Balancing supply against demand for food is one measure that can be taken to reduce food waste, since the surplus often goes straight to the bin. Other ways of preventing waste can relate to infrastructure, in which case kitchens and organisations would need to balance the costs of implementing the measures against benefits these measures might provide.
It would be unrealistic to propose that all organisations should abandon the concept of satellite kitchens and convert them to production units. A better option would be (when feasible) to supply satellite kitchens with equipment to handle surplus food, such as cooling or heating equipment. This infrastructure-related change, together with forecasting, could be a promising approach for kitchens to match portion provision to the actual number of guests. In Paper IV, the four different interventions tested were chosen according to best available technology accepted by the staff in a catering organisation. One of these interventions (forecasting) aimed at reducing serving waste, whereas the other three measures primarily targeted reduction of plate waste.
6.4.1 Awareness campaigns
Awareness campaigns are simple measures by which organisations can
either educate their staff about issues with food waste or educate guests
who eat in the organisation’s premises. In practice, this often involves
displaying message boards so that staff or guests easily receive the
message. The canteens that implemented the awareness campaign in Paper
IV did so with the ambition of lowering their guests’ plate waste. In the
canteens where the awareness campaign was deployed, plate waste was
reduced from 37 g to 24 g per guest, corresponding to a 35% reduction.
Influencing guest behaviour is no easy task, however, and is often met with varying success. For instance, Visschers et al. (2020) examined the effects of an awareness campaign with the effects of reducing plate size in two university canteens. They found that reducing plate size achieved an actual reduction in food waste, whereas the awareness campaign alone was not sufficient to reduce food waste. This may indicate that campaigns affect age groups differently depending on how they are designed, or if other or prior measures have influenced the guests. Another approach suggested by Filimonau et al. (2022a) is for preschools and parents to work together to raise children’s awareness of food waste. This may have the potential to lower food waste not only in the educational setting, but also in families (Liz Martins et al., 2020)
6.4.2 Tasting spoons
The fairly simple measure of providing tasting spoons in canteens gave a reduction in plate waste of 22%, to 21 g of food waste per guest. However, there was also an observed tendency for a shift towards more serving waste.
Tocco Cardwell et al. (2019) concluded that providing tasting spoons, together with clear and consistent instructions regarding portioning, can reduce the edible food waste fraction significantly. Providing tasting spoons together with awareness campaigns is a cheap tool that canteens can implement easily as a starting point, since it has a low entry barrier and requires very little in terms of material and time from kitchen staff.
6.4.3 Plate waste tracker
The canteens in Paper IV that used the plate waste tracker had normal levels of plate waste (median levels of around 19 g/guest) and saw a reduction to 12 g per guest, which represents a very low level of plate waste. To put this in perspective, the plate waste fraction for primary schools that contributed data to Papers I and V was around 21 g per guest.
This is in line with plate waste values reported by the Swedish National
Food Agency food in a mapping covering the year 2020, where the median
plate waste in primary school was 20 g per guest. A larger reduction in
plate waste might have been observed if the plate waste tracker had been
placed in canteens with higher initial plate waste levels. However, in the
canteens that used the plate waste tracker, the level of serving waste was
that there is a need to monitor waste processes that are not the intended target for the intervention, so that potential spill-over effects can be detected (as in the case of the tasting spoons, where there was a tendency to shift waste to the serving waste fraction). The plate waste tracker also enabled the staff to get an understanding of why meals were wasted on a daily basis, making it possible to adjust their meal planning.
The plate waste tracker could possibly also be used to monitor the level of food waste achieved by other simultaneous measures, such as pedagogic meals, where teachers devote some lesson to food waste or a similar topic connected to food and also eat with their pupils. Since the plate waste tracker enables waste to be displayed when it is recorded, there is potential to connect this information to the teaching schedule, to get a rough idea of classes taking part in the pedagogic meal and whether this educational concept reduces food waste.
6.4.4 Forecasting guest attendance
Paper III demonstrated that by using forecasting techniques, it was possible to predict quite accurately the number of guests that would attend a certain meal. The overall best-performing forecasting model identified, based on neural network, was tested in Paper IV. The forecasting intervention, which targeted serving waste, was successful overall, with a reduction of 49% in the two canteens that implemented this intervention.
This was higher than the anticipated 20-40% reduction indicated in Paper
III, although it was difficult to isolate cause-effect relations in the type of
quasi-experimental set-up used in Paper IV. A drawback with forecasts is
that they are sometimes wrong and that they underestimate actual demand
quite often if they include no margin, which would lead to shortages of
food. Shortages of food in canteens are undesirable and therefore there
needs to be some safety margin in place, along with some sort of backup
food option. Figure 11 showed margins associated with a forecast and
number of portions that would be need to taken from a backup source and
how often. Even with a 10% margin, kitchens that serve many guests would
still need to have 30+ portions ready for 22-43 days of a school year
according to the findings in Paper III. For kitchens that have a steady
stream of guests and less attendance variability, it would be sufficient to
have a 10% margin (or even lower) to their forecast and achieve zero days
of underestimation. However, it is possible to have an ‘optimal’ margin in
place, as illustrated in Paper III, which optimised the margin from an economic perspective. Today, kitchens do not seek to identify the optimal number of portions to make because the system is designed in most cases to produce meals for all students enrolled, due to fear of shortages. One way of overcoming this fear could be by step-wise adjusting the portion numbers downwards, applying appropriate margins, knowing approximately how many times a shortage is likely to occur and having a backup stock of food ready, which would hopefully lead to less food waste.
A solution to deal with food shortages might be to meet unexpected demand using the food stock in the contingency plan that is now being established across strategic public catering establishments (Swedish National Food Agency, 2020a). This stock is intended for emergency situations, but if replaced instantly could meet two purposes, namely reducing food waste in daily operations and ensuring that the emergency stock is fresh and ready to use. Another policy-oriented solution could involve making it more expensive to throw away food, as indicated by the waste penalty cost and the sensitivity analysis in Paper III.
6.4.5 Reference canteens
In Paper IV, the canteens that had none of the intervention measures in
place reported total waste of 58 g per guest. This was reduced to 41 g per
guest (a reduction of 29%) after the interventions were rolled out. The plate
waste fraction for the reference canteens decreased from 15 g to 7 g per
guest, a reduction of 53%, and the serving waste decreased from 39 to 28 g
per guest, a reduction of 28%. This indicates that the ongoing systematic
work to reduce food waste in the participating organisation, which has
quantification data since 2014, seems to be having an effect and that further
interventions should be targeted at specific canteens that have identified
potential problems. The questionnaire used in Paper IV asked three
questions: What process generates the most waste? How many dinners do
you anticipate? and What size of portions do you serve?. It was thus fairly
simple, but could be used to identify areas for further improvement that
could be targeted by specific interventions. This would make
implementation of the intervention more precise and would potentially
maximise the potential for waste reduction, making it a helpful tool for
kitchen staff and managers. In particular, canteens that have greater
better understand how many guests might arrive, would benefit from
targeting these areas and get the most out of their food waste reduction
