Amadeus' low carbon recipe analysis with Foodsteps

Foodsteps has produced the below methodology which assesses the impacts of all food sold at the NEC. The report shows how “Low Carbon” labels were created for food items having a ‘Very Low’ or ‘Low’ carbon footprint.

This work was performed in collaboration with a study conducted by the West Midlands Combined Authority and the University of Birmingham on food sustainability labelling and behaviour change. Foodsteps received data on approximately 3,800 recipes from Amadeus, catering services provider for the NEC venue, with which they used a modified methodology to perform a full life-cycle assessment of the recipe’s carbon footprints. Assessed recipes receiving an ‘A’ or ‘B’ Foodsteps rating were given a custom “low carbon” food label alongside short explanatory text on the menus explaining to customers the meaning behind this label.

The Foodsteps Rating

The five impact ratings, ranging from A to E, are assigned according to the carbon footprint per kilogram of each food item - otherwise known as the carbon intensity. They indicate a carbon intensity score from A (Very Low) to E (Very High).

The A-E ratings are allocated according to a food carbon budget of 5 gigatonnes of COe per year for the global population. This global scientific target has been set by the EAT-Lancet Commission to achieve healthy diets and sustainable food production, and ultimately help to meet the Paris Agreement pledge of keeping global warming to well below 2°C, aiming for 1.5°C.

Foodsteps uses this budget to establish an individual daily carbon intensity allowance, and therefore measure a food item’s level of impact. A-rated foods have a ‘Very Low’ carbon footprint, and align to Foodsteps’ individual daily allowance for carbon intensity. B-E rated foods fall outside of this allowance, with each moving further and further from helping to meet the Paris Agreement pledge.

Image for Amadeus' low carbon recipe analysis with Foodsteps
An example of the carbon labelling at the NEC's Edge restaurant

Why are the carbon footprint rating and carbon footprint value measured differently?

Foodsteps has chosen to include both the carbon footprint per kilogram and carbon footprint per serving on their carbon labels, as they provide equally important information about a food item’s impact. The carbon footprint per serving demonstrates the food item’s actual carbon footprint, taking into account its portion size. The carbon footprint per kilogram, reflected in the Foodsteps A-E rating system, enables consumers to compare between products which may be served in different portion sizes. This mirrors the reporting of nutritional information on food packaging.

How was this modified for the NEC?

Labels provided by Foodsteps for food items sold at the NEC display a simple “Low Carbon” message rather than the A to E impact rating. The “Low Carbon” label corresponds to foods receiving either an ‘A’ or ‘B’ impact rating and assessed as having a “Low” or “Very Low” carbon footprint.

We opted for simplified carbon labels here because:

  • The project required a singular icon, legible in context at a high footfall event on digital screens, conforming to all accessibility requirements.
  • The data available didn't always include full ingredient breakdowns, so we excluded on a precautionary basis, and filled gaps with similar ingredient information from our database, meaning a less detailed indicator is more appropriate at dish level.


Unit of Analysis

The unit of measurement for the carbon footprint is kilograms of carbon dioxide equivalent (kg CO2e). A 100-year global warming potential (GWP) factor was applied to GHG emissions and removals data to calculate the inventory results in units of CO2e.

The functional unit of the assessment is a single serving of each recipe, as served to the customer upon purchase. For meals with multiple choices of sides and add-ons, each recipe component was assessed individually. Labels therefore correspond only to each food in isolation (i.e. a burger or a side of fries) unless the recipe is specifically sold as a fixed combination (i.e. fish and chips).


Though Foodsteps received roughly 3,800 recipes for assessment, many of these recipes fell outside of scope for analysis. Due to ingredient data gaps, all third-party branded food products and recipes, such as Coca-Cola products and Starbucks branded meals, were excluded from assessment. This analysis also focused exclusively on food items and omitted any beverage products, due to both third-party conflicts and limitations on data availability.

After these initial exclusions were made, Foodsteps chose to omit from assessment recipes containing ruminant meats, such as beef and lamb, and coffee products. This is due to the extremely high carbon intensity of these ingredients, and thus corresponding low frequency with which recipes containing these ingredients achieve ‘A’ or ‘B’ impact ratings. Of all recipes in the Foodsteps database containing beef or lamb, 0% and 3%, respectively, achieve these ratings. All of the roughly 1,800 remaining recipes were assessed in-line with the methodology detailed below.

Inventory Type

The Foodsteps carbon footprint assessment uses a cradle-to-grave inventory. This covers the carbon footprint from farming inputs through to product decomposition by various disposal methods. See a more detailed breakdown of the different stages below (Table 1).

Table 1: Life Cycle Stage Breakdown

Data Sources

In this project, Foodsteps used recipe ingredient and quantity data provided by Amadeus for all food products sold between 28th July and 8th August 2022 to perform recipe carbon footprinting. Recipe ingredients were matched to the Foodsteps database of lifecycle emissions factors. Due to the limitations on the data provided and timeline of the project, emissions factors for the varying lifecycle stages were calculated using the Foodsteps Database. A description of the specific assumptions made for different stages is included below:

Farming: As ingredient origin information was not provided, ingredients are assumed to have average UK sourcing.

Packaging: As product-specific packaging information was not provided, food item packaging is based on averages from the Foodsteps database. This includes primary (final), secondary (shelf) and tertiary (distribution) packaging where appropriate.

Cooking: The carbon footprint of cooking is based on instructions provided by Amadeus, where applicable, and assumptions made regarding the type and preparation of recipes, where instructions were not provided. The impacts of cooking were only applied to recipes where it was not improbable that the addition of cooking impacts could change their carbon intensity to an extent where they would no longer be rated as "Low Carbon". These recipes were all of those for which carbon intensity, prior to the application of cooking impacts, were above 84% of the upper bound of the "Low Carbon" rating.

Foodsteps Secondary Data

Farming to End-of-Life: Secondary data derives from existing carbon footprint values in the Foodsteps database. The Foodsteps database consolidates a large number of peer-reviewed academic Life Cycle Assessment (LCA) studies. These studies span more than 100 different countries, and Foodsteps uses international trade data to map the average UK sourcing of ingredients.

Gap Filling

Where an exact ingredient match does not exist in the Foodsteps database (or in the wider academic literature), we aim to find the closest proxy item. Ingredients with a comparable production system can be used to estimate the carbon footprint. Where appropriate, additional penalties for packaging and/or processing of ingredients are also applied.

Assurance: The outputs from this project - a “low carbon” collection of food items - were produced by first party review only.

Standard: ISO14040

Methodology: Aligned to the GHG Protocol Life Cycle Accounting and Reporting Standard.

Limitations of Use (in line with The Green Claims Code):

  • Environmental claims can only be made for meals included in the scope of the assessment.
  • Carbon footprint information (such as the Foodsteps carbon labels) cannot be publicly displayed without accompanying information to verify the claims being made. This could include a QR-code link to methodological information provided by Foodsteps, for example.
  • Meals should not be described as ‘eco’ or ‘environmentally-friendly’, or other similar terms. This is only appropriate where products have an overall beneficial impact on the environment.
  • Only A-rated (‘Very Low’) meals can be claimed to align with the Paris Agreement pledge. B-rated (‘Low’) meals do not fall within the EAT-Lancet food carbon budget, and therefore cannot be claimed to align with the Paris Agreement pledge.
  • Comparisons cannot be made to items produced outside of the UK, or competitor products assessed using a different LCA inventory.