Regulatory compliance tells you a fire suppression system is permitted. It does not tell you it is the right one. Each area of a vessel presents a distinct fire hazard profile, and the agent that protects a machinery space well may be the wrong choice a bulkhead away.
Fire suppression on a vessel is not a single decision. It is a series of decisions, one per protected space, each governed by what is burning, what else is in the compartment, and whether anyone might still be inside when the system operates. Carbon dioxide, foam, and water mist all satisfy SOLAS in the right application. Choosing between them means understanding what each agent actually does to a fire, and what it does to everything else in the room.
The three dominate marine fixed-system practice for good reason: each attacks a different leg of the fire triangle. That difference, rather than cost or familiarity, should drive the specification.
Three Agents, Three Mechanisms
Carbon Dioxide: Proven, Precisely Regulated, Unforgiving
CO₂ remains the reference standard for total flooding of Category A machinery spaces, and the FSS Code specifies its quantity in unusually concrete terms. For machinery spaces, the carbon dioxide carried must give a minimum volume of free gas equal to the greater of two figures: 40% of the gross volume of the largest machinery space protected, excluding the part of the casing above the level at which the horizontal area of the casing is 40% or less of the horizontal area of the space, or 35% of the gross volume of that space including the casing.
For cargo ships of less than 2,000 gross tonnage, those percentages reduce to 35% and 30% respectively. For the purposes of the calculation, the volume of free carbon dioxide is taken at 0.56 cubic metres per kilogram.
Quantity alone does not extinguish a fire. The FSS Code requires that the fixed piping be arranged so that 85% of the gas can be discharged into a machinery space within two minutes.
A system holding the correct mass of CO₂ but discharging it too slowly allows the atmosphere to dilute through leakage and ventilation paths before an extinguishing concentration is reached. Slow discharge is a failure mode that a cylinder-weight check will not reveal.
The operational cost of CO₂ is that it works by making the compartment uninhabitable. The same mechanism that starves the fire starves anyone left inside. The FSS Code addresses this with procedural safeguards rather than engineering ones: protected spaces require audible and visual alarms warning of release, operating for long enough to evacuate the space and in no case for less than 20 seconds, and a warning notice must be displayed at the point of entry alerting personnel to the asphyxiation hazard.
The 20-second minimum is a floor, not a target. It is the shortest delay the Code will accept, not an assessment of how long your crew needs to clear a specific engine room. A space with restricted egress, working platforms at multiple levels, or routine hot work in progress warrants a longer pre-discharge delay, established by drill rather than assumed.
Foam: Effective on Fuel, Now Constrained by Chemistry
Foam earns its place wherever the credible fire is a liquid fuel in depth or spreading across a surface: tank decks, cargo pump-rooms, helidecks, and machinery-space bilge areas. By blanketing the fuel it suppresses vapour production, which is what actually sustains a hydrocarbon fire. No other agent in common marine use does this as directly.
What has changed is not the physics but the chemistry. The SOLAS Chapter II-2 amendment adopted by IMO resolution MSC.532(107) in June 2023 entered into force on 1 January 2026, prohibiting fire-extinguishing media containing perfluorooctane sulfonic acid above 10 mg/kg, equivalent to 0.001% by weight.
The prohibition covers stored foams in both fixed and portable equipment. Ships constructed on or after 1 January 2026 must comply from delivery. Existing ships must comply no later than the first survey falling on or after that date, which means a substantial share of the world fleet is working through this at its next scheduled attendance.
A foam stock that was fully compliant in 2024 may not be compliant at your next survey. The determining factor is the concentrate’s PFOS content, which is a documentation question before it is a procurement one.
Operators should be able to produce, for each foam stock on board, evidence of its composition against the 10 mg/kg threshold. Where that documentation does not exist, the foam should be treated as suspect until tested or replaced.
Replacement is not a straight swap. A different concentrate may carry a different induction rate and different compatibility with the existing proportioning equipment, so the system, not just the drum, needs reviewing.
We covered the prohibition and its survey implications in more detail in our earlier analysis of the PFOS ban.
Water Mist: Equivalence Proven by Testing, Not by Formula
Water mist occupies a different regulatory position from the other two, and this is the point most often misunderstood. There is no equivalent of the CO₂ volume formula for mist. Systems installed under Chapter 7 of the FSS Code are approved in accordance with the guidelines for equivalent water-based fire-extinguishing systems for machinery spaces and cargo pump-rooms, set out in MSC.1/Circ.1165, adopted in 2005 and subsequently amended.
Approval under those guidelines rests on full-scale fire testing at an accredited laboratory, demonstrating that the system achieves an outcome equivalent to CO₂ total flooding in a Category A machinery space. The consequence is significant: a water mist system is certified as a tested configuration, not as a calculated quantity.
A CO₂ system can be verified against a formula. A water mist system can only be verified against the test configuration it was approved under. Change the nozzle spacing, the pump capacity, or the compartment geometry, and the approval no longer describes what is installed.
That distinction has practical consequences during modification and repair. Substituting a nozzle type, relocating heads to accommodate new equipment, or altering the protected volume can invalidate the basis of approval even when every component remains individually certified. Any change to a mist installation should be referred to the class society before the work is done, not after.
In exchange, mist offers what CO₂ cannot: it is survivable. Crew can remain in or re-enter a space during discharge. It causes far less water damage than a conventional sprinkler or deluge arrangement, and it does not require the space to be sealed to the same degree as a gas system, since it does not depend on maintaining a compartment-wide concentration.
Matching the Agent to the Space
The specification question is rarely which agent is best in the abstract. It is which agent suits a particular compartment, given its hazard, its occupancy, and its boundaries.
- Can the space be sealed reliably? Gas systems depend on holding a concentration. Where ventilation dampers, cable penetrations, or door discipline are unreliable, a gas system’s real-world performance diverges from its design case.
- Is anyone likely to be inside? Continuously or intermittently manned spaces push the decision away from CO₂ and towards mist, whatever the volume calculation permits.
- Is the credible fire a pool or a room? A liquid fuel hazard in depth favours foam. A three-dimensional machinery fire favours total flooding or mist.
- What does re-entry look like? CO₂ requires the space to be ventilated and proven safe before anyone returns, which extends the incident well past extinguishment.
- What is the cost of the agent itself? Water damage, residue, and the loss of sensitive equipment are part of the total cost of a discharge, not incidental to it.
In practice most vessels carry a combination, and the combination is the correct answer rather than a compromise. A machinery space under total flooding, a pump-room under foam, and accommodation under sprinkler or mist is not indecision. It is three hazards being addressed on their own terms.
What Actually Drives the Decision
Three factors decide most specifications, and none of them is the agent’s headline extinguishing performance.
Occupancy is the first and most decisive. A system that cannot be discharged while people are present is a system that will be hesitated over, and hesitation during a machinery-space fire is measured in minutes the fire does not give back.
Verifiability is the second. A system whose correct condition can be confirmed at a routine inspection will stay in a known state. A system whose condition depends on records held ashore, or on a test configuration nobody on board has seen, will drift.
Consequence of discharge is the third. The agent that extinguishes the fire but writes off the switchboard has solved one problem and created another. This is where mist has changed the calculation for many operators over the past decade, particularly on vessels where downtime is the dominant commercial risk.
Frequently Asked Questions
Can water mist replace an existing CO₂ system in a machinery space?
It can, where the mist system holds approval as an equivalent water-based system under MSC.1/Circ.1165 for the relevant space category, and where the installation matches the configuration that approval was granted against. This is a class-approved conversion, not a like-for-like swap, and the compartment geometry has to fall within the tested envelope.
Does the PFOS prohibition apply to portable extinguishers as well as fixed systems?
Yes. The prohibition covers fire-extinguishing media containing PFOS above 10 mg/kg in both fixed and portable equipment, including stored foam concentrate. A compliant fixed system with non-compliant portable units is not a compliant vessel.
How is the required CO₂ quantity calculated?
For machinery spaces it is the greater of 40% of the gross volume of the largest protected machinery space excluding the qualifying part of the casing, or 35% of that volume including the casing, reducing to 35% and 30% for cargo ships under 2,000 gross tonnage. Free carbon dioxide is taken at 0.56 cubic metres per kilogram.
Is a 20-second pre-discharge alarm always sufficient?
Twenty seconds is the regulatory minimum, not a recommendation. The Code requires the alarm to operate for the length of time needed to evacuate the space. Where egress is restricted or the space is worked at multiple levels, the appropriate delay should be established from evacuation drills rather than defaulted to the floor.
Why does a fixed system pass inspection and still fail in a real fire?
Most commonly because the inspection confirmed the agent was present but not that it could be delivered. Cylinder weight, pressure, and certification dates say nothing about discharge rate, pipe obstruction, damper closure, or boundary integrity. Those are the parameters that decide whether a design concentration is actually reached.
Sources: IMO International Code for Fire Safety Systems (FSS Code), Chapters 5, 6 and 7; IMO resolution MSC.532(107) on the prohibition of fire-extinguishing media containing PFOS; IMO MSC.1/Circ.1165, Revised Guidelines for the Approval of Equivalent Water-Based Fire-Extinguishing Systems for Machinery Spaces and Cargo Pump-Rooms; DNV, Lloyd’s Register and Bureau Veritas class guidance on the 1 January 2026 PFOS entry into force.
Continue reading
-
compliance The IMO FSS Code: A Practical Compliance Guide for Shipowners
-
compliance Lifesaving Appliance Inspection and Maintenance Under SOLAS
-
compliance Lifeboat Systems: Types, Release Mechanisms & SOLAS Requirements Explained
-
compliance SOLAS Compliance for Life-Saving Appliances: What Ship Operators Get Wrong
-
compliance Top Fire Safety Non-Compliance Issues on Ships: PSC Findings & How to Fix Them