The CarbonCloud methodology

Goal

CarbonCloud helps actors in the food and beverage industry assess the climate footprint of their products in accordance with ISO 14067 and GHG protocol Product Life Cycle Accounting and Reporting Standard.

Climate footprint of food

Rapid climate change caused by human activities such as food production has severe negative impact on the physical environment, the ecosystems and us as human species. However, climate change is by no means the only negative externality associated with food production. Food production is also the main driver for antibiotic resistance, animal welfare issues, unsustainable water extraction, eutrophication, biodiversity loss from pesticide usage, habitat destruction and overfishing. There are also important public health and worker-safety issues related to food production. This is not intended as a comprehensive list of food production related externalities.

The focus of CarbonCloud is solely climate change. This focus is chosen without any ranking of the importance of climate change relative to any other of the negative externalities associated with food production.

Attributional life cycle assessment

CarbonCloud uses the attributional (book-keeping) approach to life cycle accounting. This means that all significant activities and their emissions in the production are considered, and their combined climate impact is attributed to the product. The attributional approach only accounts for emissions and removals of greenhouse gases generated during a product’s life cycle and not avoided emissions or actions taken to mitigate released emissions. This contrasts with the consequential approach, which is used to assess the climate impact from changing the level of output of a product. The consequential approach focuses on marginal effects linked to the production of a product.

Stakeholders

Climate footprint studies can be conducted on the CarbonCloud platform by actors in the food and beverage industry who want to assess their products. These footprints are third-party reviewed by CarbonCloud. The target audience can be both internal and external stakeholders, such as end consumers of the product. The studies can be intended for both in-house and external applications. These include decision support to internal decision-makers at identification of climate impact hotspots, marketing, climate labeling, and comparison with other products on the CarbonCloud platform.

Scope

Unit of analysis

The unit of analysis is one kilogram (kg) of food product.

System boundaries

All food products modeled on the CarbonCloud platform are categorized according to their stage in the supply chain. The categories are farm gate product, factory gate product and store shelf product. The climate footprint includes all steps of the life cycle as applicable from the production of agricultural inputs, through agriculture or aquaculture, production of non-agricultural ingredients, fishing or hunting activities, transports, processing, packaging, storage and distribution up until the product reaches its current stage in the supply chain.

Mechanisms included

All mechanisms that are generally considered within the system boundary are listed in this section.

Agricultural mechanisms

CO2emissions from organic soils
Nitrous oxide (N2O) emissions from organic soils
CO2 emissions from deforestation
CO2 emissions from production of fertilizers
N2O emissions from production of fertilizers
N2O emissions from soil organic processes. Specifically, direct N2O emissions, indirect N2O emissions from volatilization of N and from leeching and runoff of N. These are caused by the application of both synthetic and organic fertilizers, and from nitrogen in crop residues left in the fields
CO2 emissions from application of lime
CO2 emissions from application of urea
CO2 emissions from pesticide production
CO2 emissions from use of farm equipment
CO2 emissions from drying of cereals, pulses and other crops typically dried at the farm
Methane (CH4) emissions from rice cultivation
Energy consumption associated with irrigation
N2O and CH4 emissions from manure management
CH4 emissions from enteric fermentation of ruminants
Emissions of CO2, N2O and CH4 from feed production and grazing

Mechanisms regarding packaging

Extraction of raw materials
Production of raw materials
Production of packaging from raw materials
Recycling of packaging
Transportation of packaging
Oxidation and release of fossil carbon stored in the material, from incineration or decomposition of packaging materials

Transportation and distribution mechanisms

Emissions from extraction, production, transportation and combustion of fuels
Fuel consumption for all transportation stages within the system boundary of the study, such as transportation/distribution:
- - from farms to food
  - between factories
  - to warehouses
  - distribution from factories or warehouses to markets
The following aspects of transport are considered:
- - distance
  - temperature controlled transportation
  - leakage of refrigerant for temperature controlled transportation
  - fuel consumption as a function of capacity utilization of the vehicles
  - empty returns of vehicles during distribution
The high-altitude climate effects of aviation

Food processing mechanisms

Direct emissions of fossil carbon or other greenhouse gases from ingredient reactions
Energy consumption for food processing
Food waste during production
Overhead operations (e.g. facility lighting, ventilation, air conditioning)
Leakage of refrigerants
Waste treatment

Mechanisms for wild aquatic products

Fuel consumption for fishing vessels
Fuel consumption for refrigeration
Leakage of refrigerants
Allocations of emissions considering edible portion of the products

Mechanisms excluded

Mechanisms explicitly excluded as out-of-scope:

Maintenance of farm equipment
Commute of personnel to and from the farms
Housing of personnel working at the farms
Albedo changes due to the production of crops
Transportation from store shelf to consumers
Energy consumption for preparation of food products by consumers
End-of-life treatment of products and packaging

Mechanisms excluded unless it is expected to have significant impact on the result of the study:

Manufacture of capital goods (e.g., machinery, trucks, infrastructure)

Mechanisms not considered:

Corporate activities and services (e.g., research and development, administrative functions, company sales and marketing)

Time period

Agricultural production data averages the most recent 5-year period. For other agricultural input data, the latest available is used. Process data represent the most recent full year, but may be proxied with older data when deemed reasonably representative and newer data cannot be obtained.

Carbon storage in products and packaging materials

Biogenic uptake of carbon stored in agricultural products is not considered since the carbon is released again upon digestion, decomposition or incineration. Delay of emissions is not taken into consideration due to the short time scales involved.

It is assumed that all carbon stored in packaging material (biogenic or fossil) eventually reaches the atmosphere. For biogenic carbon, the emissions cancel out the carbon uptake from the atmosphere when the raw material is grown. For fossil carbon, the emissions are included in the climate footprint. Delay of these emissions is not taken into consideration.

Climate footprint indicator

The impact category assessed is climate change. The metric used is Global Warming Potential over a 100-year time horizon (GWP100). The metric integrates the radiative forcing of events (such as greenhouse gas emissions) over a 100-year time period and gives values relative to those for the reference gas CO2. Thus, the climate footprint indicator is carbon dioxide equivalents (CO2e). The calculations include emissions to the atmosphere of carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O). Sulfur hexafluoride (SF6) is indirectly included in the emission factors for electricity. Perfluorocarbons (PFCs), and hydrofluorocarbons (HFCs) emissions are included in the climate impact of chilled transport, if applicable. The so-called high-altitude effect of aviation is also accounted for.

All greenhouse gases are weighted with the latest values of GWP100 including climate-carbon feedback effects given by the Intergovernmental Panel on Climate Change (IPCC; Myhre et al., 2013). For methane and nitrous oxide, a GWP100 of 27.2 and 273 kg CO2e/kg are used, respectively.

Allocation

When a process generates more than one product, the climate impact from the process needs to be allocated between the products. Allocation (instead of system expansion) is part of the traditional attributional approach to life cycle assessment. Allocation can be based on physical characteristics of the co-products such as mass or energy content or based on their relative market values (economic allocation). As a general principle on the CarbonCloud platform, economic allocation is applied. This means that the climate impact from a process is allocated between the products in proportion to their economic value.

In order to get the most correct calculation, each process should be modeled as accurately as possible. However, it can be difficult to measure the specific energy consumption of individual processes. It is often the case that data regarding energy consumption only is available on a factory level, with no way to tell different processes apart. In such cases, the climate impact from the factory’s energy consumption is allocated between all the products that goes out of the factory according to their economic value.

If the economic value of by-products is unknown, it is conservatively assumed that they have no economic value, so that the whole climate impact is allocated to the main product.

For recycling the “simple cut-off” approach is used which allocates the entire climate footprint of material production to the primary user of the material.

Land use and land use change (LULUC)

Climate impact of land use includes greenhouse gas emissions from cultivation, growth and harvesting, which is all considered in the assessment and attributed to the crop. All types of crop growth cause biological soil organic processes that emit nitrous oxides (N2O). Farming on drained wetlands (organic soil) such as swamps, also causes large amounts of carbon leakage in the form of carbon dioxide (CO2).

Land use change is when there is conversion of land use from one type into another, for example from forest to agricultural land. Land use change, largely in the form of deforestation, is responsible for significant amounts of carbon dioxide emissions each year. These emissions can be linked to crops that are grown close to the deforestation frontier.

Studies conducted on the CarbonCloud platform uses data from (Pendrill, 2020) to estimate emissions from deforestation for specific agricultural products. Pendrill et al. use remote sensing data to identify deforestation on a per country basis and allocate it among products through a land-balance model based on mainly FAOSTAT data for expansion of cropland, pastures and forest plantations. For Brazil and Indonesia, accounting for 40% of forest loss, a subnational analysis has been performed.

Modeling procedures

The modeling of climate footprints are conducted on the CarbonCloud platform. The following procedures are essential to achieving the goal of the study: screening, data collection and data quality improvement, and revision.

CarbonCloud online software

The CarbonCloud platform (online software) provides the framework for climate footprint assessments of products within the scope of the study. The software gives access to a comprehensive library of agricultural products from different countries and regions. The climate footprints of these products are calculated with the CarbonCloud agricultural model. The library also contains a wide range of representative ingredients, various types of packaging forms and packaging materials as well as different energy carriers. The software facilitates modeling of climate footprints specific to food products. Upon data entry to the model, the software carries out the climate footprint calculation considering all mechanisms listed in the scope of the study.

¹ The CarbonCloud agricultural model is a model of farm level emissions based on the IPCC guidelines (IPCC, 2019), complemented with estimates of emissions due to production of inputs (for a list of mechanisms included and excluded, see section Scope). When primary data from suppliers are not available, the input to this model is taken from a database of activity estimates constructed with representativeness as main goal. For example, estimates of fertilizer inputs are made by starting with FAO and IFA data on total fertilizer use within a country and/or crop group (FAOSTAT, 2021; Heffer, 2017) and attributing that to different crops in proportion to their nitrogen needs (modeled by using a mix of allometric data and specific factors such as nitrogen fixating mechanisms combined with production data).

Screening

The assessment starts with a screening process, performed on the platform. For screening, a draft model is created to assess the climate footprint of the product. The goal of the screening process is to identify components contributing significantly to the result. This is realized by mapping all ingredients, processing steps, transportation steps, packaging, storage and energy consumption within the scope of the study into the draft model. The CarbonCloud platform calculates the climate footprint of the product based on the data inputs to the draft model. The software then presents a result with breakdowns of emissions from different sections and visualizes the hotspots in the production processes.

The initial data inputs in the draft are from the following sources:

Primary data², if readily available
Data inputs through deploying ingredients, packaging and energy carriers provided in the CarbonCloud online software library
Remaining secondary data³ collected by CarbonCloud

² Primary data can be process activity data (physical measures of a process that results in GHG emissions or removals), direct monitoring, stoichiometry, mass balance, or similar methods) from a specific site, or data that is averaged across all sites that contain the specific process. Primary data can be measured or modeled, as long as the result is specific to the process in the product’s life cycle.

³ Secondary data can come from external sources (e.g., lifecycle databases, industry associations, etc.) or can be data from another process or activity in the supplier control that is used as a proxy for the process. This data can be adapted to the process or can be used “as-is”. Examples of secondary data include:

Average number of liters of fuel consumed by a process, obtained from a life cycle database
Kilowatt-hours consumed by another similar process used as a proxy in the studied product’s life cycle
Industry-average kilograms of material input into a process
Industry-average GHG emission from a process’s chemical reaction
Amount spent on process inputs, either specific to the process or a company/industry average

Data collection and data quality improvement

After the screening process, all data needs are identified in the draft model. The conductor of the study continues by collecting primary data for the processes under their ownership or control to replace the secondary data input in the draft model. The components of the life cycle with the largest contribution to the climate footprint are given priority. To ensure the quality of primary data, only most recent data that are geography and technology specific are used as inputs in the model.

When no primary data that is sufficiently representative of the given process in the product’s life cycle is available, the use of conservative secondary data collected by CarbonCloud is allowed for filling the data gaps.

Revision

Finally, CarbonCloud performs a revision on each version of the climate footprint study. Comments addressing possible improvements and concerns are communicated through the platform. Depending on the nature of comments, additional iterations of comments, recommendations and responses may be necessary. Providing all comments are addressed, the study is then approved for publication. All changes made in response to reviewer comments are documented. The revision ensures that the data used are appropriate and reasonable in relation to the goal of the study, and that the study report is transparent and consistent.

Life cycle inventory

Ingredients

All transports are calculated with the CarbonCloud online software. The model is largely inspired by Network for Transport Measures (NTM; Swahn, 2008). Parameters and model implementation are described in the table below.

Parameter	Model implementation	Comment
Transport mode	A good can be transported by road, rail (container or bulk), water (container or bulk) or air.
Vehicle type	Road: Semi-trailer (ICE or electric) or distribution truck (ICE or electric). Rail: Train (Electric or diesel). Water: Small ship (container or bulk), medium ship (container or bulk), large ship (container or bulk). Air: Cargo plane or Passenger plane (belly freight)	For air travel it is assumed that there is a correlation between distance and size of the aircraft. The user of the software may not choose size of aircraft. This is done automatically.
Capacity utilization	Each vehicle type has a default capacity utilization. The capacity utilization of vehicles for road transport can be modified by the user of the software.	For maximum capacity, both mass and volume are considered as limiting factors. In practice it is often the volume that is the limiting factor.
Fuel consumption	The fuel consumption depends on vehicle type and the weight of the cargo. This is determined by the capacity utilization and pallet density. Additional fuel consumption due to use of thermo equipment is taken into account. An average (transport mode specific) default factor times the general fuel consumption is used. There is no difference in scaling factor between frozen and refrigerated goods.
Empty returns	If there are known empty returns of vehicles in road transport, these are taken into account and their climate impact allocated to the transported good.
Transport distance	Transport distance is entered into the model by the software user and may be given either by the transport service provider or estimated by online services such as google maps (road transport), sea-distances.org (water transport) or NTMCalc Basic 4.0 (rail transport). \t
Allocation for goods transport based on load-limiting factor	A hypothetical transport is modeled, where one single type of good makes up the total cargo load. The total climate impact of the transport is therefore allocated to the good in question.
Fuel	Emission factors for each fuel is calculated with a life cycle perspective (well-to-wheel).
Global warming potential	Non-CO₂ effects of aviation is included in the assessment.	Aviation has impacts on climate change through both its CO₂ emissions and non-CO₂ effects.

Transport infrastructure and additional resources and tools such as cranes and transporters are not included unless if it significantly impacts the results of the study.

High-altitude effect of aviation

For high-altitude effect of aviation, we use the following characterization factors:

Continental and intercontinental flights:

GWP_HA = 0.9 * D * E_cruising * F_TTW

Regional flights:

GWP_HA = 0.4 * D * E_cruising * F_TTW

GWP_HA: Global warming potential of the high-altitude effects of a particular flight [kg CO₂e]

D: total distance of flight [km]

E_cruising: energy consumption while cruising [MJ/km]

F_TTW: CO₂ content of fossil jet fuel, tank to wheel [kg CO₂/MJ]

The high-altitude effects include emissions of nitrogen oxides (NO_x), aerosols and their precursors (soot and sulphate), and increased cloudiness in the form of persistent linear contrails and induced-cirrus cloudiness. The main source for this characterization factor is Lee et al (2010) who finds that the 100-year GWP of aviation corresponds to 1.9 times the CO₂ emissions from combustion of the fuel. This is an average number, and shorter flights have smaller climate impact from non-CO₂ effects, mainly since these flights do not reach so high altitudes. The lower multiplier for regional flights is taken from Kamb (2018).

There are still large scientific uncertainties about non-CO₂ effects of aviation. However, it is certain that they are not zero.

Processing

The climate footprint of each processing step is calculated through the refinement function in the software using energy and ingredient input for the specific process. The climate footprint of the process is allocated to the main product and by-products in proportion to their economic values (economic allocation).

Electricity emission factors

Emission factors for electricity are selected according to the following principles in hierarchical order

emission factors that represent specific electricity suppliers if there is a direct link to the supplier, for instance through a contract or agreement, and if there is supplier specific data available
emission factors that represent the residual mix of the country in question when there is no direct link to a specific supplier, and when there is data on the residual mix available
emission factors representing the average mix of the country (consumption based if possible, i.e., taking imports and exports into account) if none of the above applies

For electricity taken from the grid, only emission factors representing supplier-specific electricity is used if the supplier can guarantee through a contractual instrument that the electricity produced is assured with a unique claim and is tracked and redeemed, retired or cancelled by or on behalf of the reporting entity. As far as possible upstream emissions (occurring to extract and transport fuels) and power losses along the grid are considered. The emission factors are taken from G20 (2020), Moro & Lonza (2018), Association of Issuing Bodies (2019), IVA (2016), EDP of Electricity from Vattenfall’s Wind Farms (2019), EDP of Electricity from Vattenfall’s Nordic Hydropower (2021) and Parra (2020).

Fuel emission factors

The emission factors for solid, liquid and gaseous fuel consider a life cycle perspective, including production, shipping, refining, distribution and combustion. The emission factors are based on Skone (2015), Fransson et al. (2020), Bengtsson et al. (2011), Edwards et al. (2014), Energimyndigheten (2020), Prussi (2020), Bernstad Saraiva et al. (2017), Naughton (2016) and Eggleston et al. (2006).

Packaging

For packaging, emissions from the following are considered: packaging material (virgin or recycled), production/recycling process and the transport chain. It is assumed that all carbon stored in the packaging material (biogenic or fossil) eventually reaches the atmosphere and these emissions are included in the climate footprint. Delay of these emissions has not been taken into consideration. Apart from this, no other activity related to end-of-life treatment is considered. The scope is thereby consistently applied to both the product and its packaging.

For recycling, the cut-off approach is used. In this approach, the entire climate footprint of material production is allocated to the primary user of the material. Recycled materials bear only the impacts of the recycling process itself. The logic here is to be consistent with our philosophy of using economic allocation. Consumers are seldom compensated economically for leaving their used packages at the recycling station. In this sense, used packages have no market value. This means that it is assumed that only the primary use of packaging material is what drives the extraction of the raw material.

The following equations summarize the principle of accounting emissions for packaging materials:

Virgin packaging material:

Ev = Ecarbon + Eproduction + Etransportation

Recycled material:

E_{r</subr = E_recycling+ E_{transportation}}

E_v: climate footprint of virgin packaging [kg CO₂e/kg]
E_r: climate footprint of recycled packaging [kg CO2e/kg]
E_carbon: oxidation and release of carbon stored in the material [kg CO₂e/kg]
E_production: emissions from production of the packaging [kg CO₂e/kg]
E_recycling: emissions from recycling of the packaging from well to packaging factory gate [kg CO₂e/kg]
E_{transportation}: emissions from transportation of the packaging [kg CO₂e/kg]

Fossil resource

This category accounts for any fossil material that is used as an ingredient in the food product, or as a catalyst in the production process. Examples are petroleum derivates that are used to produce food colorings, vitamins, food-grade paraffin wax and other different types of food additives. The climate impact associated with this category represents the greenhouse gas emissions that result from the production and decomposition of these products.

Results

The climate footprint is given with unit kg CO₂e/kg. The footprint is also broken down into the following fundamental activities:

Agriculture
Fossil material extraction and use
Transports
Processing/storage
Packaging
Refinement
Uncategorised

References and data sources

Association of Issuing Bodies. 2019. European residual mixes 2018.

Bengtsson, S., Andersson, K., Fridell, E., 2011. A comparative life cycle assessment of marine fuels: liquefied natural gas and three other fossil fuels. Proceedings of the Institution of Mechanical Engineers, Part M: Journal of Engineering for the Maritime Environment 225, 97–110. https://doi.org/10.1177/1475090211402136

Bernstad Saraiva, A., Valle, R. a. B., Bosquê, A.E.S., Berglin, N., v Schenck, A., 2017. Provision of pulpwood and short rotation eucalyptus in Bahia, Brazil – Environmental impacts based on lifecycle assessment methodology. Biomass and bioenergy.

EDP of Electricity from Vattenfall’s Nordic Hydropower. Registration number: S-P 00088. 2021-05-05

EDP of Electricity from Vattenfall’s Wind Farms. Registration number: S-P-01435. 2019-01-31

Edwards, R., Larivé, J-F., Rickeard, D. & Weindorf, W. (2014). Well-to-Wheels Analysis of Future Automotive Fuels and Powertrains in the European Context. JRC Technical Reports, Report EUR 2637 EN. ISBN 978-92-79-33888-5

Eggleston, H.S., Buendia, L., Miwa, K., Ngara, T., Tanabe, K., 2006. 2006 IPCC Guidelines for National Greenhouse Gas Inventories. Volume 5. Chapter 4. Biological treatment of solid waste.

Energimyndigheten (2020). Drivmedel 2019. Redovisning av rapporterade uppgifter enligt drivmedelslagen, hållbarhetslagen och reduktionsplikten. ER 2020:26

Food and Agriculture Organization of the United Nations. “FAOSTAT Statistical Database.” Accessed February 6, 2021. http://www.fao.org/faostat/en/#data.

Fransson, N., Lundblad, M., Lätt, A., 2020. Emissionsfaktorer för bränslen till el- och värmeproduktion (No. B 2398). VL Svenska Miljöinstitutet.

G20 (2020). Climate transparency report 2020. Country profiles

Heffer, P., A. Gruère, and T. Roberts. “Assessment of Fertilizer Use by Crop at the Global Level 2014-15.” International Fertilizer Association & International Plant Nutrition Institute, 2017. https://api.ifastat.org/reports/download/12246.

IPCC. 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories. Edited by E. Calvo Buendia, K. Tanabe, A. Kranjc, J. Baasansuren, M. Fukuda, S. Ngarize, A. Osako, Y. Pyrozhenko, P. Shermanau, and S. Federici. Switzerland: IPCC, 2019. https://www.ipcc-nggip.iges.or.jp/public/2019rf/index.html.

IVA Kungliga ingenjörsvetenskapsakademien (2016). Framtidens el – så påverkas klimat och miljö. En delrapport

Kamb, A., Larsson, J., & Åkerman, J. (2018). Klimatpåverkan från svenska befolkningens flygresor 1990–2017. Chalmers, Göteborg.

Lee, D. S., Pitari, G., Grewe, V., Gierens, K., Penner, J. E., Petzold, A., … & Sausen, R. (2010). Transport impacts on atmosphere and climate: Aviation. Atmospheric environment, 44(37), 4678-4734.

Moro, A., & Lonza, L. (2018). Electricity carbon intensity in European Member States: Impacts on GHG emissions of electric vehicles. Transportation Research Part D: Transport and Environment, 64, 5-14.

Forster, Piers, Trude Storelvmo, Kyle Armour, William Collins, Jean-Luis Dufresne, David Frame, Daniel J. Lunt, et al. “The Earth’s Energy Budget, Climate Feedbacks, and Climate Sensitivity.” In Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change, edited by Valérie Masson-Delmotte, Panmao Zhai, Anna Pirani, Sarah L. Connors, C. Péan, Sophie Berger, Nada Caud, et al. Cambridge University Press, 2021. https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Chapter_07.pdf

Naughton, C.C., 2016. Modeling Food Security, Energy, and Climate and Cultural Impacts of a Process: the Case Study of Shea Butter in Sub-Saharan Africa. University of South Florida.

Parra, R. E. N. É. (2020). Contribution of Non-renewable Sources for Limiting the Electrical CO2 emission factor in Ecuador. WIT Trans. Ecol. Environ, 244, 65-77.

Pendrill, Florence, U. Martin Persson, and Thomas Kastner. “Deforestation Risk Embodied in Production and Consumption of Agricultural and Forestry Commodities 2005-2017.” Zenodo, November 9, 2020. https://doi.org/10.5281/zenodo.4250532.

Prussi, M., Yugo, M., De Prada, L., Padella, M. and Edwards, R., JEC Well-To-Wheels report v5, EUR 30284 EN, Publications Office of the European Union, Luxembourg, 2020, ISBN 978-92-76-20109-0, doi:10.2760/100379, JRC121213.

Skone, T.J., 2015. Life Cycle Greenhouse Gas Emissions: Natural Gas and Power Production.

Swahn, M. (2008). DRAFT Additional CO2e-factors in goods transport. Network for Transport Measures. https://www.transportmeasures.org/wp-content/uploads/2015/11/Additional-energy-use-in-transport-sector-20080611-rev-20102411.pdf