In ocean freight, decarbonization is widely discussed, yet maritime Scope 3 emissions are still managed with inputs that were never built for operational decisions. Emissions outcomes shift with routing, vessel behavior, port disruption, and utilization, while most reporting approaches smooth that execution reality into assumptions.
To highlight this, we will be analyzing a shipper’s calculations for their carbon footprint based on a carrier’s schedule under three scenarios and how each scenario’s outcome compares to the actual carbon footprint of the voyage.
The three scenarios are:
- Scenario A: Clean Cargo Calculation
- Scenario B: OOCL Calculation Tool
- Scenario C: 2024 Voyage Data of the Vessel on the Same Route
In one real long-haul voyage, commonly used approaches - like GLEC framework and Clean Cargo data - diverged from recorded performance by minus 8.5 kg CO2e per TEU on one method and by as much as plus 416 kg CO2e per TEU on another. That is not noise. It is the difference between a manageable variance and a material misstatement at scale.
The above graph presents the exposure of a shipper with annual shipments of 1,000 TEU across the scenarios. Under Scenario A, emissions are overestimated by 8,500 kg; under Scenario B, they are underestimated by 415,900 kg, and under Scenario C by 158,200 kg.
Per container inaccuracies multiply fast. At portfolio volumes, the same per TEU deltas can shift reported totals by hundreds or thousands of tonnes, change supplier comparisons, and complicate audit narratives. To settle this uncertainty, many companies resort to Book & Claim mechanisms, often at substantial cost, only to discover that the underlying emissions baseline is still too approximate to prove real reductions with confidence.
Many teams feel covered because they use carrier calculators or programs such as Clean Cargo. These are valuable and have advanced standardization, but they are often misunderstood. In most cases, the output is not primary execution data for your specific shipment or voyage. It is an average under a defined methodology, which cannot reliably reflect what actually happened on a given sailing, including the route taken, distance sailed, schedule recovery, and utilization swings. When averages are treated as shipment truth, emissions become reportable but not manageable. You cannot explain variance, defend results with confidence, or identify real reduction levers without guessing.
A major routing difference is today recorded for vessels leaving Asia destined for Europe. Most of those vessels choose the longer and safer Cape of Good Hope routing and do not cross the Suez Canal. However, this is not true for all voyages; many carrier schedules deploy ships that pass through the Suez Canal.
This information is excluded when calculating your carbon footprint using a route-average approach.
Figure 1 Routing through Suez (Blue) vs routing around the Cape of Good Hope (Green) for a vessel leaving Shanghai (China) and traveling to Rotterdam (Netherlands). Source: AIS signal of the same vessel across different time periods (2023 vs 2024).
This article tests common calculation approaches against high-granularity voyage performance to show how quickly reasonable methods diverge from actual outcomes and why that gap creates regulatory, financial, and strategic risk. The aim is practical: replace averages with voyage intelligence so emissions can be managed like any other logistics KPI, measured, explainable, and actionable, even when operations deviate from plan.
The Truth Behind Carrier Schedules
To illustrate why industry averages and carrier calculators fall short, let's examine a real-world voyage and specifically a service connecting China (Shanghai) with Europe (Rotterdam).
OOCL, CMA CGM, and Evergreen are alliance partners sharing capacity on the same vessel for this service. Yet each carrier published different schedules for the identical voyage. The schedules show minor discrepancies in ETAs (Estimated Time of Arrival) and ETDs (Estimated Time of Departure).
|
Carrier |
Port |
ETA |
ETD |
|
OOCL |
Shanghai |
11/21/2025 |
11/21/2025 |
|
OOCL |
Ningbo |
11/24/2025 |
11/24/2025 |
|
OOCL |
Singapore |
11/30/2025 |
11/30/2025 |
|
OOCL |
Rotterdam |
12/30/2025 |
12/30/2025 |
|
CMA-CGM |
Shanghai |
11/21/2025 |
11/22/2025 |
|
CMA-CGM |
Ningbo |
11/24/2025 |
11/25/2025 |
|
CMA-CGM |
Singapore |
11/30/2025 |
12/1/2025 |
|
CMA-CGM |
Rotterdam |
12/30/2025 |
1/2/2026 |
|
Evergreen |
Shanghai |
11/21/2025 |
11/22/2025 |
|
Evergreen |
Ningbo |
11/24/2025 |
11/25/2025 |
|
Evergreen |
Singapore |
11/30/2025 |
12/1/2025 |
|
Evergreen |
Rotterdam |
12/30/2025 |
1/2/2026 |
Scope 3 Emissions Calculation: 3 Scenarios Analysis
To calculate Scope 3 emissions, many shippers use industry averages, which underestimates the impact of vessel and voyage characteristics on the carbon footprint.
To highlight the severity of these characteristics and the extent to which ignoring them leads to miscalculation of emissions, we will consider the following calculation scenarios:
- Scenario A: Clean Cargo Calculation
- Scenario B: OOCL Calculation Tool
- Scenario C: 2024 Voyage Data of the Vessel on the Same Route
Thereafter, we will analyze the actual voyage corresponding to the specific schedule.
Scenario A: Clean Cargo Calculation
The 2024 Clean Cargo report provides specific intensities for each route. A shipper sending cargo from Shanghai to Rotterdam would refer to the provided intensities. Assuming the shipper optimally uses those intensities, he would use the schedule data to decompose the voyage into legs and calculate the carbon footprint of each leg separately.
Based on Clean Cargo, the intensity for the first two legs of the voyage is 94.5 g CO2e per TEU km (Intra Northeast Asia) and 94.4 g CO2e per TEU km (from Northeast Asia to Southeast Asia). Thereafter, the intensity is 40.8 g CO2e per TEU km for voyages destined for Rotterdam.
Regarding distance calculations, Clean Cargo relies on historical voyage data to determine emissions intensities. However, when shippers calculate their Scope 3 exposure independently, they must source voyage distance data themselves. In practice, this often means relying on the Minimum Feasible Distance (MFD) required to complete the voyage. This approach does not account for deviations from standard routes, such as rerouting via the Cape of Good Hope, which can significantly increase actual sailing distance. As a result, emissions calculations based solely on MFD may underestimate real-world emissions. Under this scenario, using Clean Cargo methodology, emissions would amount to approximately 1,020.2 kg per TEU transported.
|
Scenario A |
Clean Cargo Calculation |
Intensity (g CO2e per TEU km) |
Distance (km)* |
Emissions (kg)/TEU |
|
Shanghai - Ningbo |
94.5 |
241.6 |
22.8 |
|
|
Ningbo - Singapore |
94.4 |
3,836 |
362.1 |
|
|
Singapore - Rotterdam |
40.8 |
15,569.7 |
635.2 |
|
|
Total |
- |
19,647.3 |
1,020.2 |
*Distance is calculated based on 2024 historical data
Scenario B: OOCL Calculation Tool
Many carriers offer online emissions calculation tools. However, those tools also suffer from accuracy limitations. First, they assume full capacity utilization, without accounting for vessel size or actual utilization. Second, they typically use Minimum Feasible Distance (MFD) rather than actual voyage distance data in their calculations. Even if they change their formula to account for the Cape of Good Hope distance, some vessels could still cross the Suez Canal, and the formula would again yield inaccurate results for those voyages.
According to OOCL, the voyage leg between Shanghai and Ningbo is 301.9 km, the voyage leg between Ningbo and Singapore is 3,894.8 km, and the final leg between Singapore and Rotterdam is 15,601.2 km, for a total distance of 19,797.6. Based on the intensities provided by the tool (as shown in the table below), total emissions are 595.8 kg per TEU transported.
|
Scenario B |
OOCL Calculation Tool |
Intensity (g CO2e per TEU km) |
Distance (km) |
Emissions (kg)/TEU |
|
Shanghai - Ningbo |
54.68 |
301.9 |
16.5 |
|
|
Ningbo - Singapore |
54.28 |
3,894.8 |
211.4 |
|
|
Singapore - Rotterdam |
23.58 |
15,601.2 |
367.9 |
|
|
Total |
- |
19,797.9 |
595.8 |
Scenario C: 2024 Voyage Data of the Vessel on the Same Route
To illustrate how methodologies that do not account for voyage- and vessel-specific characteristics, but instead make assumptions about vessel capacity and utilization, fail, we will compare the above scenarios with historical data on this vessel’s voyages on that route in 2024. The service analyzed is common, and in 2024, multiple voyages were completed by the same vessel on that route.
The average utilization exceeded Clean Cargo's assumed 70% but remained well below the full-capacity utilization assumed by carrier tools. It also deviates significantly per leg, which is not accounted for in any of the previous scenarios.
|
2024 Voyage Data |
AVG Utilization |
|
Shanghai - Ningbo |
81% |
|
Ningbo - Singapore |
87% |
|
Singapore - Rotterdam |
98% |
On average, voyages were highly efficient, with emissions reaching 6.6 kg per TEU for the first leg of the voyage (Shanghai-Ningbo), 147.7 kg per TEU on the second leg (Ningbo-Singapore), and 699.2 kg per TEU on the third leg (Singapore-Rotterdam).
Total emissions averaged 853.5 kg per TEU for voyages on this exact itinerary by this specific vessel. Compared to the 2024 voyage data, Scenario A overestimates emissions by 16.3%, while Scenario B underestimates them by 43.2%. Remember, the assumptions of both Scenario A and B are drawn from 2024 voyage data. Therefore, even within the same time period, both Scenario A and Scenario B fail to reflect the complex reality of actual vessel performance in a voyage.
|
Scenario C |
2024 Voyage Data |
Intensity (g CO2e per TEU km) |
Actual Distance (km) |
Emissions (kg)/TEU |
|
Shanghai - Ningbo |
27.23 |
272 |
6.6 |
|
|
Ningbo - Singapore |
38.51 |
3,894.9 |
147.7 |
|
|
Singapore - Rotterdam |
44.91 |
21,924.7 |
699.2 |
|
|
Total |
- |
26,091.6 |
853.5 |
However, this is not what VesselBot offers. All three scenarios above share a fundamental flaw: they rely on historical data that is no longer relevant to current operations. Some also rely on assumptions, averages, and incomplete data. Now, let us compare them to the actual voyage data.
Actual Voyage Data: The VesselBot Approach
Data accuracy and veracity in VesselBot’s approach are the highest among all scenarios. We use actual voyage data from AIS vessel signals, our digital twin model, and data enrichment to provide accurate information on vessel size and utilization, speed at each point in the voyage, and the distance covered, among other voyage characteristics.
The emissions we computed for the different legs of the voyage are not averages; they are the actual emissions based on the recorded intensity and the distance covered for each TEU transported. When the next rotation of this service occurs, those numbers will change because voyage conditions are highly dynamic. This is precisely why we can’t base emission calculations and decarbonisation strategies on averages.
First, let’s look at vessel utilization across the voyage’s legs. From Shanghai to Ningbo, the vessel was loaded at just 56%, well below the historical 2024 average (81%), the Clean Cargo average (70%), and the carriers’ assumed full-capacity utilization.
|
Actual Voyage Data |
Utilization |
|
Shanghai - Ningbo |
56% |
|
Ningbo - Singapore |
89% |
|
Singapore - Rotterdam |
92% |
Small changes in capacity can significantly affect Well-to-Wake intensity, as the vessel carries fewer TEU and produces less transport work. Indeed, compared to the 2024 average for this vessel, intensity is 83.6% higher in the first leg, 7.4% higher in the second leg, and 20.3% higher in the third leg.
Remember, we are comparing this voyage with neither the industry average, nor the vessel average across all its voyages, nor the route average. We are comparing actual voyage data with the average voyage data for the same vessel on the exact same itinerary and route one year prior. The differences recorded are not small and highlight that even granular historical data remains outdated for current operations.
|
Actual Voyage Data |
VesselBot |
Intensity (g CO2e per TEU km) |
Actual Distance (km) |
Emissions (kg)/TEU |
|
Shanghai - Ningbo |
50 |
443.9 |
12.1 |
|
|
Ningbo - Singapore |
41.3 |
3,981.3 |
158.6 |
|
|
Singapore - Rotterdam |
54 |
21,927.1 |
841.1 |
|
|
Total |
- |
26,352.3 |
1,011.7 |
On this voyage, total emissions reached 1,011.7 kg for each TEU a shipper would send to Rotterdam from Shanghai. A shipper using Clean Cargo calculations would overestimate their carbon footprint. Instead of 1,011.7 kg per TEU, they would get 1,020.2 kg per TEU.
What Real-Time Data Actually Enables
This voyage between Shanghai and Rotterdam demonstrates why industry averages and carrier calculators systematically underestimate emissions. Scenario A overestimated emissions by 8.5 kg per TEU, Scenario B underestimated emissions by 416 kg per TEU and Scenario C underestimated them by 158 kg per TEU. These aren't outliers. They're the inevitable result of using approximations instead of actual voyage data.
Even the difference recorded within Scenario A and the actual voyage data is major. At first, 8.5 kg might seem insignificant, but it is 8.5 kg per TEU transported. A shipper sending 15 TEU per quarter from Shanghai to Singapore on this vessel would overestimate their Scope 3 emissions exposure by 510 tons of CO2e.
|
Actual Voyage Emissions Compared to Each Scenario |
|
|
Scenario A |
-8.5 kg CO2e/TEU |
|
Scenario B |
+416 kg CO2e/TEU |
|
Scenario C |
+158 kg CO2e/TEU |
Clean Cargo Intensities are averages that do not account for vessel size. Those intensities have been calculated using an average well-to-wake emissions intensity of 70%. As our real-life example showed, intensities differ significantly from the average. In our white paper, we further analyze the claim that Clean Cargo intensities rest on more than one problematic assumption. Apart from the assumed 70% utilization, intensities do not differ between voyages originating from the same area and terminating at a specific port, even when the origin ports differ. A voyage originating from Shanghai, destined for Rotterdam, will have the same intensity as one originating from Singapore and heading to Antwerp.
Even if Clean Cargo provides different intensities for each major carrier, these intensities are inconsistent for the same reasons and also because:
- Carrier schedules are dynamic: A vessel might be deployed to a different route in 2026 than it was in 2024. But a shipper basing his calculations on Clean Cargo today will be using the 2025 report, which is based on 2024 data.
- Routing varies across carrier services: routes between the same ports may differ. One vessel might cross the Suez Canal to reach Rotterdam, while another will go around the Cape of Good Hope. For example, the vessel we analyzed in this blog crossed the Suez Canal, called at Piraeus, and then called at Rotterdam. Today, it skips the Suez Canal and heads straight to Rotterdam.
The above highlights that even when Clean Cargo sources data from carriers, this data includes assumptions, such as that utilization is 70% on every voyage. After that, Clean Cargo computes aggregated estimates for each route/carrier. Those aggregates even out crucial vessel and voyage-specific details, such as the port of origin and destination, and the vessel capacity.
For the primary data sourced from carriers to be actionable, it would need to be real-time; the carrier would need to disclose all voyage-related information in real-time. Among other things, that would include:
- Vessel size
- Voyage-specific vessel utilization
- Voyage-specific routing
- Voyage-specific fuel consumption
In reality, this does not happen; ergo, the data shippers get from Clean Cargo and the Carrier’s Calculation Tool are not real-time or accurate.
VesselBot's Supply Chain Sustainability Platform was built to solve this problem. The methodology tracks actual vessel movements via AIS signals, applies digital twin modeling to calculate real fuel consumption, and enriches the data with vessel-specific characteristics including size, utilization, speed variations, and operational events like weather delays. This is the actual emissions profile of each voyage your cargo takes.
The platform provides three indispensable capabilities:
- data granularity that captures voyage-specific variables,
- accuracy and veracity grounded in actual operations rather than assumptions, and
- real-time visibility into supply chain performance as it happens.
This combination transforms reactive supply chain management into proactive optimization. Actual voyage data reveals which carriers consistently deliver promised performance, which routes generate unexpected costs, and where operational decisions create measurable financial impact. That distinction determines whether your supply chain runs on agreegates or intelligence and whether it can adapt to disruption or merely report it afterward. Building resilient, regenerative transportation networks requires visibility into what actually happens, not what averages suggest should happen.
This blog introduces the key findings. Our white paper presents a comprehensive benchmark across additional scenarios, between carrier tools, industry methodologies, historical data, and actual voyage emissions. Download the complete analysis here.