At What %LEL Should a Car-Deck Gas Alarm Trip?
Not at %LEL for early warning. Vent gas has an LFL near 10% vol, so a 10% LEL trip is ~1% gas, thousands of ppm: an explosion setpoint, not detection.
A %LEL setpoint is the wrong instrument for early warning and the right one for explosion prevention — and a vehicle deck needs both. Lithium-ion vent gas has a lower flammability limit of roughly 10% by volume at ambient temperature, falling toward 6% when the gas is hot. A combustible-gas detector tripping at 10% LEL is therefore reading about 1% vent gas — around 10,000 ppm — long after an electrochemical sensor would have seen the first hundred.
The distinction matters because the two questions get merged in specifications. An owner asks for 'gas detection on the car deck'; a supplier quotes an industrial %LEL instrument built to IEC 60079-29-1; an underwriter reads the certificate and assumes early warning. The instrument will do what it is certified to do — flag a mixture approaching explosive — and it will do it minutes after the window that matters for stopping vehicle-to-vehicle spread has closed.
What the LEL of vent gas actually is
About 10% by volume for a high-SOC cell at room temperature, and it moves with state of charge, chemistry and temperature. Ma, Liu and Yu (ACS Omega 2020, 5(43):28096) computed the limits for NCA vent gas across state of charge: at 0% SOC the mixture is 95% CO₂ and its lower limit sits near 12%; at 75–100% SOC — roughly 49% CO, 23% H₂, 7% CH₄, 20% CO₂ — it falls to about 10%, with an upper limit near 62%. Baird, Archibald, Marr and Ezekoye (Journal of Power Sources 2020, 446:227257) put the NMC range at 7.6–9.0% across SOC and found LFP vent gas has a higher lower limit than NCA or LCO, with lower flame speed and overpressure.
Temperature is the number a deck engineer should keep. The same ACS Omega study reports the lower limit dropping about 1.8 percentage points for every 150 K of initial temperature: roughly 10% at 298 K, 8.5% at 450 K, 6% at 600 K. Vent gas leaves a pack at several hundred degrees. Near the source, before it cools and mixes, the mixture is flammable at a lower fraction than the room-temperature figure suggests — which is one reason a jet ignites at the vent and becomes a fire rather than a cloud.
How much gas one car puts on the deck
Tens of cubic metres from a single pack — enough to make a flammable layer, not enough to make a flammable deck. A 75 kWh NMC pack is on the order of 20,000 Ah of cell capacity. At the roughly 0.07 mol/Ah reported for NMC cells (Batteries 2025, 11(9):323) that is about 1,400 mol, or 35 m³ of gas at ambient temperature; at the 3 L/Ah upper bound in the venting literature it is nearer 60 m³. Taking a 10% lower limit, 35 m³ of vent gas forms about 350 m³ of flammable mixture. Fully mixed through an enclosed deck of, say, 150 m × 30 m × 2.2 m — around 10,000 m³ — the same gas sits at 0.35% vol, about 3.5% LEL, and no combustible-gas detector on the deck would trip.
The two numbers are the whole argument. A single vehicle cannot bring a ventilated ro-ro space to its lower limit as a whole, so a %LEL alarm set for the space will not see it. But the same vehicle can put a 350 m³ flammable layer under a deckhead, in a corner, or in the pocket a spray pattern creates — and that layer is what a deflagration needs. Whether it forms is a question of mixing, not of quantity, and the case that water systems can worsen it is made in the sibling blog on DNV's sprinkler finding.
What the ventilation rule buys and what it does not
SOLAS II-2/20.3.1.1's 10 air changes per hour for closed vehicle and special-category spaces on passenger ships (six on cargo ships) is a dilution rate, and dilution is a race against release rate. Ten changes of a 10,000 m³ space is about 1,700 m³ per minute of fresh air against a pack that releases 35–60 m³ over the several minutes of a violent vent — roughly a 30- to 50-fold dilution if mixing were perfect. It is not perfect, which is why the layer argument survives the ventilation rule.
DNV GL's 2020 maritime battery study — run with the Norwegian and Danish maritime authorities and MARAD — found the limit from the other direction: with around 4,000 Ah of cells failing simultaneously in a battery room, even 100 air changes per hour could not hold overpressure below the roughly 0.5 barg at which bulkheads fail. A single car is not 4,000 Ah at once; a bay of eight cars in a spread fire is.
Why a %LEL instrument cannot be the early-warning layer
Because of what it measures and how. A combustible-gas detector reports fraction-of-LEL, so its resolution is set by the LEL of the calibration gas. Calibrated to hydrogen (LEL 4% vol), a 10% LEL alarm is 4,000 ppm; calibrated to methane (5%), it is 5,000 ppm; calibrated to the vent mixture itself, around 10,000 ppm. The electrochemical H₂ and CO channels that the corpus's gas-signature article treats as the early markers are working in tens to hundreds of ppm — one to two orders of magnitude below the point a %LEL instrument begins to move. The two are not competing products; they measure different stages of the same event.
Detection principle compounds the gap. Infrared combustible-gas detectors, the low-maintenance choice on tankers, are blind to hydrogen and carbon monoxide — neither diatomic absorbs in the hydrocarbon band — so an IR %LEL point on a car deck is watching the CH₄ and C₂H₄ minority of the vent, roughly 7–10% of it at high SOC. Catalytic-bead detectors do respond to H₂ and CO, but at %LEL resolution, and with the poisoning and drift behaviour the corpus covers separately. Neither is a fault; both are the instrument doing what its certification says.
So where should the setpoints sit?
On two different instruments, for two different actions. The early-warning layer belongs to ppm-class sensing — solvent vapour, H₂, CO, CO₂ — with thresholds set from bench and staged-test data against the deck's own baseline, and its action is verification and response, not ventilation. The explosion-prevention layer belongs to a %LEL instrument with a low alarm at 10–20% LEL and a high alarm at 25% LEL, and its action is the NFPA 69 logic: force ventilation, hold water back from the pocket if the fixed system allows it, and treat the space as an explosion hazard for the crew. The MDPI Batteries 2025 result that real LFP vent gas ignited below its theoretical LEL in a 1.73 m³ rig (Batteries 2025, 11(10):352) is the reason the explosion alarm should sit at the conservative end of that range, not the reason to push the detection alarm up to meet it.
For the specification that lands on an owner's desk this reduces to one question to ask a supplier: what concentration, in ppm, of which gas, does your first alarm correspond to? If the answer is given in %LEL, the system being offered is an explosion-prevention system. It may be a good one. It is not the early-warning layer, and the deck still needs one.
How RoRoSAFE helps
An %LEL setpoint is an explosion alarm, not early warning. RoRoSAFE's gas sensing works at the level this article argues for: electrolyte-vent VOCs at the vehicle, far below LEL, fused with per-vehicle thermal readings. The crew is alerted while the gas is a warning sign, long before it is an explosion hazard.
Pilot: one deck · installed alongside the berth · no drydock · 6 months of dashboard access
Sources
- 1. Ma, B., Liu, J., Yu, R. — 'Study on the Flammability Limits of Lithium-Ion Battery Vent Gas under Different Initial Conditions', ACS Omega 2020, 5(43):28096–28107 (open access via PMC7643196): NCA vent-gas composition by SOC; LFL ~12% (0% SOC) to ~10% (75–100% SOC), UFL to ~62%; LFL ~8.5% at 450 K and ~6% at 600 K; limits rise with pressure.
- 2. Baird, A. R., Archibald, E. J., Marr, K. C., Ezekoye, O. A. — 'Explosion hazards from lithium-ion battery vent gas', Journal of Power Sources 2020, 446:227257 (OSTI 1574807): LFL, laminar flame speed and maximum adiabatic overpressure by chemistry and SOC; NCM LFL 7.6–9.0%; LFP higher LFL, lower flame speed and overpressure than NCA/LCO.
- 3. DNV GL — 'Technical Reference for Li-ion Battery Explosion Risk and Fire Suppression' (report dated 12 January 2020; partners Norwegian Maritime Authority, Danish Maritime Authority, MARAD, Corvus Energy, Kongsberg, ABB, Stena, Scandlines, Damen), as reported by The Maritime Executive (7 January 2020): 4,000 Ah simultaneous failure defeats 100 ACH; flaming cells produce as little as half the gas of non-flaming; sprinklers can pocket gas and worsen explosion risk.
- 4. NFPA 855 (Standard for the Installation of Stationary Energy Storage Systems) and NFPA 69 (Standard on Explosion Prevention Systems): gas detection with exhaust ventilation activated at 25% LFL, or deflagration venting to NFPA 68, for lithium-ion ESS rooms.
- 5. Batteries (MDPI) 2025, 11(9):323 — per-Ah vent-gas volume by chemistry (~0.02 mol/Ah LFP, ~0.07 mol/Ah NMC) and composition; Batteries (MDPI) 2025, 11(10):352 (Gill, Buston, Howard et al., published 27 September 2025) — real LFP prismatic-cell vent gas ignited and deflagrated in a 1.73 m³ rig at concentrations below the theoretical LEL of the homogeneous mixture.
- 6. SOLAS Chapter II-2, Regulation 20.3.1.1 — ventilation capacity for closed vehicle spaces, closed ro-ro spaces and special-category spaces: 10 air changes per hour on passenger ships carrying more than 36 passengers, 6 on cargo ships and other passenger ships. IEC 60079-29-1 — performance requirements for flammable-gas detectors reading in %LEL.
Questions, answered
What is the lower explosive limit of lithium-ion battery vent gas?+
About 10% by volume for a high-state-of-charge cell at room temperature, with published ranges of roughly 7.6–9% for NMC and higher for LFP. It is not fixed: at 0% SOC the gas is mostly CO₂ and the limit rises toward 12%, and with initial gas temperature the limit falls by about 1.8 points per 150 K — near 6% at 600 K.
Is a %LEL gas detector enough for EV fire detection on a car deck?+
No. A 10% LEL alarm on vent gas corresponds to roughly 1% gas, around 10,000 ppm, whereas the early markers — electrolyte solvent vapour, hydrogen, carbon monoxide — are detectable at tens to hundreds of ppm. A %LEL instrument is an explosion-prevention device; it can confirm a hazardous accumulation but cannot provide the minutes of lead time early-warning detection is meant to give.
Can one EV fire create an explosive atmosphere on a ro-ro deck?+
Not deck-wide, but locally yes. A 75 kWh pack releases on the order of 35–60 m³ of vent gas, enough for a few hundred cubic metres of flammable mixture. Mixed through a 10,000 m³ ventilated space that is a few percent of LEL; collected in a layer under the deckhead or a pocket created by spray, it can reach it. Mixing, not quantity, decides.
Does SOLAS ventilation prevent a vent-gas explosion?+
It reduces the probability for a single vehicle but is not a guarantee. SOLAS II-2/20 requires 10 air changes per hour on passenger ships (6 on cargo ships), which dilutes one pack's release many times over if mixing is good. DNV GL's 2020 maritime study found that with about 4,000 Ah of cells failing at once, even 100 air changes per hour could not prevent explosion-magnitude overpressure.
Continue the thread
H₂, CO, CO₂: The Thermal-Runaway Signature
The order matters more than the shortlist. Solvent vapour beats H₂ by about fifteen minutes — and LFP and NMC invert which gas dominates.
Does the Off-Gas Ever Leave the Battery Pack?
An EV pack is sealed to IP67 and releases only through a calibrated burst path. A deck sensor sees what survives that path, then dilution.
LFP vs NMC: Tuning Detection Thresholds
LFP vents about a third the gas of NMC and burns far cooler, but proportionally richer in hydrogen. A detector tuned only on NMC under-warns on LFP.
How Fast Does a Gas Sensor Drift at Sea?
Electrochemical cells can lose up to 20% of sensitivity a year. On a vehicle deck, the threshold you set at commissioning is not the one you have later.
Does Deck Ventilation Delay Fire Detection?
Yes. SOLAS II-2/20 mandates continuous ventilation on closed vehicle decks, and that airflow dilutes smoke and off-gas before detectors see it.
Can Water Suppression Worsen an EV Blast?
DNV found sprinklers can make a battery explosion worse by pocketing vent gas — a sharp question for the water-based systems now mandated on vehicle decks.