wrap.org.uk/sites/files/wrap/HH_food_
waste. J. Clean. Prod. 76, 106–115. waste_collections_guide_section_1_context_%20and_background.pdf, Accessed
Quested, T., Parry, A., 2017. Household Food Waste in the UK, 2015 [Online]. WRAP Avail- date: 1 August 2018.
able. www.wrap.org.uk/sites/files/wrap/Household_food_waste_in_the_UK_2015_ WRAP, 2017a. Gate fees report 2017. [Online]. Waste & Resources Action Programme
Report.pdf (Accessed August 2017). (WRAP) Available:. www.wrap.org.uk/sites/files/wrap/Gate%20Fees%20report%
Quested, T., Easteal, S., Ingle, R., 2013. Methods used for Household Food and Drink Waste 202017_FINAL_clean.pdf, Accessed date: 1 August 2018.
in the UK 2012. [Online]. WRAP. Available:. www.wrap.org.uk/sites/files/wrap/ WRAP, 2017b. What is Courtauld 2025?. [Online]. Waste & Resources Action Programme
Methods%20Annex%20Report%20v2.pdf, Accessed date: 1 April 2019. (WRAP) Available:. www.wrap.org.uk/content/what-courtauld-2025, Accessed date:
Scholes, P., Areikin, E., 2014. A survey of the UK Anaerobic Digestion industry in 2013. 1 August 2018
[Online]. WRAP. Available:. www.wrap.org.uk/content/survey-uk-anaerobic-diges- WRAP, 2018a. Gate Fees 2017/18 Final Report. [Online]. Waste & Resources Action Pro-
tion-industry-2013, Accessed date: July 2018. gramme (WRAP) Available:. www.wrap.org.uk/sites/files/wrap/WRAP%20Gate%
SEPA, 2017. Household Waste – Summary Data 2016 [Online]. Scottish Environment Pro- 20Fees%202018_exec+extended%20summary%20report_FINAL.pdf, Accessed date:
tection Agency Available. www.sepa.org.uk/media/320744/household-waste-sum- 1 August 2019.
mary-data-and-commentary-2016.pdf (Accessed July 2018). WRAP, 2018b. Kerbside Costing Tool. [Online]. Waste & Resources Action Programme
Slorach, P.C., Jeswani, H.K., Cuéllar-Franca, R., Azapagic, A., 2019a. Environmental and eco- (WRAP) Available:. laportal.wrap.org.uk/ICPToolHome.aspx, Accessed date: 1 August
nomic implications of recovering resources from food waste in a circular economy. 2018.
Sci. Total Environ. 693, 133516. Zero Waste Europe, 2017. Sweden's recycling (d)evolution. [Online]. Zero Waste Europe
Slorach, P.C., Jeswani, H.K., Cuéllar-Franca, R., Azapagic, A., 2019b. Environmental sustain- Available:. https://zerowasteeurope.eu/2017/06/swedens-recycling-devolution/,
ability of anaerobic digestion of household food waste. J. Environ. Manag. 236, Accessed date: 1 October 2018.
798–814.
StatsWales, 2017. Waste managed (tonnes) by management method and year. [Online].
Welsh Goverment. Available:. statswales.gov.wales/Catalogue/Environment-and-