TL;DR: A vessel’s emissions figure is its fuel consumption multiplied by a published emission factor, which means almost all of the uncertainty sits in the fuel quantity rather than the conversion. Two separate things then get called verification: independent verification of the model that produces the figure, and third-party assurance of the report you publish. ZeroNorth’s AI-enabled fuel model is verified by DNV at 93.8% accuracy, which settles the first and makes the second a shorter conversation.
By the time an emissions figure reaches a report it looks like a single number. Behind it sit a fuel quantity, a conversion factor and a set of decisions about which voyages were counted and how.
Most of the time nobody asks how that number was built. The question tends to arrive later and from outside: a charterer comparing your figure against their own, or an assurance provider working through the methodology line by line.
Which is a good reason to know the answer before it is asked. That means two things: how the figure is actually calculated, and what the word verified covers when a vendor uses it.
How a vessel’s emissions figure is built
It is fuel consumption by fuel type multiplied by a standard emission factor for each fuel, so nearly all of the uncertainty sits in the fuel quantity rather than in the conversion.
Emission factors are the easy half. They are published constants for each fuel grade, they are not in dispute, and every credible system uses the same ones. If two organisations disagree about a vessel’s emissions, the factors are almost never the reason.
The fuel quantity is the whole of the work. It arrives from three sources with different characteristics: noon reports, which are human-entered daily summaries; fuel flow meters, which sample continuously but drift and occasionally gap; and bunker delivery notes, which record what came aboard rather than what was burned. Reconciling those into one defensible figure per voyage is the actual task, and it is why continuous emissions monitoring is treated as a data problem before it is treated as a reporting one.
Where emissions reporting software does the work
Its job is the validation step in the middle, because a correct conversion applied to an unchecked fuel figure produces a precise number that is still wrong.
That middle step is the part worth paying for, and it is more specific than it sounds. ZeroNorth emissions reporting software collects fuel, distance and cargo data from noon reports, flow meters and bunker delivery notes, validates each figure against physical limits and against the vessel’s own past patterns, then converts the validated fuel to CO2 and compares it against the thresholds that carry a cost: CII bands, EU ETS exposure and FuelEU balances.
The sequence matters more than any individual feature. Validation has to happen before conversion, because once a figure has been converted and aggregated into a fleet total, an error in it is no longer visible as an error. It is simply part of the number.
A useful question when you evaluate: ask what the system does with a reading that fails validation. Whether it is corrected, flagged for review or silently excluded is a design decision with different consequences at audit, and vendors answer it differently.
Two different things are called verification
Verification of a calculation model and assurance of a published report are separate exercises, and a claim about one tells you nothing about the other.
Model verification is about the method. An independent body examines the way a system turns operational data into a fuel or emissions figure, tests it against measured reality, and states how closely the two agree. This is a statement about the tool, and it travels with the tool to every vessel using it. ZeroNorth’s AI-enabled fuel model has been verified by DNV at 93.8% accuracy, which is a figure established by someone outside ZeroNorth rather than asserted by us.
Report assurance is about your disclosure. An assurance provider examines the boundary you drew, the completeness of what sits inside it, the controls around your process and the consistency of your method year on year. This is a statement about your organisation, and no vendor can supply it on your behalf.
The two are related in one direction only. A verified model does not deliver assurance, but it removes a whole category of question from the assurance conversation, because the derivation of the underlying figure is already independently established and does not have to be defended from first principles. Our overview of maritime emissions reporting covers where that sits in the wider reporting cycle.
Which level of verification your reporting needs
The level you need follows from who reads the figure, not from how large the fleet is.
Read across the bottom row and the dependency becomes clear. Assurance takes the model as an input, so a model that has already been independently verified shortens the exercise rather than replacing it.
The bottom line
Judge an emissions figure by how its fuel quantity was derived and who checked the derivation, not by how precisely it is presented.
The vendor-neutral question to put to anyone you are evaluating, including us, is this: where does the fuel figure in this number come from, what happens to a reading that fails validation, and has the method that produced it been examined by anyone outside your company?
For ZeroNorth the answer is three specifics. Fuel, distance and cargo data are validated against physical limits and the vessel’s own history before any conversion is applied. The conversion uses the same published emission factors as everyone else, so the difference sits in the input rather than the arithmetic. And the model producing the fuel figure has been verified by DNV at 93.8% accuracy, so its derivation is established by a third party rather than by us.
If you are working through this now, see how Emission Analytics handles validation and reporting, review pricing, or book a demo and bring a figure you have been asked to explain.
.png)
.png)
