I would recommend looking at the Gibraltar City GHG inventory report - we produce this on an annual basis and shipping emissions is a HUGE consideration there, as a major international port. https://www.gibraltar.gov.gi/new/sites/default/files/HMGoG_Documents/20170601-Gibraltar_City_Inventory_Report_Published.pdf
In a nutshell, we obtained fuel sales data but it was patchy and didn't separate fuel sold as bunkers (i.e. to international ships just passing though - there is no tax so many many ships fill up but don't really stop there) to those calling at the port for a purpose, like ferries, cruises or freight.
We obtained a similar databset to yours from the port authority, on class of vessel, tonnage, purpose of call, origin and destination (from the ship manifests) for thousands of ships.
We split the ships by purpose of call so we knew which were outside of scope and which we wanted to consider.
For all ships, we then estimated distances between Gibraltar and origin/destination (only departing journeys were ultimately reported, consistent with the GPC). This was quite time consuming as it was manual, although many ships follow the same routes (we used http://ports.com/sea-route to estimate the distance, note it's in nautical miles soconvert to km).
We then used the details of ship type and tonnage, combined with these calculated distances, to estimate fuel consumption for each ship. There are factors in the European Environment Agency Emission Inventory Guidebook for this, reproduced in the above report and available at EMEP/EEA air pollutant emission inventory guidebook — European Environment Agency.
From that we could generate implied fuel use, which can then be multiplied by an emission factor for the fuel type.
We used the fuel sales as a verification check, but it was so patchy and inconsistent that it didn't really verify much...
The Bunkering ships and ships not calling for a local purposes (just for fuel) were reported in 'other scope 3'.
We could have then assigned only a proportion of the emissions from the ships to Gibraltar, based on % of total tonnage of freight, or % of passengers, but because we had already separately reported the non-Gibraltarian ships to other scope 3, we decided the remaining really were calling at Gibraltar primarily for local purposes so reported these 100%.
Hope that makes sense and gives you some ideas on how to proceed. Ultimately, shipping (excluding unless local waterborne e.g. ferries on rivers in the city) is mostly Scope 3 so there is quite a bit of flexibility in how to report. I think as long as the approach to allocate emissions to the city seems sensible and is transparent, and the methodology follows approaches reported in e.g. the EEA/EMEP guidebook or IPCC, then do what makes sense for the city.
Hi Rose, thank you so much for your response. Fong from WRI made me aware of the GHG inventory for Gibraltar, good work! We are currently considering following your approach to calculate the fuel use and compare this to the emissions we have estimated so far. In our earlier approach we converted the tkm distance directly to emissions, but this seems to be too large. Will be interesting to see what results we get following your approach.
Our main issue at the moment is how to apportion the emissions. The Port of Tauranga services a much larger area than the city. Unfortunately the port has no statistics on the origin of their exported products; all we have is a split by tonnage of exported products (i.e. logs, other forest products, dairy products, kiwi fruit & other goods). Have you worked with any city where you had to allocate emissions from their ports or airports to a wider area?
If we were to allocate all of the shipping related emissions to the city of Tauranga, the emissions would likely be larger than any of the other emission sources combined (i.e. it’s a massive port for a small city). If you had any further advice, that would be very much appreciated.
Many thanks and kind regards,