Full-scale fire

One problem associated with discussing flame retardants is the lack of a clear, uniform definition of flammabiHty. Hence, no clear, uniform definition of decreased flammabiHty exists. The latest American Society for Testing and Materials (ASTM) compilation of fire tests Hsts over one hundred methods for assessing the flammabiHty of materials (2). These range in severity from small-scale measures of the ignitabiHty of a material to actual testing in a full-scale fire. Several of the most common tests used on plastics are summarized in Table 1. 

The German Federal Institute for Material Testing (BAM) carried out full-scale fire tests on commercial liquefied-propane storage tanks. Tank volume was 4.85 m in each test (Schoen et al. 1989 Droste and Schoen 1988 Schulz-Forberg et al. 1984). Unprotected and protected tanks filled with propane (50% filled) were exposed to a fire. In some tests, the propane was preheated. 

Droste, B., and W. Schoen. 1988. Full-scale fire tests with unprotected and thermal insulated LPG storage tanks. J. Haz. Mat. 20 41-53. 

Note that this index only produces a relative number. Two products with widely different values of the index might be equally safe if, in fact, neither impedes escape. Conversely, two products with apparently similar values may produce different hazard levels if both products are close to the margin of safety. Thus, the scale for any index must be "calibrated", and it may well be different for each building or type of occupant. Generally, this will require a more complete hazard analysis and/or full-scale fire tests. Protocols for doing this are currently under consideration. 

This presentation covers some of the basic data and derived results are discussed. The gases species of oxygen, carbon monoxide and carbon dioxide and nitrous oxide have been measured for all the tests. In the full scale fire tests hydrogen chloride and hydrogen cyanides were measured. Hydrocarbons and their relative abundance were determined by collecting gas samples on absorbent tubes for later analysis on a gas chromatograph and a mass spectrometer. 

In a joint research project in Sweden under the main title "Fire hazard - Fire growth in compartments in the early stage of development (pre-flashover)" (1, 2) a number of different factors have been studied. In the process of developing a full-scale fire test method - "room-corner" configuration - for surface lining materials, Nordtest NT-FIRE 025, the emission of smoke and gas was studied. That study covers data from thirteen different single and... 

In the full scale fire tests some additional gaseous species were studied specifically, i.e. formaldehyde. Not all gas species were studied in every test. Hydrogen cyanide and hydrogen chloride have only been studied in situations where evolution of these species were suspected. HCN and HC1 have only been studied as collective (2, 5 or 10 minutes) samples for each fire test. It is most preferable to follow the concentrations with direct reading instruments. This has been the case for carbon monoxide, carbon dioxide, oxygen and in three out of four cases for nitrous oxide. Drager tubes were used for measurements of nitrous oxides in the DIN 53436 test. 

For the small scale fire test methods it was possible to determine the mass of the sample burnt. In the full scale fire test this could not be done. To make gas emissions comparable between the fire models, the emissions of gases in the small scale fire tests have been reduced by the amount of material burnt in each case. 

Results full scale fire tests. The analytical results from six of the thirteen materials investigated in the full scale fire test art presented in Table II. The integrated amount of each gas from the start of the experiment until flash over in the room has occured i ... 

A qualitative comparison of the results from the gas chromatography - mass spectrometer study of the different hydrocarbons from the wood materials, did not show significant differences between results from one method to the other. As far as can be judged it is mainly the amount of each component that differs between the small scale test methods and full scale fire tests. 

Results from thermally desorbed samples taken from the fire effluents of the different materials in the full scale fire experiments are presented in Table VI. The different species of hydrocarbons have been grouped together and presented in three different cathegories T< M and R representing Trace. Medium and Rich concentration. In this way it is possible to get an idea of the amount of contribution of different species of hydrocarbons to the fire effluents of each material. These results agree well in principle with results obtained by other researchers (6. 7). 

A comparison of results for fire effluents from full scale and small scale fire tests has to be done in steps. A full scale fire is a developing event where temperature and major constitutions changes continously. A small scale fire test either take one instant of that developing stage and try model that or try to model the development in a smaller scale. On a priority one level rate of heat release, temperature, oxygen concentrations and the ratio of C02/C0 concentrations have to be similar for a comparison. The full scale fire experiments reaches a temperature of 900 C at the moment of flashover, while the small scale fire tests are reaching temperatures just above 400 °C for NT-FIRE 004 and the cone experiments. For the DIN 53436-method the temperature was set to 400 °C. 

In the DIN 53436 experiments the oxygen concentration is fairly stable over each experiment and varies between 13 to 18 % for the different materials. For the full scale fire test the oxygen concentrations stays close to the ambient almost all the way up to point of flashover. 

This study demonstrate similar ranking of each material independent of test method used. At this stage it is premature to choose one test as a better small scale model of full scale fires. Each method needs further elaboration. 

Sundstrom, B. Full Scale Fire Testing of Surface Materials Technical Report SP-RAPP 1986 45, Boras, Sweden, 1986 p 117. 

A large number of procedures are now available for measuring fire properties, but many of them are of little interest since they represent outdated technologies. Thus, in order to obtain a realistic estimate of fire hazard for a scenario it is essential to measure relevant fire properties. Furthermore, the appropriate instruments have to be used, viz. those yielding results known to correlate with full scale fire test results. 

It has already been stated that the principal toxicant in a fire scenario is carbon monoxide, generated when all carbonaceous materials burn. Moreover, the carbon monoxide concentration in full scale fire scenarios depends heavily on fire load (i.e. how much material is burning, per unit volume) and on geometrical arrangements, including ventilation, while the dependence on materials is of a lower order. 

This secondary effect of materials is illustrated by the difficulties encountered, in a recent study [54], when attempts were made to correlate CO concentrations measured in small scale and full scale fire tests. The same small scale equipment (typically the cone calorimeter rate of heat release test) could predict adequately a number of very important full scale fire properties, including ignitability, rate of heat release, amount of heat release and smoke obscuration. It could not, however, be used to... 

Babrauskas, V., Bench-Scale Methods for Prediction of Full-Scale Fire Behaviour of Furnishings and Wall Linings. SFPE Technology Report 84-10, Soc. Fire Prot. Engineers, Boston, MA, 1984. 

Hill, R. G., Eklund, T. I., Sarkos, C. P., Aircraft Interior Panel Test Criteria Derived from Full-Scale Fire Tests. DOT/FAA/CT-85/23, September 1985. 

Smoke has usually been measured in the NBS smoke chamber. Such results cannot be correlated with full scale fire results and do not predict fire hazard. Rate of heat release (RHR) calorimeters (e.g. NBS Cone (Cone) and Ohio State University (OSU)) can be used to determine the best properties associated with fire hazard, as well as smoke. Results from the Cone RHR correlate with full-scale fire results. The best way to determine the fire hazard associated with smoke, for materials which do not burn up completely in a fire, is by using RHR to measure combined smoke and heat release variables, such as smoke parameter or smoke factor. 

However, the majority of small scale tests actually used to measure fire properties are incapable of determining either more than a single property or combined properties. Furthermore, there is, often, no attempt to investigate whether the test results are can be related to results to be expected in full scale fires. This is incompatible, thus, with modern concepts of fire hazard. 

There are many inadequacies in this procedure, both theoretical and experimental, Table I [9-13]. The most important deficiency is the lack of correlation of the results with those from full scale fires. This is exemplified in Table II, for both obscuration and soot [14]. 

Results do not correlate with full-scale fires. 

The Cone calorimeter yields smoke results which have been shown to correlate with those from full scale fires [10, 15-18]. The concept of a combined heat and smoke release measurement variable for small scale tests has been put into mathematical terms for the cone calorimeter smoke parameter (SmkPar) [10]. It is the product of the maximum rate of heat release and the average specific extinction area (a measure of smoke obscuration). The correlation between this smoke parameter and the smoke obscuration in full scale tests has been found to be excellent [10]. The corresponding equation is ... 

It has already been shown that the Cone calorimeter smoke parameter correlates well with the obscuration in full-scale fires (Equation 1). At least four other correlations have also been found for Cone data (a) peak specific extinction area results parallel those of furniture calorimeter work [12] (b) specific extinction area of simple fuels burnt in the cone calorimeter correlates well with the value at a much larger scale, at similar fuel burning rates [15] (c)maximum rate of heat release values predicted from Cone data tie in well with corresponding full scale room furniture fire results [16] and (d) a function based on total heat release and time to ignition accurately predicts the relative rankings of wall lining materials in terms of times to flashover in a full room [22]. 

The principal interest are, of course, real fires. Since Cone calorimeter results correlate well with those from full scale fires, it is essential, thus, to check whether OSU calorimeter results correlate well with Cone calorimeter results. There are several aspects of this, but the present paper will focus mainly on measurement of smoke. 

This work does not give definitive proof, however, that the results from the OSU calorimeter correlate well with those from full scale fire tests. It is likely that this will happen, but the product of two good correlations cannot be guaranteed to give another good correlation. Thus, correlation between OSU calorimeter and full scale fires still remains to be firmly established. 

Smoke obscuration, as measured in the NBS smoke chamber, does not correlate well with full-scale fire hazard nor with smoke as measured in heat release calorimeters. In particular, those materials which do not burn up completely in full-scale fires are being treated excessively harshly by this method. 

Azone model calculatesthe fire environment by dividing each compartment in the model into two homogeneous zones. One zone is an upper hot smoke zone that contains the fire products. The other zone is a lower, relatively smoke-free zone that is cooler than the hot zone. The vertical relationship between the zones changes as the fire develops, usually via expansion of the upper zone. The zone approach evolved from observations of such layers in full-scale fire experiments. While these experiments show some variation in conditions within the zones, the variations are most often small compared to the difference between the zones themselves. 

---