A change of fuel changes what must be measured

Decarbonizing a ship is not simply a matter of changing the fuel fed into its engine. LNG, methanol, ammonia, hydrogen and batteries differ in storage state, leak behavior, effects on people, ignition characteristics and the way a fire develops. Merely adding a few gas detectors to the fire-detection system of a conventional oil-fueled vessel is therefore insufficient. The fuel system and instrumentation must be redesigned together, including what to detect, where to place sensors, and which signals to link to ventilation, valves and fuel shutdown.

The IMO introduction to the IGF Codedescribes the international safety framework for equipment, arrangement, control and monitoring systems on ships using gas or other low-flashpoint fuels. An important point is that the Code is goal-based. Its current detailed mandatory provisions focus on natural gas; for other fuels, separate interim guidelines or the SOLAS alternative-design procedure, flag-state approval and classification-society rules must be considered together. As of September 2026, the IMO’s list of safety guidelines for alternative fuelsseparately identifies the IGF Code for LNG and guidelines for methanol, ammonia, hydrogen and fuel cells. This is why one instrumentation design cannot be applied unchanged to every green ship.

The common foundation is risk assessment and safety integration

Although fuels differ, instrumentation design poses several common questions. First, during normal operation, bunkering, maintenance, purging, loss of ventilation, pipe rupture, or after a collision or grounding, which substances could be released as liquids or gases? Second, where will those substances travel and accumulate in the leak-source space, adjoining spaces, ventilation inlets and outlets, and around deck structures? Third, should a detection signal trigger only an alarm, or should it also initiate automatic fuel isolation and an emergency shutdown?

The IGF Code requires fixed gas detection in relevant spaces and areas, with quantities determined by the space's size, arrangement and ventilation, and locations established through gas-dispersion analysis or physical smoke testing. In other words, placement cannot be finalized with a single statement such as 'light gas always means the ceiling, and heavy vapor always means the floor.' The initial jet from a low-temperature leak, evaporation, ventilation direction, equipment obstructions, ship motion and temperature conditions all alter the distribution. Actual sensor reachability must be confirmed through dispersion analysis for each leak scenario and commissioning tests.

The detection system is not a stand-alone accessory; it is an input to safety functions. Alarm setpoints, two-sensor voting logic, ventilation failure, valve closure, fuel-supply shutdown, electrical isolation, and indication on the bridge, in the central control room and locally should be connected in a cause-and-effect matrix. Redundancy and conditions for bypass operation should also be defined so that a failed sensor or one taken out of service for calibration does not create a safety blind spot. Specific setpoints and shutdown logic, however, must be finalized according to the applicable IMO instruments, flag state, classification society, and approval conditions for the engine and fuel-supply system.

LNG: address both methane explosion hazards and cryogenic leaks

The main detection target on an LNG-fueled ship is vaporized natural gas, chiefly methane. The full text of the IGF Coderequires fixed detectors in tank connection spaces, fuel-pipe ducts, machinery spaces containing gas equipment, compressor and fuel-preparation rooms, enclosed and semi-enclosed spaces where gas may accumulate, and, when necessary, at ventilation inlets to accommodation and machinery spaces. Machinery spaces protected by ESD require redundant detection systems. In general areas, it specifies visible and audible alarms at 20% LEL and safety-system activation at 40% LEL when detected by two detectors, with separate criteria for certain ventilation ducts. These values are specific LNG Code requirements and must not automatically be copied to other fuels.

Methane is lighter than air, but a cryogenic LNG leak may initially create a cold, dense vapor cloud that spreads low. Sensor height should therefore be reviewed for each scenario at the leak source, low points, ceiling pockets and ventilation outlets rather than fixed at a single upper location. The type of flammable-gas sensor also matters. Infrared sensors are useful for detecting hydrocarbons and have little dependence on oxygen, but indication errors associated with the target gas and calibration gas must be checked. Catalytic-combustion sensors can be affected by oxygen deficiency and catalyst poisoning. LNG storage and supply systems need integrated monitoring not only of gas concentration, but also tank pressure and liquid level, leakage in double-wall spaces, low temperature, ventilation airflow and fire.

Methanol: separate vapor detection from liquid-leak detection

Methanol offers the advantage of being a liquid that can be handled at ambient temperature, but it is both flammable and toxic. Because liquid may pool on a floor or in a cofferdam and then evaporate, assuming that flammable-vapor detection will catch every early leak may result in a delayed warning. IMO Interim Guidelines for Ships Using Methyl/Ethyl Alcohol as Fuelrequire liquid-leak detection in protective cofferdams around tanks, fuel-pipe ducts, fuel-preparation rooms and similar locations, and require leakage in the annular space of double-wall fuel pipes to be monitored and linked to an alarm and shutdown of the corresponding fuel supply.

The same guidelines call for fixed vapor detectors in double-wall fuel-pipe spaces, machinery spaces containing fuel equipment, fuel-preparation rooms and spaces where vapor may accumulate, with an alarm at 20% LEL and safety-system activation at 40% LEL on detection by two detectors. They also explicitly require toxicity to receive special design consideration. Installing only LEL sensors does not resolve worker-exposure concerns. The measurement principle and selectivity of the fuel-vapor sensor, expected interferents and performance at low concentrations must be reviewed, while suitable portable instruments and separate permit-to-work procedures are needed for bunkering and maintenance. Fire detection should also account for the fact that a methanol flame may be difficult to see.

Ammonia: do not miss toxic concentrations while focusing on explosion

Ammonia contains no carbon, but its high toxicity is the central design hazard. Its odor does not replace safety instrumentation. Olfactory sensitivity differs among people, high-concentration exposure is itself dangerous, and odor cannot be relied upon while wearing protective equipment. The movement of a toxic cloud must be assessed with consideration for evaporation and cooling of leaked liquid, reaction with water, and discharge locations of ventilation and scrubbing systems.

IMO Interim Guidelines for Ships Using Ammonia as Fuel, MSC.1/Circ.1687provide for continuous detection in secondary enclosures of fuel pipes, tank connection spaces, fuel-preparation rooms, bunker stations, spaces where ammonia vapor may accumulate, ventilation inlets required by the risk assessment, safe-haven air inlets, and outlets of tank pressure-relief devices. Detector locations are to be determined by gas-dispersion analysis, and the quantity in each space must enable voting logic.

The guidelines establish ammonia-specific stages, including local indication at entrances to enclosed spaces at 25 ppm, visible and audible alarms at 110 ppm, and safety-system activation at 220 ppm. Liquid leakage is detected separately at the lowest point of secondary enclosures, tank connection spaces, fuel-preparation rooms and bunker stations. This shows why directly applying LNG %LEL logic may fail to warn of toxic risk early enough. Sensor selection should verify the ammonia measurement range, response and recovery times, effects of humidity and temperature, cross-sensitivity, recoverability after high-concentration exposure, calibration gas and material compatibility of sample lines.

Hydrogen and fuel cells: respond to rapid rise and dispersion and low ignition energy

Hydrogen is extremely light and disperses rapidly, but flammable mixtures can form in pockets beneath ceiling structures or cable trays. Its small molecules create many potential leak paths, and its ignition energy is low. Liquid hydrogen adds scenarios such as cryogenic leakage, loss of vacuum insulation, air condensation and oxygen enrichment. Some infrared sensors intended for common hydrocarbons cannot detect hydrogen directly, so selection cannot be based on the label 'flammable-gas detector' alone. The ranges and environmental suitability of candidates validated for hydrogen—including catalytic, thermal-conductivity and electrochemical types—must be compared.

Approved in 2026, the IMO Interim Guidelines for Ships Using Hydrogen as Fuel, MSC.1/Circ.1701distinguish between storage and supply of liquefied and compressed hydrogen and require continuous leak detection in compressed-hydrogen tank connection enclosures. The detection strategy should consider combining different detection principles to cover diverse leak scenarios, and it provides for an alarm at 20% LEL and safety action at 40% LEL upon detection by two detectors. Sensors must jointly target potential leak points, accumulation locations and ventilation outlets.

On a ship using fuel cells, the primary fuel and the reformed fuel may differ. IMO Interim Guidelines for the Safety of Ships Using Fuel Cell Power Installations, MSC.1/Circ.1647require fixed hydrogen detectors at potential hydrogen leak points such as valves, flanges and seals, and address a 20% LEL alarm in fuel-cell spaces and spaces where flammable gas may accumulate. For a methanol-reforming system, therefore, the possibility of methanol vapor, hydrogen-rich reformate and process by-products such as CO must be assessed together with the system supplier’s hazard analysis.

Battery ships: move from fuel-gas detection to monitoring early signs of thermal runaway

Battery-electric propulsion reduces bunker-fuel leakage but does not eliminate instrumentation. An internal short circuit or overheating in a lithium-ion battery can progress from abnormal cell temperature or voltage to thermal runaway, producing flammable and toxic gases, smoke and flame. The first monitoring layer is the battery-management system’s supervision of cell voltage, current, temperature, insulation, cooling and communications integrity. At the space level, fire, smoke and heat detection and ventilation status should be combined, where needed, with off-gas or flammable-gas detection selected for the cell chemistry and test data.

No single off-gas, however, should be declared a universal leading indicator for every battery chemistry. Generation composition and timing vary with cell type, state of charge, failure mode, module and rack structure, and ventilation. The battery annex to ClassNK’s 2025 Rules for the Survey and Construction of Steel Shipsaddresses fixed fire detection in compartments containing accumulator-battery systems and requires risk assessments to consider the composition, volume and release rate of gases emitted by cells during thermal runaway, as well as explosion risk. Sensor selection should be based on cell- and module-level test data and classification approval conditions.

The maturity of international regulation also varies by fuel. The IMO stated that in 2026 it was still pursuing regulatory work on the safety of lithium-ion batteries and new technologies for shipsThis is not grounds for predicting future obligations as settled facts; it means that the latest rules and approval conditions of the applicable flag state and classification society should be confirmed early in the current project. A risk assessment must also determine how battery-room detection results are linked to charge and discharge restrictions, contactor opening, cooling and ventilation, compartment isolation, firefighting and crew response.

Multi-fuel ships also manage detection systems by operating mode

During the transition, two or more energy sources coexist in configurations such as LNG–diesel, methanol–diesel, ammonia–pilot oil and battery hybrids. Shutting down one system can overload another system or cause a loss of propulsion. For each fuel mode and state—bunkering, port stay, charging and maintenance—the active sensors, alarm settings, automatic-shutdown scope and acceptable degraded conditions must be defined. Even when every signal is consolidated on a common display, the units and meanings of the hazards must remain distinct. Toxicity alarms in ppm, explosion alarms in %LEL, oxygen concentration, liquid leakage, low temperature, smoke and cell temperature are not substitutes for one another.

Portable instruments should not be treated as miniature versions of fixed systems. Pre-bunkering checks, confined-space entry, re-entry after a leak, maintenance and verification of a gas-free condition each have different measurement purposes, locations and equipment ranges. When a sample hose is used, consider adsorption and transport delay for reactive substances and verify that the pump, filter and hose are suitable for the target substance and low-temperature conditions. Personal detectors, emergency-escape respiratory protection and protective clothing have different roles, so an alarm must not be described as a substitute for protective equipment.

Items shipyards and owners should include in instrumentation specifications

After selecting the fuel type, at least the following items should be documented together with their design basis.

  • Physical properties and toxicity, flammability and cryogenic hazards of fuel, reformate, inert gas and battery off-gas

  • A list of leak sources and dispersion or fire scenarios during normal, abnormal, maintenance, bunkering and charging conditions

  • Target substances, principles, ranges, accuracy, response times, and environmental and cross-sensitivity limits of fixed and portable sensors

  • Arrangement drawings, ventilation analysis, gas-dispersion analysis or demonstration tests supporting sensor locations and quantities

  • Alarm stages, voting logic, cause-and-effect matrix, and interlocks with ESD, valves, ventilation and electrical isolation

  • Redundancy and degraded-operation criteria during loss of power or communications and sensor failure or calibration

  • Type approval, explosion-protection and classification requirements, calibration gas, functional-test intervals, spares and crew-training plans

  • Detection reachability, alarm and automatic-shutdown tests to be performed during commissioning using actual gas or a safe alternative

Lloyd’s Register guidance on its 2026 rules for ships using gases or other low-flashpoint fuelsalso addresses methanol, ammonia and hydrogen in separate appendices. Early in the contract, the flag state, classification society, engine manufacturer, fuel-supply-system supplier, battery supplier and sensor supplier should review the same hazard list and cause-and-effect matrix to reduce late design changes.

Conclusion: begin with failure scenarios, not fuel names

Instrumentation requirements for green ships are not changing only by increasing the number of sensors. Designs should focus on methane explosion hazards and cryogenic leaks for LNG; liquid leaks, flammable vapor and toxicity for methanol; low-concentration toxic alarms and liquid leaks for ammonia; rapid dispersion and hydrogen-specific detection for hydrogen; and cell condition, thermal-runaway off-gas and fire for batteries. On multi-fuel ships, these hazards overlap according to operating mode.

A sound specification therefore does not begin with a 'universal detector for green fuels.' It begins by confirming the applicable IMO codes and interim guidelines, flag-state decisions, classification rules and manufacturer data, then developing vessel-specific leak, dispersion and fire scenarios before placing sensors and assigning safety actions. In areas where regulation is still evolving, it is more practical to record current approval criteria and design assumptions and establish a management-of-change plan than to assert future obligations. Accurately reflecting fuel-specific differences in the instrumentation system is essential to the safety of maritime decarbonization.

Sources

Cover photo

Original photo on Pexels