Maritime emissions measurement has a structural problem. The methodologies that most shippers rely on today were built for reporting, not for operational decision-making. They are constructed from carrier-reported historical data, aggregated into trade-lane averages and converted using standardized factors, then applied as a single intensity figure, regardless of which vessel carried the cargo, how loaded it was, how fast it traveled, or how long it spent waiting in port. That design made sense when emissions data was primarily a compliance input. It is increasingly inadequate when shippers aim to reduce Scope 3 emissions, defend carrier selection decisions with measurable evidence, and treat logistics performance as a financial variable.
The gap between what the aggregates report and what actually happens on a given voyage is not a minor calibration issue. It is the difference between a number that satisfies an audit and a number that informs a decision. This blog uses actual 2025 ocean voyage data, sourced from VesselBot’s Logistics Intelligence Platform, to trace that gap across four levels of granularity: trade-lane average, port-pair average across all carriers, port-pair average for a specific carrier and port-pair average for a specific vessel. At each level the error narrows. The conclusion that data leads to is unambiguous: accuracy requires shipment-grade execution data.
The findings have direct implications for three decisions every supply chain and sustainability team faces: how to calculate Scope 3 emissions exposure accurately, how to select carriers based on executed efficiency rather than reported estimates, and how to identify the operational levers that produce real supply chain emissions reductions rather than adjustments to how those emissions are counted.
This is becoming increasingly critical as regulatory pressure on supply chain emissions disclosure intensifies across jurisdictions, including the EU Corporate Sustainability Reporting Directive (CSRD), California’s Climate Corporate Data Accountability Act (SB 253), and emerging legislation such as New York’s S3456.
Key Takeaways
- Port-pair analysis narrows the gap but still overestimates by 29.6% even when carrier identity is known.
- The same vessel, on the same route, produces a 17.3% emissions intensity variance across three voyages in a single year because of speed, distance, and port call differences.
- Scope 3 accuracy, carrier selection, and decarbonisation strategy all depend on voyage-level execution data, not aggregated estimates.
- The data in this article is drawn from VesselBot's Logistics Intelligence platform, which tracks actual vessel movements via AIS signals and models fuel consumption through digital twin technology.
How Supply Chain Emissions are Miscalculated in Practice
To make the argument concrete, this blog works through a single commercially representative route: Singapore to Rotterdam, one of the most traded port pairs connecting Asia and Northern Europe. Using 2025 execution data from VesselBot’s platform, covering AIS-tracked voyages across the major carriers operating this route, the analysis moves through four levels of granularity. It starts with the trade-lane average for the full China/East Asia to Northern Europe corridor. It then narrows to the port-pair average for Singapore to Rotterdam across all carriers, then to the port-pair average for Evergreen specifically, and finally to the average for a single vessel. At each stage, the question is the same: how much does the calculated emissions intensity differ from what actually happened on a given voyage?
Consider what that looks like for a single shipment. A shipper sends cargo from Singapore to Rotterdam. A compliance report calculates the Scope 3 footprint using an aggregates derived from actual voyages. The number looks reasonable. Depending on which methodology was used, it is wrong by anywhere between 9.4% and 47.6%.
That is not an outlier result. It is what the data consistently shows when aggregate estimates are measured against actual voyage execution. The ensuing analysis traces why, step by step.
The Limits of Trade-Lane Averages in Maritime Emissions
One of the most widely adopted industry frameworks, the Clean Cargo Initiative, assesses maritime emissions by assigning a single intensity factor to an entire trade lane, derived from carrier-reported aggregate data. These methodologies are aligned with the GLEC framework for logistics emissions accounting. The first step of the analysis was to follow a similar, yet more precise approach. Using VesselBot’s digital twin technology, we gathered data about the actual execution of all containership voyages in the trade lane China/East Asia to Northern Europe and calculated the Well-to-Wake (WTW) emissions intensity for 2025, which was equal to 54.9 g CO2e per TEU km. It applies regardless of which carrier operated the voyage, which vessel was used, how full it was, or how fast it traveled.
The next level of granularity is to examine how the trade-lane average varies by carrier. Across all major carriers, the members of the Ocean Alliance were the most efficient; COSCO recorded an emissions intensity of 50.2, followed by Evergreen (50.7) and CMA CGM (52.1). Hapag-Lloyd voyages, on the other hand, were less efficient, at 58.5 g CO2e per TEU km.
|
Well to Wake Emissions Intensity (g CO2e/TEU km) |
China/East Asia - Northern Europe |
|
COSCO |
50.2 |
|
Evergreen |
50.7 |
|
CMA-CGM |
52.1 |
|
Hapag-Lloyd |
58.5 |
Hapag-Lloyd's trade-lane intensity is 6.5% above the corridor average. Evergreen and COSCO sit below it. These differences matter for carrier selection. But they still rest on aggregated data. They reflect the average outcome in the trade lane by carrier. That rarely matches the outcome of a specific voyage on the trade lane.
Port-Pair Analysis in Maritime Emissions: More Precise, Still Insufficient
Narrowing the analysis to a specific port pair, Singapore to Rotterdam, the most commercially significant connection between Asia and Northern Europe, produces a different picture. The carrier rankings shift. Evergreen moves to first position with a Well-to-Wake intensity of 48.21 g CO2e per TEU km. COSCO follows at 50.9, CMA CGM at 53.1, and Hapag-Lloyd at 60.7.
|
Well to Wake Emissions Intensity (g CO2e/TEU km) |
Singapore - Rotterdam |
|
COSCO |
50.9 |
|
Evergreen |
48.2 |
|
CMA-CGM |
53.1 |
|
Hapag-Lloyd |
60.7 |
The shift in rankings between trade-lane and port-pair analysis is not a marginal adjustment. COSCO, Evergreen, and CMA CGM are Ocean Alliance partners and share capacity on the same vessels on this route. Among the alliance carriers, CMA CGM and Evergreen show a 9.2% performance gap at the port-pair level. That gap is lower in the trade-lane view. It widens when the analysis is anchored to actual routes.
What Drives Maritime Emissions Intensity at the Voyage Level
Port-pair averages reflect trends. They do not explain variance. To understand what actually drives emissions intensity, the analysis needs to reach the voyage level.
Across 2025 voyages from Singapore to Rotterdam, the carriers with the lowest Well-to-Wake emissions intensity shared a consistent profile:
- Large vessel capacity with high utilization, generating more transport work per unit of fuel consumed
- Voyage distances close to the minimum feasible (MFD) distance number
- Short port call duration, as time spent in port adds to total voyage emissions without contributing to transport work.
The table below shows carrier averages for this port pair in 2025, using execution data from VesselBot's platform:
|
|
AVG Emissions (tons) |
AVG WTW Intensity |
AVG Time in Port (days) |
AVG Utilization |
AVG TEU Carried | AVG Distance (km) |
|
Evergreen |
14,277 |
48.2 |
3.2 |
94% |
19,649 |
21,896 |
|
COSCO |
13,452 |
50.9 |
3.3 |
94% |
18,044 |
21,920 |
|
CMA-CGM |
12,774 |
53.1 |
3.3 |
93% |
16,737 |
21,889 |
|
Hapag-Lloyd |
8,818 |
60.7 |
3.2 |
87% |
9,888 |
21,414 |
Evergreen operated the largest vessels on average, carrying 19,649 TEU per voyage at 94% utilization, tied with COSCO. It also covered the second-longest distances, averaging 21,896 km per voyage, and spent an average of 3.2 days in port. Despite generating the highest average emissions per voyage at 14,277 tons, it produced the largest average transport work at 431.6 million TEU-km, resulting in the lowest emissions intensity on the route (48.2 g CO2e per TEU km).
Hapag-Lloyd's divergence is significant. Smaller average vessel capacity, lower utilization (87%), and shorter average distances combine to produce a WtW intensity of 60.7 g CO2e per TEU km, a 26% gap versus Evergreen on the same port pair. A shipper selecting carriers based on trade-lane averages alone would not see this difference.
Why Vessel-Data Is Not Enough for Accurate Maritime Emissions
Knowing which vessel carried the cargo is more precise than a trade-lane or carrier average. Some carrier online calculation tools use this approach, applying vessel-specific parameters but relying on assumed capacity and fixed utilization rates. As the data below shows, this is still insufficient.
In 2025, Evergreen operated 93 voyages from Singapore to Rotterdam using more than 40 different vessels. The intensity range across vessels on the same route is already substantial:
|
Vessel Name |
WtW Intensity (g CO2e/TEU km) |
Utilization |
TEU Carried |
Distance (km) |
Speed (knots) |
Transit Time (days) |
Time in Port (days) |
|
Ever Arm |
37.2 |
100% |
23,992 |
22,125 |
16.2 |
30.9 |
7.8 |
|
Ever Atop |
39.5 |
99% |
23,649 |
21,968 |
16.3 |
30.4 |
2.2 |
|
Ever Aria |
39.9 |
99% |
23,764 |
21,962 |
16.6 |
29.9 |
11.7 |
|
Ever ALP |
40.2 |
100% |
23,992 |
22,191 |
15.8 |
31.7 |
4.7 |
|
Ever Act |
40.3 |
100% |
23,764 |
22,044 |
17.1 |
29.0 |
2.8 |
|
Ever Ace |
43.3 |
100% |
23,764 |
21,948 |
17.9 |
27.6 |
3.4 |
In January 2025, Ever Arm arrived at the Port of Rotterdam from Singapore after covering 22,125 km and carrying 23,992 TEU. Total transit time was 30.9 days at an average speed of 16.2 knots. The vessel arrived at a congested port and spent 7.8 days there. The Well-to-Wake emissions intensity of the voyage was 37.2 g CO2e per TEU km.
In October of the same year, Ever Ace departed from the same origin port and arrived at the same destination. It covered 21,948 km at an average speed of 17.9 knots over 27.6 days, spending 3.4 days in port. Its Well-to-Wake emissions intensity was 43.3 g CO2e per TEU km.
Even though the port of departure and destination did not change, the voyage-specific variables did, and those differences directly affected the Well-to-Wake intensity. This illustrates the need for vessel- and voyage-specific data to ensure accuracy. Even knowing which vessels are deployed on each trade is not sufficient. This is the approach used in some carrier online calculation tools, and it is one of the reasons those tools produce inaccurate results.
Even, as the next section will demonstrate, identifying the vessel transporting your cargo, and applying an average WtW intensity based on data from that vessel’s operations on the relevant trade route, is still insufficient.
The Same Vessel, the Same Route, Three Different Outcomes
Ever Arm completed the Singapore to Rotterdam voyage three times in 2025. The number of TEU transported was identical across all three voyages: 23,992. Everything else changed.
|
Vessel Name |
WtW Intensity (g CO2e/TEU km) |
TEU Carried |
Distance (KM) |
Speed (kn) |
Transit Time (days) |
Time in Port (days) |
|
Ever Arm |
37.2 |
23,992 |
22,125 |
16.2 |
30.9 |
7.8 |
|
Ever Arm |
41.4 |
23,992 |
21,959 |
18.1 |
29.9 |
2.8 |
|
Ever Arm |
43.6 |
23,992 |
21,990 |
18.4 |
27 |
3 |
Across these three voyages, the differences compound:
- Distance decreased by 0.6% from the most efficient voyage out of the three to the least efficient one
- Average speed increased by 14%, from 16.2 to 18.4 knots
- Transit time decreased by 12.6%, from 30.9 to 27 days
- Transport work fell by 0.6%
The combined effect of these individually small differences led to a 17.3% increase in Well-to-Wake emissions intensity. No average, whether trade-lane, port-pair, carrier-level, or vessel-specific, captures this. Only voyage-level execution data does.
The Impact of Inaccurate Scope 3 Emissions on Supply Chain Decisions
The cumulative effect of using averages instead of execution data becomes concrete when measured against a real voyage. Take the Ever Arm voyage from Singapore to Rotterdam in January 2025, which recorded a WtW intensity of 37.2 g CO2e per TEU km. A shipper who sent cargo on that voyage would calculate Scope 3 emissions exposure as follows under each methodology:
|
Methodology/Data Source |
WtW Intensity (g CO2e/TEU km) |
Overestimation vs. Actual Voyage |
|
Trade-lane average (China/East Asia – Northern Europe) |
54.9 |
+47.6% |
|
Port-pair average (Singapore – Rotterdam, all carriers) |
51.6 |
+38.7% |
|
Port-pair average (Singapore – Rotterdam, Evergreen only) |
48.2 |
+29.6% |
|
Port-pair average (Singapore – Rotterdam, Ever Arm vessel only) |
40.7 |
+9.4% |
Each step down the granularity ladder closes the error, but none of them reaches zero. The overestimation at the trade-lane level would cause that shipper to overstate Scope 3 emissions exposure by nearly half. That is not a rounding error. It translates directly into misinformed carrier selection, inflated carbon credit purchases, and decarbonization investments allocated based on a fictional baseline.
The same data foundation that corrects these errors is also the foundation for optimization. When supply chain emissions are calculated at the voyage and shipment level, they become a variable a logistics team can actually act on: selecting carriers based on executed efficiency, adjusting routing decisions around port performance, and identifying where speed, utilization, or vessel size is creating avoidable cost and carbon exposure.
This shift from aggregated reporting to execution-level measurement is increasingly reflected in industry benchmarks, including independent evaluations such as Drewry’s Emissions Measurement Providers Comparison Guide, which assess the underlying data methodologies used across the market.
Averages and aggregates are a starting point. They are not a decision tool. The gap between them and execution-grade data is where supply chain strategy either succeeds or stalls.
SOURCES
- Clean Cargo Initiative
- EU Corporate Sustainability Reporting Directive (CSRD)
- California Climate Corporate Data Accountability Act (SB 253)
- US Senate Bill S3456
- GLEC Framework for Logistics Emissions Accounting (Smart Freight Centre)
- Drewry 2024 Emissions Measurement Providers Comparison Guide
About the authors Manos Charitos is Data Analyst at VesselBot, where he works with AIS-tracked voyage data, digital twin models, and shipment-level execution records to measure and interpret maritime emissions across global supply chains. The analysis in this article is his own work, drawn from 2025 voyage data covering the Singapore to Rotterdam route across all major carriers. His background combines Mathematics and Shipping, giving him both the quantitative foundation to model emissions at the voyage level and the operational context to understand what those numbers mean for shippers making carrier and routing decisions. Maria Bena is Communications Manager at VesselBot, where she develops the content and communications strategy that brings freight intelligence and supply chain emissions data to the executives who need to act on it. Working directly with data analysts, logistics experts, and sustainability leaders, she translates shipment-level insights into strategic narratives for C-suite audiences across manufacturing and global logistics. Her background spans over 15 years in communications across the private, public, and non-profit sectors, with a focus for the past four years on the intersection of logistics data, Scope 3 reporting, and supply chain decision-making.