neqsim-self-heating-ignition

NeqSim Self-Heating and Spontaneous Ignition Skill

Screening for low-temperature self-heating leading to spontaneous ignition — the mechanism behind lagging fires, where a combustible liquid soaks into porous thermal insulation and ignites with no external ignition source.

When this skill applies

Reach for this skill when you see any of these signals:

• A fire or smoke event on insulated pipework, vessels or valves with no identified ignition source.
• A leak of a combustible liquid — glycol (MEG/DEG/TEG), heat-transfer oil, lubricating oil, hydraulic fluid, amine, diesel — onto or into porous insulation (mineral wool, glass wool, calcium silicate, perlite).
• Questions phrased as "can this liquid self-ignite?", "is this insulation safe?", "how thick can the lagging be?", "why did it catch fire hours after the leak stopped?"
• A request to interpret oven / basket / hot-storage test results, or to extrapolate small-scale test data to plant scale.

The one thing to get right

Self-heating is a size-dependent problem, not a temperature-only problem.

Heat is generated throughout the volume (scales with $r^3$) but lost through the surface (scales with $r^2$). So the same material at the same temperature can be perfectly stable as a thin film and ignite as a thick layer.

This is why the lumped adiabatic screening in neqsim.process.safety.reaction.RunawayReactionAnalyzer cannot answer this question — it assumes a well-stirred mass with no spatial conduction, so it has no concept of a critical thickness or a critical surface temperature. Do not substitute it. Use neqsim.process.safety.selfheating.

Equally, do not use a Gibbs/equilibrium calculation to decide whether something self-ignites. GibbsReactor will report that any hydrocarbon is fully oxidised to CO₂ and H₂O at 20 °C, because that is thermodynamically true and kinetically irrelevant. Equilibrium gives you the fuel ($\Delta H$); only kinetics give you the hazard.