Supply Chain Frontiers issue #35. Read all articles in this issue.
Calculating the size of an organization’s carbon footprint can be extremely challenging when complex supply chain networks are involved. To simplify the task, many companies use what is called an initial screening — an approximation of the emission levels generated by each component of a network. However, as an analysis carried out by Dr. Edgar E. Blanco and Anthony J. Craig at the MIT CTL shows, the simplified model of a carbon footprint may be misleading.
Blanco and Craig compared the carbon footprints of five global logistics networks by using the screening method and a more detailed model provided by the third-party logistics services provider (3PL) Damco via its proprietary tool, SupplyChainCarbonCheck™.
There were four main sources of emissions in the global logistics networks analyzed: in-country transportation, international transportation, warehouse operations, and port operations.
For the initial screening, publicly available information was used when available:
- Location information for ports, warehouses, and distribution centers was calculated.
- All domestic road distances between locations were computed using Google® Maps for road distances.
- Ocean distances were calculated directly from port to port using dataloy.com.
- GHG Protocol-recommended emission factors were used for road and ocean transportation of 72 g CO2 per tonne-km and 10 g CO2 per tonne-km were used.
- Port emissions were based on data reported by the Port of Seattle study of 2.56 g of CO2 per kg, assuming a capacity of 10,000 kg per TEU (Twenty-foot Equivalent Units).
For the selected global networks, the 3PL supplied details about container movements, including carrier information, ocean routes, ports of call, and truck road distances. In addition, Damco provided detailed information on the fuel consumption rates for certain ocean routes by vessel type.
Information on port and warehousing emission factors was based on historical data on the 3PL’s global operations. All the emission factor information was embedded in Damco’s proprietary tool, SupplyChain CarbonCheck. The MIT CTL research team verified the validity and accuracy of emission calculations as well as the overall use of emission factors.
Total emissions under the screening approach were 27% lower than the calculations that used the higher resolution information from the 3PL model. The figures for individual products also varied greatly, with the screening model numbers ranging from 57% lower to 58% higher than the 3PL model. The three main sources of discrepancies were:
- Ocean route distance. Owing to stops at intermediate ports between origins and destinations, ocean distances are on average 10% higher than direct port-to-port distance, and on some strings may be as much as 21% higher.
- Vessel-specific numbers. The public value for the environmental performance of ocean shipping assumes a single emission factor, but different vessels and voyage strings may have different emissions characteristics. These differences may vary by more than 30% from the average.
- Emission business drivers. In the maritime industry, operational decisions are based on TEUs, a volumetric measure. However, the screening level model uses weight as the main driver of the calculations.
The research team also found variations in emission levels calculated for the shipment of specific products, although there was no consistent pattern. For one product, the value derived from the Damco model was greater than the value calculated under the screening method; but for another product, the reverse was the case.
The analysis showed significant differences in total emission calculations when using a 3PL’s detailed model as opposed to the less granular screening methodology. While the screening calculation identified the major emission sources, it significantly underestimated logistics emission levels. Thus, in order to use the carbon footprint to drive operational decisions, there is significant value in using validated carbon calculations from logistics partners. The Damco model also offers an accurate calculation methodology that is automated through a standardized tool.
There are some riders, however. The more precise 3PL approach requires the construction of screening models to use as a reference to understand the differences. Also, 3PLs do not always have higher visibility to all logistics functions. In the case of Damco, warehousing and port operational data wwere biased toward the 3PL’s own operations, and did not necessarily provide higher resolution data, but rather a different set of assumptions.
This article is based on a paper entitled, “The Value of Detailed Logistics Information in Carbon Footprints,” by Dr. Edgar E. Blanco and Anthony J. Craig of MIT CTL. For a copy of the paper and more information on the Center’s carbon footprint research, contact Dr. Edgar Blanco, eblanco@mit.edu .