Thermal reactivity and fire risk in lagging systems: Influence of contaminants, lagging materials, and metals

Lagging fires pose a significant safety risk in industrial systems where organic contaminants interact with lagging (insulation) materials. This study used isothermal calorimetry to investigate factors influencing the self-heating and ignition propensity of various C18-based substances and rapeseed oil, as contaminants in the lagging. The contaminant and lagging under investigation were added to glass ampoules, and pentane was used as a solvent to distribute the contaminant on the lagging. The factors studied were lagging materials, molecular functionalities, and metal contaminants. It was found that substances with non-conjugated double bonds, particularly those containing bis-allylic hydrogen, gave rise to the greatest peak thermal powers. Noteworthy, all tested substances exhibited some level of reactivity, suggesting no substance can be considered completely safe without system-specific analysis. To evaluate different lagging materials, rapeseed oil was used. Greater peak thermal powers were observed with glass wool and stone wool treated at temperatures ≥500 °C, likely due to the degradation of the binder materials, as supported by TGA, SEM, and EDS analyses. Furthermore, it was found that metal salts (Mn, Fe, and Cu) and copper shavings significantly increased the reactivity, while stainless steel shavings had no significant effect. Mixtures of reactive substances behaved as single entities, and their peak thermal power could be estimated as a weighted average of the pure components’ peak thermal powers. The findings have practical implications for system design, material selection, and experimental protocols, aiding engineers in evaluating fire risks and developing safer insulation systems under realistic operating conditions. • Unsaturations increases the risk of a lagging fire • Some metals increase the risk of a lagging fire • Different lagging materials affect the self-heating propensity • An approach to estimate the risk of a lagging fire from mixtures of components is proposed