Particle size emissions from PVC electrical cable fires

Electrical cables are in every building and form a significant part of fire loads and can through electrical faults be the first item burnt in some fires. PVC insulated cables are still quite common in buildings and this work investigates Prysmian PVC cables. Deaths and injuries in fires are dominated by the influence of toxic smoke emissions and most of the work on the hazards of smoke are concerned with the toxic gases such as CO. However, fires are large producers of particulate material at levels over 1000 times that in controlled combustion and there is little knowledge of the role of ultra-fine particles in fires and none at all for electrical cable fires. The cone calorimeter fire material testing equipment was used in the present work, which is an ideal test procedure for particle size measurement, as controlled dilution (100/1) of the fire products occurs which enabled diluted samples to be used for particulate number measurement. The Cambustion DMS500 transient particle size analyser was used to determine the particle size distribution. The cone calorimeter uses a 100mm square test specimen and this was filled with 10 100mm lengths of the PVC cable. The test specimen was on a load cell so that the mass burn rate was determined. The cone calorimeter ignites the specimen using a conical electrical heater that is calibrated to achieve a control radiant heat flux on the test specimen, which was 35 kW/m2 in the present work. The fire occurred in a restricted air supply with an insulated air box around the 100mm square test fire. A chimney on the conical heater exit was used to obtain a raw gas sample for toxic gas analysis using a heated Gasmet FTIR. For gases dilution is undesirable as oxidation of the toxic gases may occur. For particles the chimney temperature was too low for carbon oxidation to be significant. The dilution process also condenses unburned hydrocarbons and carbonyl species, which may form nano aerosols and these may be the source of the 10nm particles measured in the present work. HCl is a major product of PVC fires and hence hydrochloric acid aerosols are likely in the particulate measurements. In previous work of the authors, PVC cable fires were investigated with free ventilation and HCl yields of about 50% were measured with Acrolein at 5% yield and Formaldyhyde at 3%. Thus there are plenty of liquid aerosol possible in the diluted products of PVC fires. The results showed a large nuclei number peak at about 10nm. The coarse particle peak only started after flaming combustion occurred and this was initially at 200nm, which increased to 300nm after 1000s. The 10nm peak was high for the first 200s, then dropped dramatically and slowly reformed later in the fire and at the end of the fire was very high with a low coarse particle peak. The FTIR gas species will be used to speculate on the likely composition of the nanoaerosols as a function of time in the fire