Carbon emissions methodology

The methodology behind our Carbon emissions dataset available in Market benchmarks is based on the data provided by our partners at Marine Benchmark, an independent maritime data analytics firm.

Index calculations

The process of calculating the data behind the emissions index consists of two steps.

Calculating CO2 per ton

First, we analyze several objective parameters to get the actual CO2 per ton data for every vessel's journey:

• Vessel specifications
• Stakeholder information
• Journey data
  • Voyage legs
• Modeled data
  • Cargo load
  • Fuel consumption

Building the index

Next, we take the resulting data and use it to build the index by:

• Mapping journey data
• Detecting outliers
• Modeling trade lane and carrier data

Vessel specifications

We have access to the S&P Global/IHS Markit ship register supported by the UN, which contains data on hundreds of vessels.

When calculating emissions data, we primarily use the following vessel specifications:

• Size
• Age
• Engine type (to calculate the engine consumption profile)
• Fuel type (to calculate the fuel consumption)
• Hull shape (to calculate the amount of power needed to move the ship)

Stakeholder information

The S&P Global/IHS Markit ship register also provides information on the owner and operator of each individual vessel, tracking any changes that occur over time.

📘

While the register only guarantees tracking for charters that last six months or longer, charters under six months are extremely rare and represent only a small fraction of all charters that occur.

Journey data

Journey data is based on the information from AIS (Automatic Identification System), a tracking system where vessels' positioning data is automatically captured and transmitted through a network of antennas and satellites.

The journey data transmitted through AIS includes:

• Vessel type
• Identification
• Length and beam (width)
• Current position
• Speed and direction
• Draft (how deep it sits in the water)
• Destination and route plan
• Total time steaming, manoeuvring and at anchor
• Number of port calls
• Date and time of departure and arrival

Voyage legs

We have mapped out around 80,000 areas (also known as polygons), which combined cover almost all ports, terminals, and berths. When a vessel enters one of these areas (as indicated by AIS data), it is registered as having arrived in a specific location.

A voyage leg is calculated from when a vessel leaves a berth to when it leaves the next berth and includes:

• Exit from port
• Sea passage
• Any activities during sea passage
• Waiting time and port time at destination port

Sequential voyages

Vessels rarely call at only two ports on their journey, so voyage legs are combined to make one sequential voyage between given port pairs. If a vessel sails from Shanghai to Hamburg, it may also call at Singapore, Jeddah, Barcelona and then Hamburg.

The CO2 data for the ship's journey from Shanghai to Hamburg would then be the sum of these separate voyage legs, with stops, waiting time, port time, and fuel consumption all taken into account.

Modeled data

Using AIS data in combination with the vessel specifications and stakeholder information allows us to model actual carbon emissions for any specific ship.

Cargo load

We model each vessel’s cargo load in tons based on the AIS draft data, combined with specific vessel parameters and TPC (tons per centimeter).

With the load data available, we can then calculate the ship's filling factor necessary to determine its CO2 emissions.

Fuel consumption

We calculate fuel consumption and CO2 emissions for each vessel individually at every leg of its journey, taking its fuel type and a large number of vessel parameters and environmental factors into account.

📘

CO2 per ton is calculated for each voyage leg — if the filling factor changes between two legs (for example, if more cargo is loaded), the results for each leg will reflect the difference in cargo load.

Journey data mapping

We use the CO2 per ton data for all voyages from the main port in the load region to the main port in the destination region and match them with the respective regions within our geo-hierarchy to show emissions information on the largest global trade corridors.

Outlier detection

Since the aim of the index is to reflect carriers' normal day to day operations, we exclude any journeys that show unusual carrier behavior, such as:

• Calls at shipyard
• Abnormally long or short distance, speed, or load
• Smaller ship than usual for a given trade lane

Trade lane and carrier data modeling

With the journey data mapped to the relevant regions in our geo-hierarchy and any deviations excluded from the calculations, we move on to the final step — modeling the emissions data per carrier per trade lane for the index.