Uncertainties in the carbon footprint of refined wheat products: a case study on Swedish pasta

Abstract

Purpose

Calculating the carbon footprint (CF) of food is becoming increasingly important in climate change communication. To design effective CF labelling systems or reduction measures, it is necessary to understand the accuracy of the calculated CF values. This study quantified the uncertainty in the CF of wheat and of a common refined wheat-based product, pasta, for different resolutions of farm-level in-data to gain an increased understanding of the origins and magnitude of uncertainties in food CFs.

Methods

A ‘cradle-to-retail’ CF study was performed on Swedish pasta and wheat cultivated in the region of Skåne on mineral soils. The uncertainty was quantified, using Monte Carlo simulation, for wheat from individual farms and for the mixture of wheat used for pasta production during a year, as well as for the pasta production process.

Results and discussion

The mean pasta CF was 0.50 kg CO₂e/kg pasta (0.31 kg CO₂e/kg wheat before the milling process). The CF of wheat from one farm could not be determined more accurately than being in the range 0.22–0.56 kg CO₂e/kg wheat, even though all farm-level primary data were collected. The wheat mixture CF varied much less, approximately ±10–20% from the mean (95% certainty) for different years. Reducing farm-level data collection to only the most influential parameters—yield, amount of N and regional soil conditions—increased the uncertainty range by between 6% and 19% for different years for the wheat mixture. The dominant uncertainty was in N₂O emissions from soil, which was also the process that contributed most to the CF.

Conclusions

The variation in the wheat mix CF uncertainty range was greater between years, due to different numbers of farms being included for the different years, than between collecting all farm-level primary data or only the most influential parameters. More precise methods for assessing soil N₂O emissions are needed to decrease the uncertainty significantly.

Recommendations

Due to the difficulties in calculating accurate values, finding other ways of differentiating between producers than calculating numerical CFs might be more fruitful and fair. When legislation requires numerical CF values, CF practitioners have little option but to continue using existing methods and data collection strategies. However, they can provide input on improvement, contribute to standardisation processes and help raise awareness and knowledge of the associated uncertainty in the data through studies like this one.