2.1 Lifecycle assessment (LCA): quantifies environmental impacts from cradle to grave of a product

2.1.1 Overview of LCA aims and process

LCA studies, following the methodology framework set by ISO14040, involve 4 phases:

  • goal/scope definition;
  • inventory analysis;
  • impact assessment; and
  • interpretation.

Each phase is described in more detail in subsequent slides, but an overview follows here.

LCA: a structured way of quantifying the environmental impact and resource use

Adapted from Hellweg & Milà i Canals (2014)

In Phase 1, the functional unit (the product, service or system upon which the LCA is based) is defined. This could, for example, be 1kg of pork. The system boundaries are then decided; these define what is measured and what is not, and may start with the production of the raw materials required on a farm and end with purchase of 1kg of pork at a supermarket.

In Phase 2, an inventory is made of all the inputs and outputs (e.g. different types of resource use and emissions) accumulated in order to produce one functional unit's worth of product, at each stage within the specified system boundaries.

In Phase 3, emissions and resource use are grouped together into specific environmental themes or impact categories, such as climate change, eutrophication, and acidification. These are referred to as mid-point indicators. Further modelling of how these impacts will affect human health, ecosystem quality and natural resource use (end-point indicators) can also be performed. Mid- and end-point indicators are discussed later in this chapter.

In Phase 4, sensitivity and uncertainty analysis are performed to determine the robustness of the results to changes in assumptions and variables used in the calculations. Conclusions and recommendations are made. An example might be the recommendation of one manufacturing technique over another, or the identification of life cycle hotspots (locations in a product's lifecycle where the most impacts occur) for the functional unit being studied.

One challenge within LCA, if the purpose is to compare different foods, is whether different studies use the same system boundaries, measure inputs and outputs in a consistent manner, and in how these results are interpreted. Increasing standardisation of methodologies, such as PAS 2050 (designed for carbon footprint analyses), is designed to reduce inconsistencies and allow comparison between research. This is important, especially within the food system, because of the many different production techniques and variables involved (see later in this chapter for more on this).

2.1.2 Phase 1: goal and scope definition

Phase 1: Goal and scope definition

Adapted from Hellweg & Milà i Canals (2014)

When measuring the environmental impact of a product, service or system, we must initially define the scope of the assessment. Life cycle assessment ideally considers all stages involved in the production of a product or service (sometimes termed “from cradle to grave”, or “field to fork”).

For example, the LCA scope for a loaf of bread could include producing raw materials for farming, production on the farm, harvesting and processing the crop, manufacturing the bread, delivery and storage at retail, impacts from consumption (including toasting), and transport and waste impacts all along the product's life cycle.

Some LCA studies deliberately limit their scope in order to focus on the impacts of a particular stage. An example might be an assessment of different production techniques for certain crops, in order to understand how different farming techniques might affect environmental outcomes.

In practice, as supply chains are often very complex, some inputs that are known to have a very small contribution to the overall environmental impact are sometimes ignored, and it is often a matter of judgement where these boundaries are set.

We also need to define what output it is whose impacts we are measuring – e.g. one kg of product, or 100 calories of product, or 1ha of farmed land and so forth. This is called the functional unit. The impacts – e.g. the quantity of carbon dioxide equivalent CO2 eq. emissions emitted in delivering this functional unit are expressed as a ratio – e.g. 0.5kg CO2 eq./kg product. The output can also be converted and expressed as CO2 eq. per average or recommended portion.

Choice of functional unit can influence results

The functional unit is the unit that is being studied. It can be an individual product (e.g. 1kg of pork or the packaging of 1 litre of soda) or a service – e.g. (to take a non-food example) delivery of a week’s supply of washed cloth nappies.

LCAs for food products normally use weight of finished food products as the functional unit (e.g. 1 kg of pork), but also use units of area (for example, the environmental impact of wheat production in terms of hectares used) or nutritional units (for example, the environmental impact of producing 100g of protein from a food product). The choice of functional unit could influence comparison between foods. For example, egg usually has a lower carbon footprint than poultry meat if the functional unit is 1 kg of product (egg or poultry meat) but if the functional unit is 1 kg of protein instead, the carbon footprint is more similar between the two as eggs only contain 12% of protein compared to 20% in meat.

The choice of the functional unit should reflect the purpose of the LCA-study. For example, if the purpose is to compare the production of pork using different animal feed inputs, a relevant functional unit could be 1 kg of pork meat at the farm gate. For another study comparing different ways of securing protein in our diet (e.g. via meat, fish, dairy, legumes etc.), a functional unit of 1 kg of protein might be better. However, as food purchases are seldom made on the basis of the nutrient content of food, a functional unit of 1 kg is often very relevant.

The relative impacts of different products can change, depending on which functional unit is used.

2.1.3 Phase 2: assessing inputs and outputs

Phase 2: Inputs and outputs for all stages are included

Adapted from Hellweg & Milà i Canals (2014)

Data are collected for inputs and outputs at each stage within the scope of the LCA. Inputs might be energy, land and water use; outputs are the products and waste leaving the system at each stage.

Emissions are estimated for each of these, for example CO₂ emissions for the energy use to produce the chosen functional unit. The data used for this purpose may be source data specific to the focus of the LCA, such as the measured energy used in a factory process. Alternatively the data may be more generic and drawn from existing databases, which contain the average impacts arising certain types of common process. Clearly, the more specific the data the more accurate the results. Whatever the source of the data, the data need to be scaled to give the amount per functional unit that is the focus of the LCA.

Production systems often:

  • Produce more than one product
  • Contain the use of recycled material

Allocating emissions and inputs between products is important but can be difficult

If a factory manufactures three different products, one of which is the functional unit of an LCA study, then the energy use in that factory needs to be allocated appropriately to the functional unit being studied (maybe 1/3 of energy use), rather than the total energy used by the factory as a whole. Sometimes allocation is a complex process. For example, a dairy system produces not just milk but also meat (male calves, dairy cows no longer lactating). Some of the overall herd emissions need to be allocated to this meat production – but on what basis? In terms of their relative economic value? In terms of the calories or protein they supply as compared with the protein or calories supplied by the milk? Different studies adopt different approaches.

All of the data included contributes to the final LCA results. Emissions and resource use at all stages that lie within the defined scope of the LCA need to be included because they all contribute to the environmental impact of the product or service in question.

How do we deal with the problem of allocating emissions to products?

  1. Avoid allocation altogether
    If possible it is preferable to avoid having to divide emissions between co-products by, for example, expanding the system boundaries to include all products produced in the system – for example, by including both the milk and the meat produced in a dairy system in the functional unit. However, if the purpose of the study is to compare milk with some other beverage, emissions needs to be split between the milk and the meat.
  2. Substitution
    One way of handling this is to consider alternatives to the co-products on the global market based on the reasoning that when a co-product of a production system enters the market the production of some other product with the same function can be ‘avoided’. For example, the beef produced in the dairy system would avoid production of other beef meat e.g. from a suckler cow system or meat from pork or poultry. The emissions from the production of milk is then calculated as the total emissions from the dairy system minus the emissions from the production of the alternative ‘avoided’ product, in this case the beef.
  3. Allocate based on physical relationships (e.g. mass or energy) or economic relationships
    Another simple way of doing this is to allocate by mass or economic value. For example, if an economic allocation method is used, then the relative economic value of the co-products is used as the basis for allocating impacts – e.g. the relative value on the market of milk and meat. The precise impacts will vary as the market prices fluctuate.

2.1.4 Phase 3: assessment of impacts and allocation to indicators

Phase 3: Impacts are assessed and allocated to indicators

Adapted from Hellweg & Milà i Canals (2014)

The total emissions and resource use from all life cycle stages are grouped into environmental themes or categories (mid-point indicators) according to the impact they cause, and are converted to common impact units. Examples of impact categories are climate change, eutrophication potential, water use and land use. Effects that these impacts can have on humans and ecosystems may also be modelled. Results are then presented using what are called end-point indicators; these refer to broader issues of concern such as human health, biodiversity/ecosystem services. The end-point indicators are fewer in number than the mid-point indicators and therefore might be easier to interpret. However, the modelling of end-point effects is associated with great uncertainties.

The results of an LCA can be shown for each different impact category, or presented in relation to particular stages in the product’s life. Thus, the water use for producing 1kg of bread can be shown for the whole life cycle and for each stage within that (such as farming, manufacture, etc.).

LCA does not just measure greenhouse gases

LCA can be used to measure many different environmental impacts, just one of which is the global warming potential or greenhouse gas emissions (GHG). Other impacts include:

Holistic assessment  avoids burden shifting,  whereby improving one  metric might adversely  affect another.

Measuring more than one factor in this way means we are able to see wider environmental impacts, and can avoid unintended negative consequences from reductions in one specific impact (sometimes called “burden shifting”).

For example, if an intervention is designed to reduce global warming impact, but as a consequence this increases land-use requirements, then it is important to know this.

These impacts can be further allocated to general impact categories (related to actual physical impacts) concerning human health, ecosystem quality, climate change and resource use.

Environmental impacts can be presented as Mid- & end-point indicators

Mid-point indicators represent impacts as specific environmental themes.

End-point indicators are categories of impact contributed to by (often multiple) mid-point indicators, representing impacts on broader issues of concern such as human health and ecosystem quality. For example, depletion of the ozone layer and increases in photochemical smog (presented as mid-point indicators) can both contribute negatively to human health; high land-use and water requirements place a burden on resources.

Modelling end-point indicators thus allows LCA studies to communicate the real impacts on the environment and people from producing a product or service, in a less abstract way. However, not all LCA studies specifically take this extra step due to the difficulties in modelling end-point effects and the increased uncertainties associated with this step.

2.1.5 Phase 4: Interpretation of results

Phase 4: The final phase is the interpretation of the results

Adapted from Hellweg & Milà i Canals (2014)

This can include drawing conclusions about the environmental impact of the product being studied, and identifying hotspots both in relation to life cycle stages (such as the stages that have the highest environmental impacts), impact categorisations (which environmental impacts are most of concern as a result of producing this product), or both (which stages contribute the most to the key environmental impacts).

Results are also analysed and tested to see how robust they are using sensitivity analysis and/or uncertainty analysis (i.e. use of different input data, comparison with model data, choice of system boundaries, functional units, allocation principles, etc.).

Environmental hotspots can be defined (i.e. life cycle stages with high impact) and conclusions are drawn about the overall impact per functional unit being assessed.

Recommendations can also be made to reduce environmental impacts, based on the hotspots identified, and may be placed in the context of wider environmental challenges. For example, the greenhouse gas emissions of a product can be put in the context of wider climate change concerns.

LCA allows us to identify environmental hotspots

Source: Basset-Mens & van der Werf (2005). Values are per kg of pig produced, expressed as a percentage of the GAP scenario.


This example illustrates the relative impacts on eutrophication, climate change, acidification, terrestrial ecotoxicity, energy use, land use and pesticide use from different pig farming systems. It is clear that the production of feed is the hotspot in all impact categories except for acidification.