The Swedish fika down the drain

Abstract

Considering coffee's significant social role, reducing coffee waste is pivotal. We quantified liquid coffee waste generated in the Swedish food service sector and explored causes and potential mitigation measures. We combined quantitative data from 76 days across six restaurants with qualitative insights. The results showed that 10% of brewed coffee is wasted daily, corresponding to 3.3 kg restaurant/day, 13 g customer/day and 739 g employee/day. When extrapolated to national scale, these findings suggest that Swedish restaurants generate approximately 17,800 tonnes of coffee waste annually. Thus liquid coffee waste represents a previously unaccounted for 21% increase in food waste, not including waste left in consumers' cups. We identified cost savings as a motivator for waste reduction and time constraints as a significant barrier. These findings provide a more comprehensive understanding of food service waste, while also highlighting the need for inclusion of liquid waste in national statistics and for policy intervention.

Full text

Resources, Conservation & Recycling 211 (2024) 107853
Available online 10 August 2024
0921-3449/© 2024 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Full length article
The Swedish ka down the drain
Christopher Malefors
a
,
*
, Rosa Hellman
b
, Amanda Sj¨
olund
a
, Mattias Eriksson
a
a
Department of Energy and Technology, Swedish University of Agricultural Sciences, Box 7032, Uppsala SE-75007 Uppsala, Sweden
b
Department of People and Society, Swedish University of Agricultural Sciences, PO Box 190, Lomma SE-23422 Alnarp, Sweden
ARTICLE INFO
Keywords:
Food waste
Coffee waste
Quantication
Municipal waste
Beverages
ABSTRACT
Considering coffee’s signicant social role, reducing coffee waste is pivotal. We quantied liquid coffee waste
generated in the Swedish food service sector and explored causes and potential mitigation measures. We com-
bined quantitative data from 76 days across six restaurants with qualitative insights. The results showed that
10% of brewed coffee is wasted daily, corresponding to 3.3 kg restaurant/day, 13 g customer/day and 739 g
employee/day. When extrapolated to national scale, these ndings suggest that Swedish restaurants generate
approximately 17,800 tonnes of coffee waste annually. Thus liquid coffee waste represents a previously unac-
counted for 21% increase in food waste, not including waste left in consumers’cups. We identied cost savings as
a motivator for waste reduction and time constraints as a signicant barrier. These ndings provide a more
comprehensive understanding of food service waste, while also highlighting the need for inclusion of liquid
waste in national statistics and for policy intervention.
1. Introduction
With a growing global population, consumption of coffee is
increasing (Quadra et al., 2020). The estimated global trade value of the
coffee industry is USD 38 billion, or 2.5% of the total trade value in
agricultural commodities (International Trade Centre, 2021). In 2020,
worldwide production of green coffee beans was approximately 11
million tons (FAOSTAT, 2023), supplying a diverse range of coffee
beverages for consumers worldwide. However, coffee production has
environmental costs, e.g. 1 kg of ground coffee powder generates an
estimated 4.0 kg CO
2
e, contributing to climate change (Eneroth et al.,
2022), with the majority of this impact originating from the production
phase. The industry also has signicant effects on biodiversity, primarily
due to deforestation to expand cultivation areas (Ahlgren et al., 2022).
Moreover, fertiliser and pesticide use in coffee cultivation is compara-
tively high, leading to greenhouse gas emissions, climate change and
eutrophication as associated environmental concerns (Ahlgren et al.,
2022;Cederberg et al., 2019;Ho et al., 2022;Moberg et al., 2020).
Coffee consumption plays an important role in social life worldwide,
and especially in Scandinavia (Kjeldgaard and Ostberg, 2007). From a
global perspective, Sweden has the third highest per capita coffee con-
sumption (International Trade Centre, 2021). Between 2010 and 2020,
yearly per capita consumption of roasted coffee in Sweden increased
from 7.5 to 8.8 kg (Swedish Board of Agriculture, 2023). Translated into
amount of coffee consumed by Swedish adults, this corresponds to on
average 280 mL of brewed coffee per day (Lundberg-Hall´
en and ¨
Ohrvik,
2015).
Given the signicant environmental impact of high coffee con-
sumption, the World WildLife Fund (WWF) included hot beverages in its
2022 consumer guide, which encourages use of organically certied
coffee and stresses the importance of minimising waste (WWF, 2022).
Wasting food means that the resources required to produce and deliver
the nal product were used in vain (Johnson, 2020). This applies also for
liquids such as coffee, as such products are considered food (European
Commission, 2002). Preventing waste of coffee beverages therefore not
only avoids waste of resources, but also avoids the environmental
impact linked to utilisation of those resources (Büsser et al., 2008).
To minimise the negative effects caused by food waste generation,
several institutions have included this aspect in their goals. In 2015, the
United Nations (UN) set the target of halving per capita global consumer
food waste by the year 2030 (United Nations, 2015). The same target is
included in the Farm to Fork Strategy introduced by the European Union
(EU) (European Union, 2020). Additionally, one of the milestone targets
in the Swedish environmental objectives system is to reduce the total
mass of food waste by 20% between 2020 and 2025 (Swedish Envi-
ronmental Protection Agency, 2023).
An essential step in achieving the goal of reduced food waste is
quantication, in order to obtain clear data on food waste levels and
* Corresponding author.
E-mail address: christopher.malefors@slu.se (C. Malefors).
Contents lists available at ScienceDirect
Resources, Conservation &Recycling
journal homepage: www.sciencedirect.com/journal/resources-conservation-and-recycling
https://doi.org/10.1016/j.resconrec.2024.107853
Received 22 May 2024; Received in revised form 28 June 2024; Accepted 31 July 2024

Resources, Conservation & Recycling 211 (2024) 107853
2
compare these with set reduction targets (Geislar, 2020;Xue et al.,
2017). Several standards have been developed to improve methodo-
logical practices in food waste quantication, but the denitions and
scope of each standard vary and liquid waste is excluded frequently
(European Commission, 2022;International Food Waste Coalition,
2022;Swedish National Food Agency, 2023;Tostivint et al., 2016;
UNEP, 2021;World Resources Institute, 2016). Excluding specic food
categories or waste streams may overlook places where signicant food
waste occurs and where preventive measures are necessary to reduce
waste and meet reduction targets.
Food waste is generated at all stages of the supply chain to different
extents and by different causes (Partt et al., 2010). United Nations
Environment Programme (UNEP) estimates of food waste levels at the
consumption stages of the supply chain indicate that households are
responsible for the greatest share of global food waste (61%), while the
food service sector accounts for the second greatest share (26%) (UNEP,
2021). However, these estimates are based on solid food waste, while
information on liquid food waste generation is limited. Estimates gained
through surveys by the Swedish Environmental Protection Agency
(2021) showed that 190,000 tons of food and beverages were wasted via
Swedish household drains in 2020 and that almost 45% of this waste was
tea or coffee. These 190,000 tons represent 23% of total household food
waste in Sweden (Swedish Environmental Protection Agency, 2022).
However, estimates of liquid food waste in other parts of the Swedish
food supply chain are lacking.
Few previous studies on the food service sector have included liquid
food waste, and the methods used and the results obtained have varied.
A study by Ahmed et al. (2018) quantied food waste, including liquids,
in a university dining hall in the United States and found that around
19% of total waste consisted of post-consumer liquid waste (beverages
and soups). A study by Sehnem et al. (2022) on the Brazilian food service
sector, using survey methodology, showed that 4% of total waste came
from drinks or desserts. Differences in scope in previous studies and
limited denitions of the food categories and waste streams included
make it difcult to compare results. Overall, however, the results indi-
cate that liquid food waste occurs and that more research is needed to
quantify this waste. As mentioned, the food service sector is a large
contributor to food waste. It also has many entities, visitors and high
levels of production, so a decrease in food waste generation could help
reduce the negative impacts of this sector (Heikkil¨
a et al., 2016). Coffee
is offered in most food service establishments, making it a relevant food
category to focus on when exploring levels of liquid food waste in the
sector.
Identifying sources of food waste is an essential initial step in pre-
vention, as it provides an indication of where efforts should be targeted.
Moreover, when developing interventions for effectively reducing food
waste, an understanding of the underlying causes of waste generation is
needed. The extent to which coffee beverages are wasted in the food
service sector is currently unknown, as are the types of waste prevention
measures needed to reduce coffee waste and minimise its negative ef-
fects. The research questions for this study therefore are:
1. How much coffee waste is generated in the Swedish food service
sector
2. What are the primary causes of coffee waste in the Swedish food
service sector
3. Why is coffee waste often neglected in waste quantication studies
2. Material and methods
This research took the form of a mixed methods case study, where
both quantitative and qualitative methods were applied. The quantita-
tive part of the study involved quantifying actual levels of coffee waste
in restaurants. The qualitative part comprised semi-structured in-
terviews with the restaurants taking part in quantications and an on-
line survey of non-participating food service establishments.
In addition to quantifying coffee waste generated in the Swedish food
service sector and identifying the underlying causes, perceived barriers
to quantication of coffee waste were examined. In Swedish restaurants,
coffee is commonly brewed using drip lter coffee machines and is sold
as individual cups or as part of a meal, e.g. lunch. In this study, coffee
waste was dened as the liquid part of lter coffee i.e. drinkable coffee.
This only refers to the volume prepared but not served and eventually
becoming waste from the coffe machines or containers. This does not
include the solid part, i.e. coffee grounds, that is normally not
consumed. This study also did not include post-consumer waste i.e.
consumer leftovers in cups that was eventually wasted. The food service
sector was dened as “establishments or actors providing complete meals or
drinks t for immediate consumption, whether in traditional restaurants, self-
service or take-away restaurants, whether as permanent or temporary stands
with or without seating. Decisive is the fact that meals t for immediate
consumption are offered, not the kind of facility providing them”(EURO-
STAT, 2008).
2.1. Quantication and analysis of coffee waste
Recruitment for the study targeted food service establishments of-
fering lter coffee to consumers, specically including caf´
es, restau-
rants, conference centres and hotels in Sweden. Selection was random,
but geographically convenient. Initial contact with potential partici-
pants was made through email, providing study details and inquiring
about their interest in participation. Non-respondents received a follow-
up reminder. Of 201 establishments contacted across ve Swedish cities,
56 declined and 139 did not reply. Ultimately, six restaurants engaged in
the waste quantication process and four of these participated in sub-
sequent interviews. Information about the restaurants participating in
the study are summarised in Table 1.
Coffee waste was quantied in the six participating food service
restaurants (A-F) for a total of 76 days between October and December
2023, with the duration of participation ranging from 5 to 22 days per
restaurant (Table 1). The variation in quantication days was due to
each restaurant’s capacity and willingness to engage in the study. Staff
members carried out daily quantications, using the instructions and
digital and printed sheets provided (Table S1 in Supplementary Infor-
mation (SI)), and a weighing scale was supplied if lacking at the estab-
lishment. Coffee waste was collected and weighed on digital scales at
restaurants A, C, D and E, while restaurant B measured it in litres. The
collected data included the volume of coffee produced daily and the
total guest count, recorded on the quantication sheets. At restaurant F,
quantication was performed by one of the authors, to serve as
validation.
Table 2.
Coffee waste was analysed using three different indicators (per guest,
per employee and in relation to the total mass of coffee brewed) and
presented as descriptive statistics. The total volume of coffee brewed
was estimated using coffee machine indicators, rather than by weight. It
was assumed that 1 kg of liquid coffee is equivalent to 1 L. For statistical
validity, only days with complete data entries, including the amount of
coffee produced, number of guests and quantity of waste, were consid-
ered in the analysis. This criterion led to the exclusion of data from one
day at restaurant C (day 7).
2.2. Understanding waste causes and quantication motivators
To complement the quantitative data, semi-structured interviews
were conducted with representatives from four of the six restaurants.
The purpose of these interviews was to identify the reasons behind
coffee waste and to explore the obstacles and incentives relating to its
quantication. The interview format included 10 open-ended questions
designed to elicit insights on the motivation of establishments for
participating in the study, perceptions on coffee waste issues and re-
actions to the levels of waste identied at their establishment (Table S2
C. Malefors et al.

Resources, Conservation & Recycling 211 (2024) 107853
3
in SI). These interviews were carried out following the quantication
phase.
In addition to the interviews, a condential online survey was
disseminated in October 2023 to the 195 restaurants that either did not
respond or declined to participate in the waste quantication step. Two
reminders were issued to increase response rates. The survey was
accessible for 50 days on Netigate, a survey platform, and comprised
both open-ended and multiple-choice questions (Table S3 in SI). These
sought to understand the reasons for non-participation, incentives that
might encourage waste quantication and perspectives on coffee waste
causes within their own operations and more broadly. Respondents had
the option to provide contact details for potential follow-up interviews.
The survey garnered 18 complete responses, with no respondents opting
to provide contact information. However, a single response could
represent multiple restaurants, especially those under the same corpo-
rate management, implying a higher effective response rate in terms of
the number of establishments represented.
Responses from the semi-structured interviews and the online survey
were systematically analysed. They were initially sorted into categories
related to the causes of liquid coffee waste, barriers to quantifying coffee
waste, and drivers encouraging quantication. These preliminary
groupings were further rened into themes (Braun and Clarke, 2006),
based on common elements. The frequency of each response type was
recorded and summarised. It is important to note that the data did not
distinguish between waste generated during preparation, during serving
or by consumers, but rather addressed coffee waste as a whole.
2.3. Extrapolating coffee waste to national level
To calculate and compare the amount of coffee waste (in tonnes)
generated in the Swedish food service sector, coffee waste per employee
(g) was multiplied by number of employees in the sector and days open per
year (Equation 1). This procedure scaled the coffee waste per employee
and yielded a value in tonnes per year, which is similar to how solid
waste in the food service sector is scaled up to national level (Swedish
Environmental Protection Agency, 2022).
Monte Carlo simulations were used to complement the calculated
value and to estimate the uncertainty range in the nal tonnes per year
factor. The simulations were performed 10,000 times, with variability
captured using different distributions. These distributions were assumed
based on data collected, literature and our own assessment.
The parameter waste per employee was modelled using various sta-
tistical distributions (Table S4 in SI), with the three distributions
demonstrating the lowest Akaike information criterion (AIC) values
((Gamma, Exponential, LogNormal)) selected for use in simulations
(Burnham and Anderson, 2004).
The simulations provided a comparative analysis against the
extrapolated coffee waste value, thereby testing the sensitivity of the
model to variations in the underlying waste per employee parameter.
Coffee waste per year (tonnes)= Number of employees in Swedish restaurants∗
Number of days open per year ∗waste (g)per employee∗10−6
(1)
Calculated deterministic coffee waste was also compared with the
solid food waste generated within the Swedish restaurant sector. Coffee
waste in tonnes was also expressed per capita, as an additional way to
compare the results for different parts of the food supply chain.
3. Results
The 76 days of waste quantication across six restaurants in various
locations in Sweden indicated that around 10% of all brewed coffee was
discarded by these restaurants, not accounting for waste from cus-
tomers. On average, coffee waste amounted to roughly 3.3 kg per
restaurant per day, 13 g per customer per day and 739 g per employee
per day. Table 3 shows descriptive statistics for these indicators.
There was great variation within the individual restaurants, where
daily coffee waste ranged from 0 to 40% (Fig. 1). Restaurant F had the
lowest median waste percentage (4%), but also recorded waste for the
smallest number of days (5). Restaurant A had the highest median waste
percentage (around 11%), recorded over 22 days. Most restaurants
exhibited occasional spikes in waste, resulting in their waste distribution
patterns having a skewed distribution.
Despite restaurant A recording the highest median percentage of
coffee waste, it produced the lowest volume of coffee (Fig. 2). Its daily
coffee production uctuated between 2 and 130 kg, while its coffee
waste varied from 0 to 13 kg. Restaurant D was the largest coffee pro-
ducer, with daily quantities ranging from 110 to 130 kg, but its waste
mass overlapped with that of restaurants E and C, which had brewing
volumes of 27 kg to 65 kg, respectively (Fig. 2). Columnar patterns were
observed for brewed coffee amount, indicated that coffee is produced in
batches. Therefore a day’s production is made up of a number of
batches, rather than the exact number of cups needed to full customer
demand.
A noteworthy nding was that throughout the quantication period
and across all restaurants, there was a complete absence of waste during
the coffee preparation process.
Table 1
Participating restaurants and their characteristics
Restaurant Participated in interviews Quantication days Business type Guests/day (average) Employees (n)
A Yes 22 Lunch restaurant and caf´
e in ofce building 63 2
B Yes 10 Lunch restaurant for seniors 144 4
C Yes 20 Campus lunch restaurant and caf´
e 192 5
D Yes 10 Staff canteen and caf´
e 600 6
E 10 Campus caf´
e 622 3
F 5 Lunch restaurant and caf´
e in ofce building 548 7
Table 2
Parameters and distributions used in calculations and Monte Carlo simulations of Swedish coffee waste (in tonnes) per year
Parameter Distribution Description
Employees in Swedish restaurants Triangular Based on statistics provided by Statistics Sweden and the SNI code 56100.
Mode: 92 695±2.5%
Number of days open per year Triangular Min: 156, Mode: 260, Max: 365
Estimated based on assumptions.
Waste per employee (g) Fitted Fitted from collected data, zeroes removed but introduced as a fraction of their occurrence in the Monte Carlo simulations.
C. Malefors et al.

Resources, Conservation & Recycling 211 (2024) 107853
4
3.1. Causes of coffee waste and drivers and barriers to quantication
There was a signicant issue of coffee overproduction, which was a
notable source of waste (Table 4). Analysis of the interview and survey
data indicated that this tendency to overproduce stemmed from con-
cerns about potential shortages of coffee before the end of the day,
coupled with an inability to forecast daily consumption. Factors
contributing to these problems included unpredictable number of guests
and the resulting coffee demand, lack of time to analyse customer ow,
and pressure that coffee must always be freshly available. The challenge
of coffee overproduction was exaggerated in restaurants where coffee
was included in the meal price or offered with free rells, as this
Table 3
Summary of descriptive statistics on observed coffee waste indicators, compiled from daily data (n =76) across all six Swedish restaurants participating in coffee waste
quantication
Indicator/per day Min Q1 Median Q3 Max Mean Std Lower 95% CL mean Upper 95% CL mean
Waste (g) 0 423 2150 5138 12500 3262 3241 2521 4003
Waste per guest (g) 0 5 9 15 83 13 16 10 17
Waste per employee (g) 0 200 520 1071 2633 739 684 583 895
Waste (%) 0 4.9 7.7 14.5 40.0 10.5 8.6 8.5 12.5
Fig. 1. Boxplot of coffee waste (%) in the six participating Swedish restaurants (A-F) based on daily quantication data (n =22, 20, 19, 10, 10, 5). Centre lines show
the median, box limits the 25
th
and 75
th
percentiles (determined by Julia software) and whiskers 1.5 times the interquartile range (25
th
to 75
th
percentile). Outliers
are represented by dots.
Fig. 2. Relationship between the amount of liquid coffee produced and daily coffee waste generation in the different restaurants (A-F).
C. Malefors et al.

Resources, Conservation & Recycling 211 (2024) 107853
5
complicated consumption predictions. This problem was summarised
thus in a comment by one interviewee: “We offer coffee with free rells,
which makes it more difcult to predict. From ten sold coffees, con-
sumption can double to twenty cups”. The trend for overproduction was
also inuenced by the cultural perception of coffee as a readily available
commodity, as reected in this statement by another respondent:
“There’s a well-established expectation for unlimited coffee consump-
tion. It’s a balance between running out of coffee or having too much at
the end of the day”.
Another aspect related to infrastructure and time. One respondent
identied a key infrastructure limitation contributing to coffee over-
production, namely that coffee machines are designed to brew only full
batches. This design constraint compels production of excess coffee, as
there is no option for smaller quantities. Other respondents echoed this
sentiment and reported additional practical difculties, such as the time-
intensive process of precisely measuring coffee grounds for anything less
than a full batch, which further exacerbates the waste issue.
Survey results showed that most entities (11 out of 18) actively
measure solid food waste, aiming to reduce waste while monitoring
costs and sales. This demonstrates a proactive approach towards waste
management. Regular measurement of solid food waste was conrmed
by two interviewees. While coffee waste quantication was deemed
straightforward by all interviewees, with two managing it with minimal
team assistance, there was variance in engagement levels due to
perceived difculties in team involvement. Those who did not engage in
measuring coffee waste cited negligible amounts or did not view it as an
issue.
Overall, there was good awareness of coffee waste among the re-
spondents, with 19 recognising its signicance (Table 4), motivated and
driven by nancial, environmental or emotional factors. In addition,
there was an expressed external push from companies, the industry or
customers towards addressing coffee waste and quantifying it. In addi-
tion to this, respondents mentioned environmental aspects and not
wasting food as important and a way to save on resources, e.g. less waste
results in less work to get rid of the waste. One restaurant recognised the
intrinsic value of waste quantication data for organisational benets,
suggesting that such metrics can inform and improve operational ef-
ciency. However, only one restaurant stated that it will continue with
quantication of coffee waste.
Time constraints were a prominent barrier to quantifying coffee
waste according to the restaurants. Respondents also cited organisa-
tional challenges, including the absence of structured teams able to
integrate new tasks effectively. There was also the aspect of people not
viewing the problem as large enough and feeling that change is difcult
to achieve. Some entities do not measure waste, believing their levels are
too low for quantication or because they feel sufciently informed
about their waste patterns. Notably, two survey responses indicated that
the use of automatic coffee machines, which are believed to generate no
waste, eliminates the need for waste tracking.
3.2. Coffee waste extrapolated to national level
When the collected data were extrapolated to national level, the
results showed that around 17,800 tonnes of coffee waste are generated
annually in Sweden. As shown in Table 5, this nationwide estimate was
compared with values obtained in simulations. The mean gures derived
from the Gamma and Exponential distribution were nearly equivalent to
the extrapolated national gure. However, the median value from the
LogNormal distribution simulations, based on identical parameters, was
about 11.5% higher (+2,049 tonnes).
Table 4
Causes of coffee waste, barriers, and drivers for quantication according to the Swedish restaurants interviewed (n=4) and surveyed (n=18) in this study
Causes of coffee waste
Theme Response Answers (n)
Production strategies Concerns regarding coffee shortages, tendency to overproduce due to demand issues 15
There must be coffee to offer 3
Measuring out coffee grounds for half-sized brews is time-consuming 1
Coffee tends to lose its avour over time, necessitating a fresh brew 1
Lack of resources Lack of attention to customer ow or time to analyse it, due to high workload 6
Coffee machine does not allow brewing less than full batches 1
Business offer Coffee is included in meal/uncharged, making customer demand difcult to understand 2
Free rell of coffee is included in the price, making customer demand unpredictable 1
Drivers of quantication
Theme Response Answers (n)
Awareness Minimising waste leads to cost savings 10
Environmental considerations, sustainability, and waste reduction 6
Feels sad to waste food 2
Able to leverage the outcomes of quantication efforts 1
Demand from outside Demand from company or industry 3
Demand from guests to prevent coffee waste 1
Resource efciency Reducing waste decreases the workload 2
Conserving resources 1
Barriers to quantication
Theme Response Answers (n)
Lack of resources Don’t have time to quantify coffee waste 19
Cost of staff to perform quantication 1
It is unreasonable and difcult to quantify waste 1
Low or non-existent waste Already know how much is wasted 3
Have low level of coffee waste and see no point in quantifying 2
Operate a small-scale coffee production and consider quantication unnecessary 1
Low waste due to serving waste being re-used 1
Scepticism about the issue Difcult to do something about coffee waste 3
Viewed as a minor and insignicant issue 3
Coffee waste is unavoidable 1
Team structure Staff variability complicates the quantication process 2
Insufcient stafng to perform quantication 1
C. Malefors et al.

Resources, Conservation & Recycling 211 (2024) 107853
6
The simulations using the Gamma and Exponential distributions
suggested that the average amount of coffee waste produced annually by
the Swedish food service sector is around 17,800 tonnes, with a margin
of error of ~340-380 tonnes (accounting for two standard deviations).
Compared with the solid food waste generated by the Swedish food
service sector annually, which amounts to 65,000 tonnes, coffee waste
represented about 21% of total food waste. Coffee waste per capita in
Sweden based on these results was estimated to be 1.7 kg/capita/year.
4. Discussion
Understanding where food waste comes from is essential to prevent
its generation. This study showed that 10% of coffee brewed in Swedish
food service establishments is never served to customers and is left in the
brewing machine to be discarded. When this gure is extrapolated to
national level, it suggests that the Swedish restaurant sector generates
roughly 17,800 tonnes (margin of error ±380 tonnes) of liquid coffee
waste every year. This is in addition to the 65,000 tonnes of solid food
waste already identied in the Swedish restaurant sector. Thus coffee
waste represents a previously unaccounted for 21% increase in waste,
not including waste left in consumers’cups. This fraction of waste
amounts to around 1.7 kg per person per year, which can be compared
with 135 kg of overall food waste per person each year in the whole
supply chain (Swedish Environmental Protection Agency, 2024) .
Considering UN Sustainable Development Goal 12.3, which aims to
cut food waste in half by the year 2030, these results suggests that the
target reduction for coffee waste should be around 5%, depending on the
volume of coffee waste left in cups. However, achieving this reduction
target for coffee waste may pose varying challenges for different res-
taurants that serve coffee, since not all restaurants share the same
operational traits. For example, those establishments that brew the
largest quantities of coffee daily do not necessarily generate the highest
rates of waste in proportional terms. Prior research on solid food waste
suggests that establishments generating the most waste (where previous
interventions have not been made) have the greatest potential for
reduction (Eriksson et al., 2019). Quantifying food waste in such es-
tablishments could be an effective rst step in addressing the issue. By
doing so, these establishments could probably achieve rapid initial re-
ductions in waste (low-hanging fruit). This scenario is probably also
applicable to liquid waste as a whole, and specically to coffee waste.
A signicant limitation of this study is that it was based on quanti-
cation data from only six restaurants, and thus the ndings only pro-
vide an initial insight into liquid waste issues. Over 200 restaurants were
invited to participate in this study, but only six agreed to do so. This
likely resulted in a selection bias where only the most interested res-
taurants participated, potentially inuencing the results (Canali et al.,
2017;Silvennoinen et al., 2019). However, such selection bias is not
unique to this study. The range of waste levels recorded was broadly
consistent with the variation reported in other studies on solid waste,
where the level ranged from 0% to 40% of the food served per meal
(Eriksson et al., 2017;Malefors et al., 2022,2019)This study showed
that quantifying coffee waste in restaurants was a barrier for many of the
staff (surveyed and interviewed), with most of the 19 respondents
claiming that they did not have time to quantify the amount of coffee
waste. However, according to feedback from 15 restaurants, the primary
reason for coffee waste is the fear of coffee shortages, leading to over-
production and ultimately to waste. This concern has been identied in
other studies as a contributing factor to food waste generation (e.g.
Malefors et al., 2021;Steen et al., 2018). Six of the restaurants reported
that they do not have time to analyse customer ows, nding it more
convenient to brew full batches instead. Adding to the complexity,
particularly in Sweden, is the cultural expectation that a coffee purchase
often includes the option for one or more rells. This tradition makes it
difcult for coffee-serving establishments to predict demand with ac-
curacy, as highlighted by one of the interviewees. However, it may be
possible to utilise forecasting methods to address this issue. By inte-
grating point-of-sale data from cash registers with records of actual
coffee consumption, restaurants could more accurately predict the
amount of coffee to brew each day. This approach would be most ad-
vantageous if it also encompassed other types of food, thereby offering a
cost-effective solution for reducing waste.
Ten of the participating restaurants indicated that their waste
reduction efforts are primarily motivated by the prospect of cost savings.
However, only six of these establishments engaged in actual waste
quantication to gain a clear understanding of their specic cases, and
four reported that their future quantication efforts would be prompted
mainly by demand from external sources. Participants in the quanti-
cation study reported that they chose to participate out of interest in the
subject. However, only one restaurant planned to continue quantifying
coffee waste, whereas the others found the process too time-consuming,
despite acknowledging that the quantication method was relatively
simple. This disconnect between expressed interest in waste reduction
and actual commitment to ongoing quantication aligns well with
ndings by Filimonau and Coteau (2019) that managers’willingness to
engage in food waste reduction is inuenced by their perceived value of
such activities. In this study, the restaurants’interest might not be
considered strong enough to justify the time investment required for
quantication. Similarly, the potential cost savings that could motivate
waste quantication did not seem to outweigh the effort involved. This
trade-off has also been observed in other parts of the food supply chain
(Pietrangeli et al., 2023)
Extrapolating the ndings to national scale underlined the signi-
cance of selecting appropriate distribution for data tting, which can
signicantly inuence the scaling results. In our case, one distribution’s
extrapolated outcome was 11.5% higher than the others. Instances
where restaurants reported no waste (a total of six days) were omitted
from the data tting. In the simulations, however, the occurrence of
zero-waste events was factored in as a proportion to generate realistic
simulation outcomes. A broader dataset with more restaurants partici-
pating in waste quantication would offer a more comprehensive un-
derstanding of the situation. The conclusion that liquid coffee waste
represents 21% of restaurant food waste is in line with ndings by e.g.
Ahmed et al. (2018) that around 19% of total waste consists of
post-consumer liquid waste (beverages and soups). However, this is
much higher than the level reported by Sehnem et al. (2022), who
estimated that 4% of total food waste comes from drinks or desserts, and
by Filimonau et al. (2019), who concluded that 4-19% of the food waste
in coffee shops comes from coffee grounds. However, both those studies
used survey methodology, potentially underestimating waste levels.
If this study had encompassed the entire liquid food waste stream,
the results might have differed. Food service establishments might be
more incentivised to participate in liquid food waste mitigation if it
included a broader range of beverages, thereby making prevention ef-
forts more impactful in overall reduction of food waste. However,
studies on other parts of the food supply chain have suggested that coffee
and tea are the main fractions to consider in liquid waste estimation
Table 5
Extrapolated coffee waste (tonnes per year, t/y) in Sweden
compared against the outcomes from Monte Carlo simu-
lations, performed 10,000 times, with variability captured
using different distributions (Gamma, Exponential,
LogNormal). The simulated values are mean ±two stan-
dard deviations for each distribution model. The gures
are not rounded, but the number of signicant digits does
not indicate high precision.
Method Coffee waste (t/y)
Extrapolated 17,812
Simulated
Gamma 17,840±339
Exponential 17,836±383
LogNormal 19,861±597
C. Malefors et al.

Resources, Conservation & Recycling 211 (2024) 107853
7
surveys (Swedish Environmental Protection Agency, 2021;Van Dooren
et al., 2019). If liquid waste would be quantied with the same level of
efforts that is used to quantify solid waste, the overall food waste sta-
tistics would potentially look very different from today where liquid
waste is normally overlooked and therefore appear to not be a prob-
lematic loss of resources.
For waste quantication to take place, the process needs to be simple
and ideally automatic (Malefors et al., 2024), but the costs of such
automation is currently higher than the potential savings (Goossens
et al., 2022), at least when it comes to liquid waste. Similarly, Filimonau
and Coteau (2019) concluded that applying evidence-based forecasts for
guest attendance and training teams could achieve waste reductions, but
would involve high initial cost and uncertain overall gains.
The interview responses indicated that restaurant staff and cus-
tomers view coffee as having low economic value, as reected in the
business model of including coffee in lunch menus and providing free
rells. This creates highly variable demand and a high expectation that
enough coffee will be available, barriers that are difcult to overcome
even though coffee waste has an economic cost and a non-negligible
environmental impact (Eneroth et al., 2022). It will be difcult for
policymakers to address coffee waste specically and, as long as the
price of resources such as coffee is low in comparison to staff costs, there
is a high risk that coffee waste will not be prioritised. Policymakers have
the power to shift the cost balance for businesses by putting higher taxes
on resource outtake rather than on staff. The rst step is to investigate
the scope of the problem and establish the necessary knowledge base for
policymakers. Under current EU regulations, waste of beverages such as
coffee can be voluntarily quantied and reported (European Commis-
sion, 2019). Considering the considerable amount of coffee waste
generated annually, including liquid food waste in national waste sta-
tistics in Sweden and other countries can be a simple approach by pol-
icymakers to address the problem of coffee waste in restaurants.
5. Conclusions
This study found that on average, 10% of coffee brewed daily in
Swedish restaurants is wasted (excluding coffee grounds and coffee left
in cups). This corresponds to 3.3 kg of coffee waste per day, 13 g per
guest per day and 730 g per employee per day. However, there was high
daily variation, with coffee waste level ranging from 0% to 40% of
brewed coffee. Waste was mainly due to concerns about coffee short-
ages, leading to overproduction of coffee, according to interviews with
restaurant staff. According to most of the surveyed or interviewed em-
ployees, cost saving was the main driver of quantication, while lack of
time was the main barrier to quantication.
Upscaling the ndings to national level revealed that Swedish res-
taurants generate 17,800 tonnes coffee per year. This highlights the
importance of including liquid waste in national statistics and of
designing policies to counteract this sustainability issue.
CRediT authorship contribution statement
Christopher Malefors: Writing –review &editing, Writing –orig-
inal draft, Visualization, Methodology, Formal analysis, Data curation,
Conceptualization. Rosa Hellman: Writing –review &editing, Writing
–original draft, Validation, Data curation, Conceptualization. Amanda
Sj¨
olund: Writing –review &editing, Writing –original draft. Mattias
Eriksson: Writing –review &editing, Conceptualization.
Declaration of competing interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Data availability
Data will be made available on request.
Acknowledgements
This work was supported by the Swedish Research Council for Sus-
tainable Development (Formas) grant number 2023-01908.
Supplementary materials
Supplementary material associated with this article can be found, in
the online version, at doi:10.1016/j.resconrec.2024.107853.
References
Ahlgren, S., Morell, K., Hallstr¨
om, E., 2022. Mapping of biodiversity impacts and hotspot
products in Nordic food consumption, RISE Report 2022:25. RISE.
Ahmed, S., Shanks, C.B., Lewis, M., Leitch, A., Spencer, C., Smith, E.M., Hess, D., 2018.
Meeting the food waste challenge in higher education. Int. J. Sustain. High. Educ. 19,
1075–1094. https://doi.org/10.1108/ijshe-08-2017-0127.
Braun, V., Clarke, V., 2006. Using thematic analysis in psychology. Qual. Res. Psychol. 3,
77–101. https://doi.org/10.1191/1478088706qp063oa.
Burnham, K.P., Anderson, D.R., 2004. Model Selection and Multimodel Inference.
Springer New York, New York, NY. https://doi.org/10.1007/b97636.
Büsser, S., Steiner, R., Jungbluth, N., 2008. LCA of Packed Food Products.
Canali, M., Amani, P., Aramyan, L., Gheoldus, M., Moates, G., ¨
Ostergren, K.,
Silvennoinen, K., Waldron, K., Vittuari, M., 2017. Food Waste Drivers in Europe,
from Identication to Possible Interventions. Sustainability 9, 37. https://doi.org/
10.3390/su9010037.
Cederberg, C., Persson, U.M., Schmidt, S., Hedenus, F., Wood, R., 2019. Beyond the
borders –burdens of Swedish food consumption due to agrochemicals, greenhouse
gases and land-use change. J. Clean. Prod. 214, 644–652. https://doi.org/10.1016/j.
jclepro.2018.12.313.
Eneroth, H., Karlsson Potter, H., R¨
o¨
os, E., 2022. Environmental impact of coffee, tea and
cocoa–data collection for a consumer guide for plant-based foods. Department of
Energy and Tecnology 1654–9406. https://doi.org/10.54612/a.2n3m2d2pjl.
Eriksson, M., Malefors, C., Callewaert, P., Hartikainen, H., Pietil¨
ainen, O., Strid, I., 2019.
What gets measured gets managed –Or does it? Connection between food waste
quantication and food waste reduction in the hospitality sector. Resour. Conserv.
Recycl. X 4, 100021. https://doi.org/10.1016/j.rcrx.2019.100021.
Eriksson, M., Osowski, C.P., Malefors, C., Bj¨
orkman, J., Eriksson, E., 2017. Quantication
of food waste in public catering services –A case study from a Swedish municipality.
Waste Manag 61, 415–422. https://doi.org/10.1016/j.wasman.2017.01.035.
European Commission, 2022. Guidance on reporting of data on food waste and food
waste prevention according to Commission Implementing Decision (EU) 2019/2000
version of June 2022.
European Commission, 2019. Commission Delegated Decision (EU) 2019/1597 of 3 May
2019 supplementing Directive 2008/98/EC of the European Parliament and of the
Council as regards a common methodology and minimum quality requirements for
the uniform measurement of levels of food waste (Text with EEA relevance.).
European Commission, 2002. Regulation (EC) No 178/2002 of the European Parliament
and of the Council of 28 January 2002.
European Union, 2020. Farm to Fork Strategy- For a fair, healthy and environmentally-
friendly food system.
EUROSTAT, 2008. NACE rev. 2. Ofce for Ofcial Publications of the European
Communities. Luxembourg.
FAOSTAT, 2023. Crops and livestock production (Region: world, Elements: production
quantity, Items: coffee green, Year: 2021 (Updated 2023-03-24). Food and
Agriculture Organization of the United Nations.
Filimonau, V., De Coteau, D.A., 2019. Food waste management in hospitality operations:
A critical review. Tour. Manag. 71, 234–245. https://doi.org/10.1016/j.
tourman.2018.10.009.
Filimonau, V., Krivcova, M., Pettit, F., 2019. An exploratory study of managerial
approaches to food waste mitigation in coffee shops. Int. J. Hosp. Manag. 76, 48–57.
https://doi.org/10.1016/j.ijhm.2018.04.010.
Geislar, S., 2020. Quantifying food waste- Food waste audits, surveys, and new
technologies. In: Reynolds, C., Soma, T., Spring, C., Lazell, J., Soma, T., Spring, C.,
Lazell, J. (Eds.), Routledge Handbook of Food Waste. Routledge/Taylor &Francis
Group, pp. 255–268.
Goossens, Y., Leverenz, D., Kuntscher, M., 2022. Waste-tracking tools: A business case for
more sustainable and resource efcient food services. Resour. Conserv. Recycl. Adv.
15, 200112 https://doi.org/10.1016/j.rcradv.2022.200112.
Heikkil¨
a, L., Reinikainen, A., Katajajuuri, J.-M., Silvennoinen, K., Hartikainen, H., 2016.
Elements affecting food waste in the food service sector. Waste Manag 56, 446–453.
https://doi.org/10.1016/j.wasman.2016.06.019.
Ho, T.Q., Hoang, V.-N., Wilson, C., 2022. Sustainability certication and water efciency
in coffee farming: The role of irrigation technologies. Resour. Conserv. Recycl. 180,
106175 https://doi.org/10.1016/j.resconrec.2022.106175.
C. Malefors et al.

Resources, Conservation & Recycling 211 (2024) 107853
8
International Food Waste Coalition, 2022. Measuring &Reporting Food Waste in the
Hospitality &Food Service Sectors.
International Trade Centre, 2021. The Coffee Guide: Fourth Edition. United Nations. 10.1
8356/9789210010511.
Johnson, L.K., 2020. Produce loss and waste in agricultural production in. In:
Reynolds, C., Soma, T., Spring, C., Lazell, J. (Eds.), Routledge Handbook of Food
Waste. Routledge/Taylor &Francis Group, pp. 81–92.
Kjeldgaard, D., Ostberg, J., 2007. Coffee grounds and the global cup: Glocal consumer
culture in Scandinavia. Consum. Mark. Cult. 10, 175–187. https://doi.org/10.1080/
10253860701256281.
Lundberg-Hall´
en, N., ¨
Ohrvik, V., 2015. Key foods in Sweden: Identifying high priority
foods for future food composition analysis. J. Food Compos. Anal. 37, 51–57.
https://doi.org/10.1016/j.jfca.2014.09.008.
Malefors, C., Callewaert, P., Hansson, P.A., Hartikainen, H., Pietilainen, O., Strid, I.,
Strotmann, C., Eriksson, M., 2019. Towards a Baseline for Food-Waste Quantication
in the Hospitality Sector-Quantities and Data Processing Criteria. Sustainability 11.
https://doi.org/10.3390/su11133541.
Malefors, C., Strid, I., Eriksson, M., 2022. Food waste changes in the Swedish public
catering sector in relation to global reduction targets. Resour. Conserv. Recycl. 185,
106463 https://doi.org/10.1016/j.resconrec.2022.106463.
Malefors, C., Strid, I., Hansson, P.-A., Eriksson, M., 2021. Potential for using guest
attendance forecasting in Swedish public catering to reduce overcatering. Sustain.
Prod. Consum. 25, 162–172. https://doi.org/10.1016/j.spc.2020.08.008.
Malefors, C., Svensson, E., Eriksson, M., 2024. Automated quantication tool to monitor
plate waste in school canteens. Resour. Conserv. Recycl. 200, 107288 https://doi.
org/10.1016/j.resconrec.2023.107288.
Moberg, E., Karlsson Potter, H., Wood, A., Hansson, P.-A., R¨
o¨
os, E., 2020. Benchmarking
the Swedish diet relative to global and national environmental
targets—identication of indicator limitations and data gaps. Sustainability 12,
1407. https://doi.org/10.3390/su12041407.
Partt, J., Barthel, M., Macnaughton, S., 2010. Food waste within food supply chains:
quantication and potential for change to 2050. Philos. Trans. R. Soc. B Biol. Sci.
365, 3065–3081. https://doi.org/10.1098/rstb.2010.0.
Pietrangeli, R., Eriksson, M., Strotmann, C., Cicatiello, C., Nasso, M., Fanelli, L.,
Melaragni, L., Blasi, E., 2023. Quantication and economic assessment of surplus
bread in Italian small-scale bakeries: An explorative study. Waste Manag 169,
301–309. https://doi.org/10.1016/j.wasman.2023.07.017.
Quadra, G.R., Paranaíba, J.R., Vilas-Boas, J., Roland, F., Amado, A.M., Barros, N.,
Dias, R.J.P., Cardoso, S.J., 2020. A global trend of caffeine consumption over time
and related-environmental impacts. Environ. Pollut. 256, 113343 https://doi.org/
10.1016/j.envpol.2019.113343.
Sehnem, S., Pereira, L.H., Santos, S., Bernardy, R.J., Lara, A.C., 2022. Management of
food waste in restaurants by way of circular practices. J. Mater. Cycles Waste Manag.
24, 1020–1036. https://doi.org/10.1007/s10163-022-01377-x.
Silvennoinen, K., Nisonen, S., Pietilainen, O., 2019. Food waste case study and
monitoring developing in Finnish food services. Waste Manag 97, 97–104. https://
doi.org/10.1016/j.wasman.2019.07.028.
Steen, H., Malefors, C., Roos, E., Eriksson, M., 2018. Identication and modelling of risk
factors for food waste generation in school and pre-school catering units. Waste
Manag 77, 172–184. https://doi.org/10.1016/j.wasman.2018.05.024.
Swedish Board of Agriculture, 2023. Direct consumption of good. Year 2000-2020.
Swedish Environmental Protection Agency, 2024. Livsmedelsavfall i Sverige 2022.
Swedish Environmental Protection Agency, 2023. Milestone targets [WWW Document].
https://www.naturvardsverket.se/en/om-miljoarbetet/swedish-environmental-obje
ctives/milestone-targets/(accessed 3.1.undenedAD).
Swedish Environmental Protection Agency, 2022. Livsmedelsavfall i Sverige 2020 (Food
waste in Sweden 2020). INFO-serien, 8891.
Swedish Environmental Protection Agency, 2021. M¨
angd mat och dryck via avloppet
från svenska hushåll 2021 (Food and drink disposed down the drain from Swedish
household 2021) (No. 6983). Swedish Environmental Protection Agency.
Swedish National Food Agency, 2023. Nationell m¨
atmetod f¨
or matsvinn i offentliga k¨
ok
(National measurement method for food waste in public kitchens) [WWW
Document]. https://www.livsmedelsverket.se/matvanor-halsa–miljo/maltider-i-va
rd-skola-och-omsorg/matsvinn-i-storkok/handbok-for-minskat-matsvinn/minska
-matsvinnet–sahar-gor-du/mat-och-folj-upp/nationell-matmetod-for-matsvinn.
accessed 3.1.undenedAD.
Tostivint, C., ¨
Ostergren, K., Quested, T., Soethoud, H., Stenmarck, Å., Svanes, E.,
O’Connor, C., 2016. Food waste quantication manual to monitor food waste
amounts and progression. FUSIONS.
UNEP, 2021. Food Waste Index Report 2021. Nairobi.
United Nations, 2015. Transforming our world: the 2030 agenda for sustainable
developement. United Nations.
Van Dooren, C., Janmaat, O., Snoek, J., Schrijnen, M., 2019. Measuring food waste in
Dutch households: A synthesis of three studies. Waste Manag 94, 153–164. https://
doi.org/10.1016/j.wasman.2019.05.025.
World Resources Institute, 2016. Food Loss and Waste Protocol. Food Loss Waste
Account. Report. Stand.
WWF, 2022. Kaffe, te och kakaos i Vegoguiden- sammanfattning av bakgrundsrapport
(Coffee, tea and cocoa in the consumer guide for plant- based products- summary of
background report).
Xue, L., Liu, G., Partt, J., Liu, X., Van Herpen, E., Stenmarck, Å., O’Connor, C.,
¨
Ostergren, K., Cheng, S., 2017. Missing Food, Missing Data? A Critical Review of
Global Food Losses and Food Waste Data. Environ. Sci. Technol. 51, 6618–6633.
https://doi.org/10.1021/acs.est.7b00401.
C. Malefors et al.