Marine CEMS Explained: from Stack Sensor to Compliance Report
A continuous emissions monitoring system (CEMS) on a ship works by sampling exhaust gas directly at the stack, converting raw sensor signals into pollutant and CO2 concentrations, and logging that data continuously so it can be reported against standards such as EU MRV and IMO DCS. The process runs in stages: gas extraction or in-situ sensing, analysis, data logging, quality assurance, and reporting. Understanding each stage helps an operator judge whether a system is actually measuring emissions or just estimating them from fuel consumption.
Where the Measurement Starts: the Stack
A marine CEMS takes its readings at the point exhaust leaves the engine or boiler, before it is diluted by outside air. Two basic sampling approaches exist. Extractive systems pull a slipstream of exhaust gas through a heated line to an analyzer cabinet elsewhere on the vessel, which protects sensitive instruments from heat and vibration but requires more plumbing and maintenance. In-situ (or direct, stack-based) systems place sensors inside or across the stack itself, measuring gas properties without extracting a sample. Direct, stack-based CO2 measurement aligned with EU MRV and IMO DCS methods is the approach SailPlan's earlier maritime monitoring work was built around, and it remains relevant background for understanding how shipboard CEMS architecture is typically laid out, even though that specific platform is no longer a SailPlan product.
How the Analyzer Converts Gas Into a Number
Once gas reaches the analyzer, whether by extraction or in-situ sensing, it passes through one or more detection methods, commonly infrared absorption for CO2 and other techniques for NOx, SOx, or particulate matter depending on what the vessel needs to track. The analyzer converts the physical signal into a concentration reading, usually expressed as parts per million or a percentage by volume. This raw concentration is not yet a usable report figure. It still needs to be corrected for stack gas flow, temperature, pressure, and moisture content before it reflects an actual mass emission rate.
Why Emissions Factors Matter and Where They Go Wrong
Many compliance frameworks allow an alternative to direct measurement: calculating emissions from fuel consumption using a standard emissions factor, rather than measuring the exhaust itself. This is faster to implement but introduces a gap between estimated and actual emissions, because a fixed factor cannot account for variation in fuel quality, engine condition, or combustion efficiency across a fleet. Direct stack measurement avoids this gap by reading what actually leaves the stack rather than inferring it from fuel burned. The trade-off is that direct measurement requires more hardware, more calibration discipline, and more data handling than a fuel-based calculation.
From Raw Signal to Continuous Log
A functioning CEMS does not just take occasional readings; it logs data continuously, typically at short, regular intervals, so that short-duration events (a load change, a fuel switch, an upset condition) are captured rather than averaged away. This continuous log is the raw material for every downstream report. Good data at this stage looks like a steady, timestamped stream with no unexplained gaps; gaps or flatlines usually indicate a sensor fault, a power interruption, or a maintenance event, and any credible reporting process needs to flag and explain them rather than silently smoothing them over.
Quality Assurance: Calibration and Drift Checks
Analyzers drift over time due to sensor fouling, temperature cycling, and component aging, so a CEMS program includes scheduled calibration checks against known reference gases. Operators should expect periodic zero and span checks, and a documented record of when calibrations occurred and what adjustments were made. This record matters for two reasons: it is often required by the reporting standard itself, and it is the only way to tell, after the fact, whether a reading reflected the exhaust or an instrument that had quietly drifted out of tolerance.
Turning the Log Into a Compliance Report
The final stage aggregates the continuous log into the format a regulatory or verification body requires. For EU MRV, FuelEU Maritime, or IMO DCS, that typically means voyage-level or annual totals, built from the underlying continuous data rather than a single snapshot. This is the stage where most scattered-data problems surface: if telemetry, fuel records, and voyage logs sit in separate, incompatible systems, assembling an accurate report means manually reconciling them, which is slow and error-prone. See SailPlan's background on FuelEU Maritime data collection challenges for more on where that reconciliation work tends to break down, and the comparison of CEMS and PEMS approaches on ships for how the two measurement philosophies differ in practice.
What This Means for Today's SailPlan
SailPlan's earlier maritime monitoring platform, which performed direct stack-based emissions measurement for cruise, naval, and commercial fleets, was acquired by Verret Marine Consulting, led by Chad Verret, and is now being applied to predictive maintenance and machinery monitoring in the offshore, LNG, and commercial marine sectors. Operators with questions about that monitoring platform should direct them to Verret Marine. SailPlan's current work is a different layer: a unified, machine-readable data model that normalizes OEM telemetry, maintenance logs, procedures, financials, warranty terms, personnel records, and procurement data into one schema that any authorized tool can query. For a fleet running a CEMS, a PEMS, or any other OEM-specific monitoring system, this matters because emissions data is just one more dialect among the many a vessel already generates. A model-agnostic schema means that telemetry from an emissions analyzer, a maintenance log entry about sensor calibration, and a warranty record for the analyzer itself can all be queried together, without a custom integration for each source, and without paying what SailPlan calls a recurring "integration tax" every time a new dashboard or model is added.
What to Expect From a Demo
Fleet operators evaluating how their existing telemetry, maintenance records, and compliance data could be unified can request a demo, which walks through how SailPlan builds a machine-readable model from the systems already in place. A team member follows up within one business day. Operators should expect the conversation to focus on what data sources exist today, from OEM telemetry platforms to maintenance spreadsheets, rather than a sales pitch for emissions hardware, since that product line now sits with Verret Marine.
Frequently Asked Questions
How is a CEMS different from a PEMS on a ship?
A CEMS measures exhaust gas directly at the stack using physical sensors, while a predictive emissions monitoring system (PEMS) estimates emissions from engine operating parameters using a model rather than a direct gas measurement. Each approach has different hardware, calibration, and data requirements.
Does continuous monitoring mean a reading every second?
It means readings are logged at short, regular intervals rather than taken occasionally, so transient events are captured. The exact interval depends on the analyzer and the reporting standard being followed, and the research behind this packet does not specify one fixed interval that applies across all systems.
Why would a vessel use fuel-based emissions factors instead of direct stack measurement?
Fuel-based calculation is simpler to implement because it does not require stack-mounted hardware, but it is also less precise, since a standard factor cannot capture variation in fuel quality or combustion efficiency across different engines and voyages.
Who do I contact about SailPlan's older emissions monitoring platform?
That platform was acquired by Verret Marine Consulting, led by Chad Verret, and is now applied to predictive maintenance and machinery monitoring in the offshore, LNG, and commercial marine sectors. Questions about that specific system should go to Verret Marine rather than SailPlan.
What does SailPlan offer fleets today if it no longer sells a CEMS?
SailPlan's current offer is a unified, machine-readable data model that normalizes OEM telemetry, maintenance logs, manuals, financial and warranty records, and institutional knowledge into one schema any authorized tool can query, starting with maritime fleet operators.
Related resources
SailPlan builds the machine-readable data model that makes every AI tool in your stack actually work. Request a demo to see it in action.