Characterization of Flaring and Non-Flaring Container Filled Votive Candles

Authors were provided two sets of 24 exemplar candles. Of these, 24 were identified as the same that the provider had observed “flaring”. Fourier Transform Infrared Spectroscopy (FTIR) analysis of the material showed that the candles that did not flare were soy based wax and that those that did flare were petroleum based wax. Burning the candles showed that melted wax in some petroleum based candles was ignited by the wick flame and produced a pool fire inside the candles’ glass container. Based on measured mass loss rate and the heat of combustion of petroleum based wax, the calculated power output of a normal burning candle is ca . 30 W. When the liquid wax pool ignited, the output was ca . 230 W, with significant flame extension above the top edge of candle’s glass container. The FTIR analysis also showed that twelve of the petroleum based wax candles had a simple hydroxyl compound, likely an alcohol, peak that was not present in the second sample of one dozen petroleum based wax candles. Burn testing showed that the candles from the sample with hydroxyl ignite a pool fire after burning for 15 to 30 min; the second sample of petroleum based wax candles, those that did not have the hydroxyl compound, did not ignite a pool fire. This suggests that the hydroxyl compound reduces the flashpoint of the wax to an ignition temperature attained after burning the candle for tens of minutes. Once the hydroxyl compound petroleum based wax melts and reaches a temperature above its flashpoint/flame point, it vaporizes sufficiently to be ignited, with flame extension of 152 mm (6 in) or more above the wax pool. This flame extension or “flaring” represents a significant fire hazard in that it may ignite nearby combustibles or fracture the candle’s glass container and release burning wax.