Fire Modeling

There are three main approaches to model compartment fires [2,3]. The simplest is to use the basic expressions and experimental correlations of the thermochemical and fluid processes occurring to produce an analytical model of the fire development. Analytical fire models are fast to set up and easy to use, because of the few mechanisms involved [2] however, the results are only correct in the order of magnitude, because coupling of the different fire phenomena is difficult in these models. Nevertheless, they can serve as a baseline for more sophisticated computer modeling. 

Of a more complete approach are the zone models [3], which consider two (or more) distinct horizontal layers filling the compartment, each of which is assumed to be spatially uniform in temperature, pressure, and species concentrations, as determined by simplified transient conservation equations for mass, species, and energy. The hot gases tend to form an upper layer and the ambient air stays in the lower layers. A fire in the enclosure is treated as a pump of mass and energy from the lower layer to the upper layer. As energy and mass are pumped into the upper layer, its volume increases, causing the interface between the layers to move toward the floor. Mass transfer between the compartments can also occur by means of vents such as doorways and windows. Heat transfer in the model occurs due to conduction to the various surfaces in the room. In addition, heat transfer can be included by radiative exchange between the upper and lower layers, and between the layers and the surfaces of the room. 

However, current CFD cannot provide good predictions of HRR evolution (i.e., hre growth) in real complex enclosures/scenarios. Fire modeling is not yet able to predict the HRR, and research efforts need to be tailored toward this issue. However, hre environments for hre safety design can still be calculated using CFD, if the HRR is an input data to the design process. This is because, current CFD tools provide good predictions of the effects of a hre (e.g., temperature held, smoke movements, etc.) once the HRR is provided. 

Every model, as an approximate representation of an actual phenomenon, has limitations that constrain its use and narrow its range of applicability. Generally, a more complete model contains fewer simplifications that imply more freedom and fewer limitations. However, fewer simplifications imply the need to know a larger number of fundamental parameters, to be extracted, in general, from experiments, which in fire science, often are associated to significant uncertainties. There has to be a consideration also toward whether the model has been validated for the particular circumstances of interest. 

---

Using fire models, locations of equipment, heat transfer calculations, and environmental qualifications of the equipment, it is possible to estimate the time to failure. Fragility cuives that relate fire durations and equipment damage while considering the probability of fire suppre.ssion are produced to relate to the overall PSA. These fragility curves and their use is simitar the methods ised for seismic analysis. 

Chung, G., N. Siu, and G, Apostolakis, 1985, Improvements in Compartment Fire Modeling and Simulation of Experiments, Nuclear Technology, 69, p. 14. 

MacArthur, C., 1981, Dayton Aircraft Cabin Fire Model Version 3, Vols 1 and 2 IJniv. of Dayton. Research Inst. 

Nelson, H. E. and S. Deal, 1992, Comparison of Four Fires with Four Fire Models, Proceedings of 3rd International Symposium, lAFSS, Edinburg, Scotland, pp. 719-728. 

E. V. Aibano. Critical behaviour of a forest fire model with immune trees. J Phys A (Math Gen) 27 L881-L886, 1994. 

The model is a straightforward extension of a pool-fire model developed by Steward (1964), and is, of course, a drastic simplification of reality. Figure 5.4 illustrates the model, consisting of a two-dimensional, turbulent-flame front propagating at a given, constant velocity S into a stagnant mixture of depth d. The flame base of width W is dependent on the combustion process in the buoyant plume above the flame base. This fire plume is fed by an unbumt mixture that flows in with velocity Mq. The model assumes that the combustion process is fully convection-controlled, and therefore, fully determined by entrainment of air into the buoyant fire plume. 

Hash-fire modeling is largely underdeveloped in the literature there are large gaps in the information base. Hardly any information is available concerning flash-fire radiation the only data available have resulted from experiments conducted to meet other objectives. Many items have not yet received sufficient attention. 

Bader, B. E., A. B. Donaldson, and H. C. Hardee. 1971. Liquid-propellent rocket abort fire model. J. Spacecraft and Rockets 8 1216-1219. 

The modem FB boiler is a compact economical design (typically more compact than SM boilers) and provides minimum combustion efficiencies of 80% for gas-fired and 83% for oil-fired models. They usually are small units, seldom exceeding 150 hp, and operate on a variety of fuels at typically 4 to 5 sq ft of surface area per hp. 

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. 

The toxicological "facts" I have used goes beyond present validated knowledge and thus indicates directions that future work might take to produce the data that can actually be used in a computer fire model. 

Mitler, H.E. and Rockett, J.A. Users Guide to FIRST, A Comprehensive Single-Room Fire Model, NBSIR 87-3595, 1987. 

Full scale tests are particularly valuable to obtain information on fire hazard. They can be used to validate small scale tests, and to validate mathematical fire models. The most important additional dimension full scale tests add are effects, e.g. radiation from the fire itself, which are difficult to simulate in a smaller scale. Full scale tests are very expensive and time consuming. It is essential, thus, to design them in such a way as to (a) make them most relevant (b) minimize their number and (c)... 

Tewarson, A., Lee, J.L., and Pion, R.F., "The Influence of Oxygen Concentration on Fuel Parameters for Fire Modeling," Eighteenth Symposium (International) on Combustion, 1981, The Combustion Institute, Pittsburgh, PA, p. 563. 

In this case a completely different approach was taken. It was decided to use a fire model, of zonal type, to predict smoke flows, temperatures and gas concentrations. The model chosen for these calculations was the NBS Fire and Smoke Transport model (F.A.S.T.), version 18.3 [33]. This model requires that the transport between rooms be in a horizontal manner. In order to achieve this, a virtual room is needed and a vent is needed in both the room and the plenum. In order, therefore, to analyse a broad variety of different fires and scenarios, the only product used was a low heat release wire coating. 

Application of a Genetic Algorithm to Estimate Material Properties for Fire Modeling from Bench-Scale Fire Test Data. 

C9. Fire Models and Analytical Tools Specific to the Petrochemical Industry 420... 

The equations presented in this chapter are intended to provide a simple tool for preliminary assessment of hazards. As such, most of the techniques are conservative. The units in this chapter are metric and have not been converted to English units because most equations used in fire modeling are based on metric units. The following sources provide more detailed information ... 

The process hazard analysis can be a starting point for the selection of fire scenarios. The process hazard analysis can be reviewed to develop a list of scenarios that result in fire as a consequence. Generic release sizes for small, medium, and large releases have been proposed as shown in Table 5-1 (Spouge, 1999). This saves time by eliminating the need to develop a detailed scenario. The analyst can use these release sizes to perform fire modeling calculations and determine the impact by moving the release point locations. The release criteria are considered to be representative of scenarios that could reasonably be expected to occur. 

Techniques are available to calculate conditions under which enclosed fires are ventilation- or fuel- controlled. Computer models are available to estimate compartment fire growth and temperature effects. In particular, the zone fire model C-FAST (Jones et al., 2000) is widely used. Additional information on models is contained in Appendix C. 

A fire model is a physical or mathematical representation of burning or other processes associated with fires. Mathematical models range from relatively simple formula that can be solved analytically to extensive hybrid sets of differential and algebraic equations that must be solved numerically on a computer. Software to accomplish this is referred to as a computer fire model. 

Computer fire models have been available in one form or another since the 1960s. They gained greater popularity in the 1970s and 1980s as more models became available and computer power increased however, for the most part... 

The most commonly used computer fire models simulate the consequences of a fire in an enclosure. Zone models, as well as computational fluid dynamics (CFD) models, are used for this purpose. While they are in wide use, enclosure models have limited application in assessing hazards in the petrochemical industry. They are briefly described in this Appendix for general reference purposes. 

There are no widely adopted techniques for assessing and validating computer fire models. ASTM has established a subcommittee on fire modeling, which developed guides that cover specific issues pertinent to computer fire modeling ... 

ASTM El 355 addresses evaluating the predictive capability of fire models... 

ASTM El 591 describes proceduresto obtain input data for fire models... 

ASTM El 895 addresses uses and limitations of computer fire models... 

Post-flashover fire models calculate the time-temperature history in a compartment by solving simplified forms of the energy, mass, and species equations. The concentration of various gaseous constituents can be monitored as well as vent flows. Some post-flashover fire models allow mechanical ventilation to be factored in the calculation. These types of models are most useful for determining the time-temperature exposure to a structure for a specific compartment and fuel load. Such time-temperature histories can be used for assessing the possibility of structural failure or fire spread to adjacent compartments. 

The input requirements for post-flashover types of models can be quite broad. Besides the compartment and vent dimensions, detailed fuel combustion characteristics are often needed. The fuel characteristics include the fraction of carbon, hydrogen, nitrogen, and oxygen that make up the fuel, the burning efficiency, and the quantity of fuel available for burning. Mechanical ventilation flow rates and the material properties of the compartment boundaries may be necessary. Some models can account for the heat transfer through the boundaries in detail, and may even allow the user to supply time-dependent material properties. An example of a post-flashover fire model is COMPF (Babrauskas, 1979). 

The egress and evacuation simulations are not truly fire models. They were developed in response to the need to evaluate impact of fires on the occupants of a building. Most egress models describe a structure as a network of paths along which the occupants travel. The occupant travel rates are derived from people movement studies and are often stochastic. Factors that are included in the travel rates are the age and ability of the occupant, crowding, and the type of travel path. 

---