We finally have the British standard for carbon footprinting of goods and services known as PAS 2050. It builds on existing life-cycle assessment methods (ISO 14040 series) to specify rigorous and consistent requirements for assessing life-cycle GHG emissions of goods and services. I'll highlight some of the more interesting points from the specification:
- Timing of emissions:
- Production and distribution emissions are modeled as a single release of emissions at the beginning of the 100-year assessment period.
- If the use and disposal phase emissions occur beyond the first year, a weighting factor is applied to model the average time the emissions are present in the atmosphere during the 100-year period. This is a significant improvement over the static, time-independent calculation of emissions.
- All emissions within the system boundary that can potentially contribute more than 1% of the total life-cycle GHG emissions.
- Carbon storage in products, such as biogenic carbon stored in wooden furniture, using a weighted average time of storage or uptake emissions.
- Non-CO2 GHG emissions, such as CH4 and N2O, from products containing biogenic carbon as well as from livestock and soils (the last two based on the highest tier approach set out in the IPCC Guidelines or employed in the country where emissions are produced).
- For agricultural crops, GHG emissions arising from land use change. This appears to be a stringent requirement, requiring worst-in-class data when the land use history is unknown.
- CO2 emissions from biogenic carbon sources, including when a product containing biogenic carbon degrades.
- GHG emissions from the production of capital goods, such as machinery, equipment and buildings. These input flows to the system are basically "cut off" and removed. No support for a hybrid LCA approach.
- Carbon sequestration in agricultural soils or carbon release from soils due to tilling practices, crop types, use of compost and other soil management. This is potentially a big omission given the amount of carbon stored in soils.
- Human or animal labor.
- Consumer transport to point of purchase.
- Employee transport to/from place of work.
- Primary activity data must be collected from processes owned/operated/controlled by the organization doing the carbon footprinting. If that organization contributes less than 10% of the upstream GHG emissions to the next downstream user, then the primary data requirements fall on the first upstream provider that does contribute 10% or more of the emissions. This can be a very difficult requirement for small companies and retailers who are not in a position to ask their big suppliers for internal operational data.
- For secondary data, PAS-compliant data gets preference as expected. There seems to be some provision for using secondary data when primary data is not available, but it is not clear if that would prevent a company from claiming compliance with the standard.
So, some coverage of dynamic carbon footprinting, no support for hybrid LCA, no consideration of carbon sequestration in or release from soils, tight system boundary (this enables a "streamlined LCA" approach), and strict data quality requirements. All in all, a good starting point, which I hope will evolve into a more complete specification. This may well become the default global standard for product/service carbon footprinting. The GHG Protocol has an independent effort focused on product life-cycle emissions and corporate scope 3 emissions -- it might be more productive for them to converge with PAS 2050 on the product life-cycle emissions.