Scoring methodology for comparing the environmental performance of food packaging

The objective of this work was to propose an environmental scoring tool for food packaging based on the assessment of three key pillars of packaging sustainability: Materials, Functionality and Post‐Usage fate. A participatory process involving relevant food‐packaging experts and end users was applied to define the relevant criteria for each pillar. Each criterion was translated into a question for users, and the answers are converted into a score between 0 (worst option) and 1 (best option) per pillar. For the Materials pillar, two scores were computed from a streamlined calculation of resource (CEENE) and carbon footprints (IPCC) while for the Functionality and Post‐Usage pillars, scores were computed from Yes/No answers provided by the users. A fourth pillar considers the potential risk of long‐term environmental pollution. Then, the packaging options for the same food are ranked according to the Borda voting rule, considering the individual rankings obtained for the various pillars. The proposed methodology was applied to three commercial (milk and sugar) and non‐commercial (strawberry) packaging case studies. The obtained ranking is discussed with respect to current knowledge in the field. The provided methodology is easy to understand, science based, and combines quantitative and qualitative assessments. The developed tool could be handled by non‐experts in environmental sciences such as food manufacturers, packaging converters and policy makers. The resulting indicators provide answers to user concerns regarding the environmental impacts of food packaging and guide their choice of the most sustainable option. The proposed scoring method considers the functionality of the packaging with respect to preserving food and reducing food waste, which is rarely considered in packaging environmental assessments.


| INTRODUCTION
By preserving food from deterioration, packaging plays a key role in minimising food loss and waste (FLW) and the enormous concomitant impact of FLW on the environmental footprint of the food industry. 1,2 However, food packaging is mostly perceived as an environmental problem because its production and disposal are associated with numerous environmental impacts, notably the unresolved issue of post-usage plastic packaging that is associated with serious health and ecotoxicology risks. [3][4][5] Food packaging prevents FLW and enables the efficient distribution of products, thereby contributing to sustainability by providing indirect positive environmental impacts. [6][7][8] These positive impacts may even counterbalance the direct environmental impacts caused by the production and disposal of packaging. 6,9,10 However, based on the pioneering works of Wikström and Williams,11,12 the balance between positive and negative impacts of food packaging is rarely included in environmental assessments, although it is necessary for directing evidence-based packaging selection and identifying the optimal trade-off between product protection and preservation and the environmental footprint.
Life-cycle analysis (LCA) is one of the most widely used standardised methodologies for assessing the environmental impacts of a product throughout its lifecycle, 13 but it generally focuses only on the direct environmental impacts of food packaging. If the indirect environmental impacts, such as FLW reduction, are included in some LCA studies, the analysis is still insufficient for the numerous cases where the FLW reduction largely compensates for or overcomes the negative impacts. Another drawback of LCA-based assessments is that they cannot be easily used in the packaging-design step because limited data is usually available (e.g., lack of information on material properties and difficulty defining the functional unit and reference flows, two essential elements of LCA methodology 14 ). In addition, the quantity and quality of input data needed to conduct a reliable LCA limit the use of this method for analysing several packaging options at the same time, which would otherwise be very useful for the decision-making and packaging eco-design steps.
To overcome the issue of LCA complexity, less resourcedemanding eco-design tools that can be used to select and/or compare different packaging material options have been developed.
An overview of these tools, not restricted to the case of food packaging, can be found in previous studies. 15,16 The development of eco-design tools is still being intensely studied, especially early-stage assessments 14 including multicriteria approaches where the economic aspect is considered in addition to environmental performance. 17 In the specific case of food packaging, packaging evaluation tools have been proposed by several research groups. However, because of the many indicators that need to be included in these tools, their application scope varies significantly. For example, the Packaging Scorecard 18 maps the performance of packaging based on key aspects of the supply chain (e.g., product protection, volume, weight efficiency and minimal waste generation). However, due to the lack of data and user expertise to properly evaluate some of the criteria, the evaluation could be influenced by subjective judgement. 19 Another study 20 used an analytical hierarchy process to analyse and assess the packaging of fast-moving consumer goods, focusing on five evaluation areas, including environmental aspects. This method did not include the functionality of the packaging that may reduce FWL. Guillard et al. 21 proposed a multi-criteria decision-support system that considers several stakeholder requirements to help the user select suitable packaging for their application. Even if this tool considers food preservation via the choice of an appropriate barrier property for the targeted food application, it is restricted to fresh food that requires a breathable packaging and includes a limited set of criteria. Along with holistic tools that include some environmental criteria, simplified evaluation tools and approaches have been proposed to assess the environmental impact of food packaging ( Table 1). The most sophisticated eco-design tools are mostly based on streamlined LCA (e.g., Piquet, Compass, Bee and ECO-Design of Packaging) and provide environmental-impact data for some packaging categories.
The most recent and complete methodology 19 evaluates packaging systems based on four categories (material production, transport, household and end-of-life). To the best of our knowledge, this is the only existing tool to include the influence of packaging on food waste in households. Although this tool provides a comprehensive overview of the environmental performance of food packaging, it does not allow the aggregation of results into a single score, which may limit the decision-making process.
Based on this background, to address the need of food and packaging companies for a simple, easy-to-handle evaluation tool that considers both direct and indirect packaging impacts (e.g., FWL reduction), this study aimed to develop a scoring methodology using a co-creation process involving experts and stakeholders in the food/ packaging chain. This study was conducted within the framework of the GLOPACK (Granting Society with Low Environmental Impact Innovative Packaging) H2020 European project (2018-2021). The objective was to consider the important life-cycle phases of the packaging including its functionality (usage benefits), focusing on the ability of the packaging to protect food and its long-term fate in the environment. The developed tool does not integrate the impact of the food itself, to avoid biasing the decision-making process. The tool should consider all kinds of materials, not solely plastics. Finally, a final single score must be obtained to easily rank the packaging alternatives and facilitate the decision-making process. No existing tools include all of these criteria.
Target users of the GLOPACK packaging score tool are food companies that use packaging materials. The tool could also be very useful for manufacturers that provide packaging to food companies and more generally for other stakeholders involved in packaging decision-making. The tool focuses on comparing existing packaging formats but could also provide guidance in the packaging optimisation steps to evaluate the sustainability of a new packaging solution.
T A B L E 1 Summary of available tools for environmental assessment of packaging. No score aggregation between performance criteria, which may limit analysis and decision-making when comparing alternatives.
Tool developed in collaboration with Orkla Foods, largest business unit for food supplies in the Nordics, Baltics, and Central Europe. Thus, the tool is restricted to Orkla Foods packaging.
Considers the most important lifecycle phases of packed products. Includes the functionality of the packaging (influence on food waste).
Publicly available (published paper). Could be used by non-experts in LCA.
Eco-score: environmental impact of packed food products (https://docs.scoreenvironnemental.com/) Procedure to calculate a single score that is printed as FOP labels on the consumer sale unit. The score can also be viewed on a smartphone by scanning the bar code. Includes LCA of the food product itself (from Agribalyse project): 14 impacts are aggregated into one global score. Additional criteria considered as bonus or malus are then used to refine the global score: (1) production mode (bonus is given to organic, fair, sustainably certified production modes), (2) origin of the ingredients (a bonus is given depending on the origin of the ingredients and the environmental policy of the origin country), impact on biodiversity (a malus is provided to ingredients identified to have an significant impact on biodiversity), and (4)  to identify a short action plan to improve packaging sustainability.
For packaging and papers. Does not provide a score but an action plan. Functionality of the packaging is not considered.
Free, easy to handle. Provide a short action plan.
Ecolabelling scheme for the Nordic countries (Denmark, Finland, Iceland, Norway, and Sweden) All kinds of goods, not only packed food products.
More than 25 topics covered (primary packaging, raw material constituents, chemical products and substances, quality and regulatory requirements). Could be used to help companies select or design more sustainable solutions.
Only 3 types of food contact materials are considered: packaging for liquid foods, disposables for food, greaseproof paper.
A lot of information and certificates of compliance are required. Functionality of the packaging is not considered.
Link to certification (for Nordic countries).
Can be used as an eco-conception tool out of certification framework. Details of the criteria and background methodology available. Another eco-score calculation approach was recently developed by a French consortium, which is intended to be used to develop front-of-package labels for communicating packaging scores to consumers. Their methodology is based on an LCA of the food product and integrates some additional bonus-malus criteria. A score for the packaging material is calculated and applied as a malus to the overall food-product score from the LCA. The packaging malus is the average of scores related to the origin of the material (e.g., percentage of recycled material in the packaging, certified sustainable paper and use of renewable resources) and its post-usage fate (recyclability, …).
This eco-score does not consider the functionality of the packaging on FLW reduction. This eco-score calculation method is based on information from the Agribalyse database that has been expanded from the farm to consumer's plates including all stages of the food cycle. 22 Most of the tools listed in Table 1 (BEE, ECOdesign of Packaging, FEEL, COMPASS and PIQUET) permit the evaluation of several aspects of the packaging lifecycle, such as resource use, manufacturing and post-usage recyclability, but generally focused on one packaging attribute (e.g., minimising the packaging weight). These tools are applicable for users who need a simple tool to quantify the environmental impact of this modification. However, these tools are insufficient for a global assessment of the environmental impact of different packaging alternatives and could introduce bias into the decision-making process. In addition, they do not consider the packaging functionality.
The tools in Table 1 with the lowest resource demand are based on qualitative or semi-quantitative guidelines (e.g., the APR Design ® Guide for Plastics Recyclability), checklists (e.g., Sustainability Checklist for Packaging), or analytical tools that focus on single aspects, such as selecting a bio-based resource (e.g., Bioplastic Tool) or assessing packaging recyclability (e.g., Recyclass). Although less complex, the assessment is oversimplified by focusing on the single criteria of packaging life, for example, resources or recyclability. Our review of existing packaging eco-design tools showed that most available tools are focused on plastic packaging. Except a recent study, 19  • score only the primary packaging, independently of the food product itself; • consider the most important packaging life phases (pillars) and evaluate both direct and indirect environmental impacts; • one final score per pillar, as a percentage or scale; • aggregation between scores to easily compare different packaging solutions; • applicable to every type of food packaging material: for example, plastics (including bio-based materials), metals, paper, cardboard, glass and multi-materials; • simple but science-based methodology; • user-friendly: the methodology must provide an environmental evaluation of food packaging from easily available information; • public availability: the methodology must be transparent and available to all stakeholders; • updatable with new knowledge and technologies: the scoring methodology must evolve along with the new developments in the packaging field to provide reliable and accurate evaluations over time; and • target audience: food and packaging manufacturers and users, packaging developers, policymakers, staff of technical centres and packaging-development projects.

| Building a list of relevant criteria (Step 2)
The second step of tool development was to build a list of relevant criteria for the environmental evaluation of food packaging (e.g., origin of the resources, recyclability rate and oxygen barrier properties) according to the specifications defined in Step 1. To this end, eight qualitative interviews among experts and stakeholders of the food/packaging chain were performed: two academics, two representatives from food companies, one representative from a food technical centre, two representatives from waste-management companies and one environmental engineer. Face-to-face, video, or phone interview sessions were organised individually with each representative.
Each interviewee was asked to freely explain their opinion about the most important criteria for evaluating food-packaging sustainability.
To help the interviewee, they were prompted with the steps in the life cycle of packaging materials (resources, usage and post-usage fate) to initiate discussions around the key pillars of packaging sustainability.
Arguments provided by interviewees to justify their choice of criteria were also collected and then discussed and validated within the GLOPACK consortium (50 people) during dedicated working sessions.
There was some overlap between responses and the arguments that were categorised during these working sessions. Only the most commonly cited criteria from the interviewees were kept.
In parallel, a careful analysis of the state-of-the-art knowledge related to the environmental sustainability of food packaging was performed to evaluate the most-common criteria cited by the interviewees considering current scientific knowledge. It was important to choose criteria that can be easily estimated and are not dependent on obscure data that is not easily available to the user. In general, the data for finalised packaging formats are easier to find than those for formats under development. It was also important to not select criteria based on preconceived ideas or misinformation about foodpackaging sustainability. Analysis of existing tools described in Table 1, including their limitations, was also used to build the list of criteria considered in the GLOPACK packaging score. We also tried to

| Building the score (Step 3)
Each criterion selected in Step 2 was converted into a question for the user and the questions were grouped into a checklist (one per category). Each criterion is evaluated either by making a quantitative assessment (if feasible and necessary, e.g., for the Materials pillar where a qualitative assessment is irrelevant) or by a qualitative assessment based on a binary Yes/No answer. The answer 'Yes' generally implies a positive environmental attribute while 'No' implies a negative attribute. In both quantitative and qualitative questions, the user has the option to answer 'I don't know', to enable the users to obtain a score even if some information is missing. The answer 'Does not apply' is available if the question is not relevant to the user.
The answers to all questions are then converted into a score between 0 (worst option) and 1 (best option) considering environmental sustainability. A reliability index is calculated for each final score, which is the uncertainty related to the amount of 'I do not know' answers to specific criteria.

| Building the score for quantitative assessment (Materials pillar)
To calculate the score for the Materials pillar, both resource and carbon footprints are calculated. The life-cycle impact assessment method CEENE (Cumulative Exergy Extraction from the Natural Environment) is used to quantify the resource footprint from the cumulative amount of exergy extracted from nature to produce the final product. 23,24 The resources are expressed in terms of exergy, which is based on the second law of thermodynamics and includes both the quality and quantity of the resource. 23 This method is one of those recommended for computing the resource footprint in terms of thermodynamics. 25 The carbon footprint is calculated using the IPCC database (v1.03) using a 100-year timespan. 26 The inventory data were retrieved from the Ecoinvent database v3.5 (Swiss Centre for Life Cycle Inventories, 2018), using the software Simapro v9.0.0.47.
In the case of plastics, the processing of the raw materials (first transformation, such as pellet production from fossil/bio-based plastic materials) and packaging manufacturing (second transformation, such as extrusion or blow moulding) appear separately, while for other materials the final product appears in the database without a distinction between the first and second transformation. Therefore, plastic processing steps were categorised into raw materials (including the first transformation, e.g., pellets) and manufacturing (including the second transformation, e.g., films).
Next, the transport of the packaging materials to the food company is considered assuming various distance ranges: short (0-299 km), medium (300-1000 km) and long (>1000 km). For long distances, 2000 km was selected as a proxy. The impact of transport was obtained from the Ecoinvent database for each transport method for a specific mass and distance. Finally, three geographical regions where processing activities (e.g., production of pellets) and the corresponding supply chain are located were considered. The options are Europe, rest of the World (outside Europe) and Global (Europe + rest of the World). The resource and carbon footprints per component of a specific packaging are computed as follows: where CEENE i (MJex/kg) is the resource footprint of component i and mass i is its mass (kg).
A similar calculation was performed for the carbon footprint: Where IPCC i (kg CO 2 eq/kg) is the carbon footprint of component i.
To translate the absolute Resource footprint and Carbon footprint into a score between 0 and 1, a function f x ð Þ (Equation 3) was used.
Here, f x ð Þ is the Resource footprint score and Carbon footprint score, x is the absolute Resource footprint and Carbon footprint values per packaging and p is a fitting parameter with no physical meaning that controls the slope of the function, that defines the decrease rate from 1 to 0 with respect to x.

| Building the score for qualitative assessments (Functionality and Post-Usage pillars)
To calculate the score for the qualitative assessments (Functionality and Post-Usage pillars), the sum of all 'Yes' answers was calculated and divided by the total of all answers: A reliability index is used to quantify the uncertainty of the score calculated for the specific pillar:

| Comparison of packaging scores (Borda methodology)
For each of the three pillars, we obtain a score between 0 and 1. This is interesting as an absolute comparison of the packaging for each pillar but does not enable an overall comparison. However, aggregating the evaluation results directly (e.g., by means of an average) does not necessarily make sense since they represent different and quite orthogonal aspects of the environmental impact.
To address this, we used a voting rule, 27 which is a function that takes a set of individual rankings of some alternatives as input and outputs a global ranking that represents a collective choice. In our context, each pillar produces a ranking of the packaging alternatives to be compared, from best to worst, where aggregation of these rankings enables a comparison of packaging alternatives considering all pillars.
From the many existing voting rules, we chose the Borda count. For each ranking, an alternative A is assigned several points equal to the number of alternatives that are ranked lower than A; summing the points across all rankings gives an overall score of each alternative and their final ranking.
Given a set of alternatives A = {A 1 , …, A n }, a ranking or preference P i is a linear preorder. For a set of rankings P = {P 1 , …, P n } where P 1 , …, P n are the individual rankings of n voters, the score of an alternative A is defined as: This rule considers the rankings in their entirety while still being quite easy to compute and tends to favour consensus among the voters (i.e., pillars in our case). Alternatives that are acceptable to all voters, rather than those preferred by the majority, tend to be selected, which makes sense in the context of environmental assessments.
For instance, assuming four alternatives A 1 to A 4 and the following three rankings: Alternative A 1 obtains three points from the first ranking, since it is ranked higher than the three other alternatives, zero points from the second ranking as it is last and one point from the third ranking (since it is only better than A 4 ). Similarly, A 2 has six points, A 3 has five points, and A 4 has two points. With these scores, we compute the final ranking: Here, A 2 wins the vote and is the collective choice of the voters.

| Testing the tool (Step 4)
The tool was finally tested on a range of food packaging already available on the market. Three food case studies were selected (1)

| RESULTS AND DISCUSSION
This section is divided into three parts. First, the selection process to determine the set of criteria used to compute the score is presented.
The second part discusses the list of questions, which are associated with the criteria for the three pillars. In the last part, the scores are presented for the three case studies and compared with the results obtained from French Agribalyse-based eco-score tool.

| Qualitative interviews
In total, 93 responses were collected from the eight qualitative interviews with experts and stakeholders in the food/packaging chain.
Anonymous raw data can be downloaded from https://doi.org/10. 57745/SVFQL2. Note that 27% of the arguments cited were linked to the functionality of the packaging (Figure 1), highlighting that this aspect is considered important, although it is often neglected in most existing assessment tools (Table 1) The second-most quoted category of arguments was the end-of-life category (26%, Figure 1), highlighting the stakeholders' concerns related to post-usage fate of the material; 'recyclability', 'potentially recyclable', 'effectively recycled', 'actually recycled' were cited half the time, followed by 'biodegradable', 'compostable', 'duration after use/persistence', 'sorting' and 'separation', which were each mentioned more than three times. It seemed important at this stage that the scoring tool differentiates recyclable and recycled.
Indeed, recyclable refers to the capability of the material to be recycled, but does not guarantee that the material will be effectively recycled. This depends on the sorting and recycling facilities available in the region or country of disposal. In contrast, recycled means that the material is recyclable and was effectively recycled in an appropriate post-treatment facility. Recycling is a broad term describing several processes, which could be mechanical recycling and refer to either a closed-loop process where uncontaminated material with properties close to that of virgin material is recovered or an open-loop process where the material is downcycled into lower-quality products.
A broad definition of recycling may also include the recovery of chemical constituents or chemical recycling and energy recovery. 28,29 In the answers of the interviewees, we did not notice any distinction between recycling types, but we clearly understood that recovery of The fourth and fifth categories of responses ( Figure 1) are respectively, the first transformation (13%) and the second transformation (14%), that is, shaping of the raw materials into final packaging. Energy consumption and its impact on climate change was the most quoted concern related to the first and second transformations.
For the sake of clarity and to simplify the tool, we decided to use three pillars and split the criteria among them as follows: • Materials: resource extraction, transformation processes (first and second) to obtain the final packaging and transport to the food company.
F I G U R E 1 Categorisation of the 93 responses provided by the interviewees.
• Functionality: beneficial characteristics of the packaging that enable a longer shelf life of the product and prevent food waste.
• Post-Usage: potential and currently applied treatments to process the material after its use as food packaging.

| Analysis of the literature
The properties) that maintain a modified atmosphere suitable for food preservation (e.g., avoid the remoistening of a crispy, dry product), 2 but also a matter of ease of use (e.g., individual-serving portions, the packaging should be easy to open, to empty). 38 Using packaging of individual serving sizes reduces the risk that leftovers from a larger family-size package are incorrectly stored after the first use and finally never consumed and discarded. 37,39 However, individual-serving packaging also uses more material proportionally to pack the same quantity of food and thus has a higher environmental impact than larger packages. 32,40 In addition to food protection, waste reduction and convenience, food packaging must provide information such as the product identification and preparation and storage conditions, which are also important to mitigate food waste at the household level. Finally, one criterion highlighted in the literature, but not mentioned by the interviewees, was 'utilisation and handling, including providing for transport and retailing'. 37 We decided to add this criterion to the Functionality pillar, resulting in the following sub-items: (1)

| Building the tool: Checklist & list of questions
The scoring tool comprises a list of questions for each pillar and its sub-items to be answered by the user. The questions should cover all the necessary topics for calculating the final score. For instance, in the case of raw materials, it is important to know the type of resource and its origin. This information could be obtained by asking two questions 'where does the material come from?' and 'what is the resource type (e.g., plastic, paper, glass)'? From the answers to these two questions, we compute the environmental cost of the resource related to extraction and processing, which depends on the country of origin of the raw material, and also the environmental cost of transport. For each pillar and sub-item, the same approach was followed to build the tool.

| Materials
Two indicators, Resource footprint and Carbon footprint, were used to quantify the environmental impact of the materials. To calculate these indicators, the minimum amount of required information was retrieved from the answers to five questions provided by the users (Table 2) (4) blow moulding; (5) injection moulding; (6) stretch blow moulding; (7) glass production; (8) aluminium production; (9) paper/cardboard production, using recycled; and (10) paper/cardboard production, using virgin material. For the origin of the material, and the transport the user can also answer 'I don't know' and the least favourable option is considered by default in the following calculations as a precautionary principle.
After all five questions in Table 2 have been answered, the corresponding resource and carbon footprint values for 1 kg of material (in MJ ex /kg and kgCO 2 eq/kg, respectively) are retrieved from the database for each packaging component and multiplied by the mass of each component (Equation 1, Equation 2) to obtain Resource footprint and Carbon footprint for the packaging system considered.
To translate these footprints into a score (0-1), Equation (3) is used with a fitted p parameter equal to 1 for the Resource footprint and 0.1 for the Carbon footprint. The p-values were identified based on an ideal best-and worst-case scenarios for food packaging. Bestcase and worst-case packaging must have a score close to 1 and 0, respectively. Based on data extracted from Ecoinvent, the best-case packaging was considered to be a recycled cardboard tray (16 g) sealed using a cellulose-based lid film (0.55 g). The Resource footprint   (Table 3). These questions are answered once for the full packaging.
The item 'Food waste/food preservation' evaluates how well the packaging material meets the food preservation and storage needs of the food. This aspect is strongly dependent on the nature of the food (e.g., sugar powder, fresh meat and vegetables do not require the same protection during storage). To evaluate how the packaging meets the food requirements without overpackaging (too much functionality compared to the requirements of the food) or underpackaging (not enough protection for the food), a first question is related to the food itself-for example, 'Can the product dry out or loose crispiness during storage?'-and a mirror question enquires about the same aspect of the packaging a second time-for example, 'Does the packaging provide protection against humidity/dryness?' We proceed similarly for five different aspects of food preservation (Table 3). Then, depending on the answers to these two mirror questions, the 'Yes' or 'No' value is assigned and the final score for the pillar is computed according to the rule explained in Table 4.
To evaluate the sensitivity of the product toward oxidation and the ability of the packaging to protect it against oxidation, we proceeded slightly differently. Two questions were asked about the T A B L E 3 List of questions for the Functionality pillar. Is the packaging solution designed at the just necessary/optimal level? (are there any attempts using modelling tools or experimental approaches to optimise the pack, reduce the number of components, or decrease the packaging thickness to achieve the same food shelf life and standard conditions of use; does the packaging contains the right amount of product) answer must be given.
One last question related to packaging was asked in the 'Food waste/food preservation' section: 'Is the packaging solution designed at the necessary/optimal level?' The purpose of this question is to assign a bonus (an additional 'Yes' value) to packaging that has been optimised. For instance, the user could have identified the range of gas-and water-vapour barrier properties required by the product to use as targeted packaging specifications. This permits the user to select optimised packaging material, that is, packaging that matches the barrier properties required by the food.
Once the user has answered all questions listed in Table 3, the final score for the pillar is computed using Equation (4)   stage. This risk is usually not considered in quantitative environmental assessments because there is currently no quantitative model to predict it. 48 The advantage of a qualitative approach is that this risk can be included with other more easily observable aspects (e.g., 'Is the packaging part monolayer?').
All questions regarding the post-usage treatments must be answered for every element of the packaging that can be separated and has a different post-treatment method (Table 5). Once the user has answered all questions listed in Table 5, the final score for the T A B L E 5 List of questions for the Post-usage pillar. To develop a user-friendly tool, no weighting of criteria was applied.
No weighting between components means that a bottle cap counts the same as a bottle in the calculation of the final score, even if the mass of the cap is 10 times lower. Similarly, the post-treatment impact is weighted the same as the weight/size of the material.

| GLOPACK packaging score outputs
The pillar approach presented above provides four different scores between 0 (worst option) and 1 (best option): two scores for Materials Example of a possible output layout of the GLOPACK packaging score calculator for strawberry no. 1 (see Table 9). 4.5 | Assessment of the GLOPACK packaging score

| First case study
The tool was first used on commercial UHT milk containers (Table 6). (4) multilayer carton with an HDPE cap; and (5) multilayer carton without a cap. The scores were calculated for each pillar for each container. The original files can be downloaded at https://doi.org/10.

57745/JRQESS.
For the Materials pillar for each container, we do not know the origin of the raw material. Therefore, we chose 'World' for all cases.
The final packaging was also assumed to travel a long distance to the food company by cargo (water, transoceanic). However, regardless of the type of transport (even intercontinental air transport), the contribution of the transport to the overall score is very low (less than 0.5% of the resource and carbon footprints) compared to resources and manufacturing (second transformation), with contributions of approximately 63% and 35%, respectively. The origin of the resources 'Europe' and 'Rest of the World' also does not have a large impact on the resource and carbon footprints in the extraction step ( Figure 3).    post-usage treatment. Therefore, the tool is very useful for optimising the packaging design with respect to its environmental impact and enables the user to objectively and transparently select the most suitable packaging material for a given application.

| Third case study
As a commercial case study, we analysed packaging for caster sugar: a paper sachet; a cardboard box with a dispenser nozzle (re-closable); and a multilayer plastic sachet with a cap ('doypack'). The first two packages contain 1 kg of sugar, while the doypack contains only 750 g of sugar (it does not exist on the French market in a 1-kg size).
All individual score results are summarised in Table 11 and the final Borda rankings are given in Table 12.
As expected, paper and cardboard-based packaging (Sugar 1 and Sugar 2) obtained higher scores for the carbon footprint, functionality and post-usage aspects. In addition, they obtained a 'green' point for the long-term fate warning. The paper sachet also obtained the highest score for the resource footprint because of its low weight compared to the two other options. The Functionality score of the multilayer plastic sachet was the lowest because it was considered as 'over dimensioned' for sugar; the performance exceeds the requirements of the food as a good oxygen-barrier performance is not necessary for sugar. Over-dimensioning is highly detrimental to the environment because it often leads to the selection of materials that are usually not recyclable (multilayers), not biodegradable (plastic), and generally more expensive. The Borda methodology (Table 12) confirms the previous discussion based on individual scores; the paper-based sachet (Sugar 1) is the best option, followed by the cardboard box (Sugar 2) and plastic sachet (Sugar 3).

| Comparison with other methods
To compare the GLOPACK packaging score method with other initiatives in the field of sustainability scoring, the three case studies were also evaluated using the French Eco-Score method (Table 1).
This approach differs from the GLOPACK one in that the impact of the food product itself is considered through LCA data. The LCA output gives a score out of 100. Positive or negative impacts of other criteria not considered in the LCA constitute some bonus or malus that influence the final score. Hence, the packaging is considered a malus, accounting for a maximum of 15 points. In this malus, two indicators are considered. First, an upstream indicator that represents the origin of the resources: (1)  We compared the packaging malus from the French Eco-score method to the ranking of materials obtained from the GLOPACK packaging evaluation method. The packaging malus (Table 7)  The ranking based on the packaging malus is not the same than the one obtained with the GLOPACK method. The opaque PET bottle (Milk 1) is considered recyclable and does not disrupt recycling, so it was given a higher score by the French Eco-Score method than by the GLOPACK method, where opaque PET is considered recyclable but also a recycling disruptor in transparent PET recycling (stakeholder testimony). Cartons with or without Al are also not differentiated by the French method (malus of À5 in both cases). The HDPE bottle + lid + cap (Milk 2) is more strongly penalised than Milk 1, with a malus of À6 compared to À5, respectively, while with the GLOPACK method, Milk 2 performs better than Milk 1. The functionality of the packaging is not considered at all in the French method. Therefore, the multilayer carton without a cap (Milk 5) has a similar performance to that of the opaque PET bottle (Milk 1), based solely on the fact that opaque PET and multilayer carton are both recyclable. Different rankings were obtained using the GLOPACK and French Eco-score methods because they do not consider the same number and type of criteria.
The packaging malus for the three strawberry packages gave the following ranking: Strawberry 2 > Strawberry 3 > Strawberry 1, which is the same as that obtained using the GLOPACK packaging score method.
For the third case study, the French Eco-score malus gave: Sugar 1 = Sugar 2 > Sugar 3 (i.e., no difference between Sugar 1 and Sugar 2 alternatives), whereas the GLOPACK method differentiated between these options.
T A B L E 1 2 Aggregation results obtained from the Borda vote methodology and final ranking (number of Borda points) for the sugar case study.

Sugar case study
Resource footprint score Sugar 1 > Sugar 3 > Sugar 2 Carbon footprint score Sugar 1 > Sugar 2 > Sugar 3 Material score Sugar 1 > Sugar 2 = Sugar 3 Functionality score Sugar 2 > Sugar 1 > Sugar 3 Post-usage Sugar 1 = Sugar 2 > Sugar 3 Long term fate warning Sugar 1 = Sugar 2 > Sugar 3 Post-usage total score Sugar 1 = Sugar 2 > Sugar 3 Final ranking Sugar 1 (4) > Sugar 2 (3) > Sugar 3 (0) The ranking obtained using the French Eco-score packaging malus and the GLOPACK method are difficult to compare because the considered criteria are different. The comparison of the sugar and strawberry case studies shows that for very different types of material (paper/cardboard versus plastic) both approaches provide similar rankings. This is not the case for plastic packaging where information on the recyclability and use of recycled resources prevailed in the French Eco-score method, which distorted the final score. Overall, the GLOPACK method is considered more complete and provides a better overview of the sustainability of a packaging solution considering both its direct and indirect impacts (including functionality).

| CONCLUSIONS AND RECOMMENDATIONS
In a context where sustainability and environmental consciousness are becoming increasingly important for both consumers and producers of food and beverages, the development of a tool that can easily calculate the environmental impact of food packaging is crucial.
In this study, we presented a method that scores food packaging considering three key pillars of its lifecycle, (1) Materials, (2) Functionality, and (3) Post-Usage fate, and does not require the collection of extensive lifecycle inventory data.
The developed tool fulfilled all the initial specifications defined by the working group. The Materials pillar does not focus only on climate change, but considers both the carbon and resources footprints, which is novel compared to existing methods, such as the EEFP tool. 19 Our tool considers the functionality of the packaging material, which is still lacking in many scoring and eco-design tools (apart from EEFP), and the long-term fate in the environment and risk of micro/ nanoparticle generation that is especially relevant for plastic-based materials. The post-usage options are regionally dependent, and our tool provides different scores depending on the country. This tool is accessible to non-LCA experts and does not require the input of data and knowledge that is generally lacking in many food companies. The intended user is a packaging or food company that needs to select suitable packaging for a given food. Our tool can help the user improve the score of existing packaging by identifying the pillars that require further improvements. Although our tool cannot be used to directly design a new packaging format, it could be useful in the The GLOPACK packaging score is evolutive because it is scalable and can be easily upgraded with new criteria to consider new regulations and scientific breakthroughs in the domain of packaging science.
The database used for the background calculations of carbon and resource footprint scores could also be extended with new materials.
Other possible improvements of the tool could include the addition of weighting among the pillars/criteria to consider the most important features in the final ranking calculation. Some criteria may be more important than the others, for example, the urgency of overcoming plastic pollution, and including importance weightings is the most important direction for improving the tool in the future. Another direction for improvement could be extending the assessment to societal impacts to better evaluate the overall sustainability of the packaging.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are openly available in