A Powerful and Effective Tool for Fire Investigators - Essay Example

Potential Applications and Limitations of Computer Modelling to Fire and Explosion Investigations Student’s Name Instructor Course Date of Submission Potential Applications and Limitations of Computer Modelling to Fire and Explosion Investigations Introduction Fire and explosive investigations are two distinguishable fields of forensic science that cover a number of sub-disciplines. As a result of the broad nature of the events that precipitate fire and explosion investigations, there are important public safety and health issues that have to be considered. Forensic investigations related to fire comprise three things: finding out its origin (the place where the fire began), cause (the factors that combined the ignition and fuel source), and finally, the responsibility for the fire[Len09]. There are a number of issues that need to be addressed during fire investigations such as failure of suppression and detection systems, failure of occupants to escape, and rapid fire spread, among others[Sut12]. After scene investigation is carried out, laboratory analyses involving engineering and chemistry are often required. Forensic investigations related to explosives are also composed of three things: determining the nature of the explosive (chemical or mechanical), determining the nature of the fuel (chemical, dust or gas explosive), and finally, the responsibility for the explosion[NFP11]. Explosions caused by terrorist bombings are classified separately, and are normally investigated by specialists investigators hired by national governments. Fire investigation procedures are extremely standardized and discussed in trade literature and forensic science, but mostly because of safety concerns, very few has been written concerning post-blast investigations. Investigating both explosions and fires require the application of scientific principles from the beginning[Len09]. For the last 3 decades, computer fire modeling application to solving fire dynamic problems has been increasing. The main focus of computer modelling is to provide scientific and mathematical research into the problems and behavior related to fire[Put13]. Since the discovery of computer fire models, several private, university and government laboratories have been helping with the advancement and improvement of the models to make sure that the mathematical equations dependably represent real-world fire behavior. Evolutions in computing and fire science and technology have led to an increasing number of great mathematical models that are applied in fire safety engineering design and analysis. Applications A fire model can be described as a mathematical or physical representation of burning or other processes related to fire[Bea09]. A computer fire model can be defined as software that is used to numerically solve hybrid set of algebraic and differential equations on a computer. The most frequently applied computer fire models simulate the magnitudes of a fire inside an enclosed place. The models used for this purpose include Zone models and CFD (computational fluid dynamics or field models[Ole03]. Enclosed fire models are also used for the simulation of smoke and fire spread in multi-room structures. The other set of computer fire models forecast how people, systems or materials respond when exposed to particular fire conditions[Har14]. The models that fall in this category include detection activation and sprinkler models. Further examples include calculation methods to evaluate the load-bearing capacity of assemblies and structural elements to fire, and models that simulate human egress and behavior during a fire incident. Computer fire models are mainly used for two reasons: 1) reconstructing and analyzing the fire, and 2) fire-safe design of a structure[0113]. Reconstructing and analyzing the fire is much easier because there is always additional information accessible such as fire department reports, eyewitness accounts, forensic evidence, among others.[Bea05] In this context, a computer fire model is most commonly used to supplement the additional information in attesting that a specific hypothesis is plausible or not[Jan02]. Computer fire models have been found to be great tools for investigators carrying out post-fire reconstruction and analysis. Modern computer fire models are capable of performing the following: 1) predict fire growth and behavior, 2) evaluate occupant egress, 3) analyze smoke control systems, 4) predict detector actuation time, and 5) provide plausible post-fire timeline of events[Har14]. A zone model can be described as a software or computer program that forecast the impacts of the development of a fire inside an enclosure. In nearly all applications, the enclosure is not completely enclosed as vents, windows, and doors are normally included in the calculation. Zone models for compartments have been developed for both multi-room and single-room configurations. This model split rooms into one or more zones. The models which are commonly used assume that a room is composed of two zones: an upper layer with heated combustion products, and a lower layer with cooler air moderately free from combustion products[Bey08]. In the two zone model, the fire creates the relation between the lower and upper layers. The two layers are presumed to be well integrated to ensure that the conditions in all the layers are constant. For instance, the predicted temperature in the hot or upper layer is assumed to be same throughout. Most of the models involve provisions for openings to other rooms or to the outside and for heat losses to the ceiling and walls[Gor08]. Model inputs generally include the fire heat release rate, room furnishings characteristics, the locations and sizes of room openings, and the building materials and room dimensions[NIS11]. Outputs of these models normally include combustion gas concentrations, the height of the surface forming a common boundary between the lower and upper layers, lower and upper temperatures, time to flashover, and prediction of fire alarm or sprinkler activation time[NFP11]. There are a number of zone fire models available commercially from several sources globally[Har14]. CFD models split an enclosure or a room into several small 3-Dimensional boxes referred to as cells. The enclosure may include so many cells ranging from centimeters to meters in size. CFD fire models are established on the fundamental physical principles of momentum, mass, and energy conservation[Gor08]. The computer computes the movement of smoke and heat between the cells over a period of time. At given point in time, there is a possibility of finding the gas concentrations, velocity, and temperature in every cell. Just like the case with zone models, the conditions in every cell are presumed to be constant. However, the large number of cells makes it possible to predict the conditions in the enclosure in much greater detail. CFD models have the capability of predicting the conditions in very small and very large spaces, in complex multi-room configurations, and in spaces with complex shapes that are not probable with zone models. Because of their complexity, CFD models need a higher level of skills and expertise to operate and are presently operated on costly computer equipment. General purpose CFD fire models are available commercially from several sources globally. Advanced CFD models, particularly designed for fire safety analysis, are currently being developed at the NIST (National Institute of Standards and Technology)[NIS11]. Detector response models primarily forecast the time to activate an initiating device. Although most of these models forecast the response of a fusible link, sprinkler or thermal detector to a fire-induce flow, a number of them compute the response of a smoke detector[Dec09]. Generally, detector response models use the zonal technique to compute heat and smoke movement, but use sub-models to find out the thermal elements in the detectors to the flow field and heat. Other models that are also applied in fire modelling include egress models, fire endurance models, as well as miscellaneous models[Put13]. Limitations Because of the broad range of computer fire models currently available, different degrees of expertise are required to appropriately apply the models to fire and explosion investigations. Investigators or those who use computer fire models should have in-depth skills and knowledge of the exact assumptions made by the models and the backgrounds of the experimental data and correlations used as inputs. Each fire model has its own specific limitations because of the assumptions made and the experimental techniques employed to derive input data and correlations[Bea09]. It is very possible for different users to produce quite different results. For instance, in a study conducted concerning the probabilistic model in gas and oil industry in Europe, it was determined that risks estimates produced by various users varied by some orders of magnitude[0113]. This means that the experience and knowledge of the investigators is very important to avoid the type of problems which occur in using fire models. One of the limitations of fire models is that they lack the reality of the numerical and theoretical assumptions. The numerical and conceptual assumptions in a computer fire model are just an estimation of the real world. For instance, a control volume model could assume one of several different models for getting into the fire plume. A CFD fire model could assume one of several turbulence models. Because of the variability in assumptions made, it will be possible to use different values for parameters since there are uncertainties in specific applications and the users will still be able to argue that the values they used are reasonable[Bea09]. Another concern with fire models is that they lack fidelity of the numerical solution techniques. Systems of equations are supposed to be solved numerically instead of solving them analytically, with an exception of a simple model having moderately simple equations. Different numerical solution methods have the capability of producing different results. A good of this is provided by the grids used in CFD fire models to create the cells. Results in these models normally rely on the resolution of the grid. Using either a fine grid or a coarse grid will result to significant differences. Additionally, results from a CFD fire model will generally rely on the time step and boundary conditions assumed, as well as other factors[Har14]. It also possible that software may have direct mistakes and this make them inaccurate representation of the numerical solution techniques and model. Errors may occur from mistakes in the computer programs. There is also the likelihood of the physical system being modelled entering a condition that is not suitable for the software. This could be possible due to lack of realism of the numerical and theoretical assumptions made in the fire model. Computer hardware may also have some faults thus contributing negatively to the model results. In most cases, computer hardware is assumed to be very reliable and the likelihood of a computer to make a mistake because of its hardware is highly ignored. Nevertheless, a fault may exist in hardware because of faults in the manufacture of micro-processors, mistakes in the design of micro-processors, or a combination of the two[Bea05]. Finally, a mistake in application is another limitation of the computer fire models. A model user or fire investigator can make an error while putting input into a fire model or while analyzing the output due to a number of reasons. Some of reasons that might contribute to such errors include: a slip in reading output or inserting input; misunderstanding of the design of the computer program or software; and misunderstanding of the fire model or its numerical solution procedures[Bea09]. Inadequate documentation may also contribute to ‘effective errors’. For instance, poor documentation may indicate that a model possess a given ability whereas in reality it does not[Bea09]. Conclusion When carrying out fire investigation, reconstruction, and analysis, it is very crucial to collect a lot of data concerning incident. Computer fire modelling is very powerful when used in fire forensics and when integrated with thorough origin and cause analysis of a fire event; it can be regarded as an extremely powerful and effective tool for fire investigators. Computer fire modelling is capable of exploring multiple fire situations cost effectively and quickly. Computer fire models in use include zone models, CFD models, detector response models, egress models, fire endurance models, as well as miscellaneous models[Put13]. Though these models are very powerful when used in fire forensics, they have some limitations such as lack the reality of the numerical and theoretical assumptions, lack fidelity of the numerical solution techniques, mistakes in software, hardware faults and mistakes in application. References List Len09: , (Lentini, 2009), Sut12: , (Sutura, 2012), NFP11: , (NFPA, 2011), Put13: , (Putorti, 2013), Bea09: , (Beard, 2009), Ole03: , (Olenick & Carpenter, 2003), Har14: , (Harrington Group, 2014), 0113: , (Lentini, 2012), Bea05: , (Beard, 2005), Jan02: , (Janssens, 2002), Har14: , (Harrington Group, 2014; Santos, 2013), Bey08: , (Beyler, et al., 2008), Gor08: , (Gorbett, 2008), NIS11: , (NIST, 2011), Dec09: , (Decker & Ottley, 2009), Read More

Computer fire models are mainly used for two reasons: 1) reconstructing and analyzing the fire, and 2) fire-safe design of a structure[0113]. Reconstructing and analyzing the fire is much easier because there is always additional information accessible such as fire department reports, eyewitness accounts, forensic evidence, among others.[Bea05] In this context, a computer fire model is most commonly used to supplement the additional information in attesting that a specific hypothesis is plausible or not[Jan02].

Computer fire models have been found to be great tools for investigators carrying out post-fire reconstruction and analysis. Modern computer fire models are capable of performing the following: 1) predict fire growth and behavior, 2) evaluate occupant egress, 3) analyze smoke control systems, 4) predict detector actuation time, and 5) provide plausible post-fire timeline of events[Har14]. A zone model can be described as a software or computer program that forecast the impacts of the development of a fire inside an enclosure.

In nearly all applications, the enclosure is not completely enclosed as vents, windows, and doors are normally included in the calculation. Zone models for compartments have been developed for both multi-room and single-room configurations. This model split rooms into one or more zones. The models which are commonly used assume that a room is composed of two zones: an upper layer with heated combustion products, and a lower layer with cooler air moderately free from combustion products[Bey08].

In the two zone model, the fire creates the relation between the lower and upper layers. The two layers are presumed to be well integrated to ensure that the conditions in all the layers are constant. For instance, the predicted temperature in the hot or upper layer is assumed to be same throughout. Most of the models involve provisions for openings to other rooms or to the outside and for heat losses to the ceiling and walls[Gor08]. Model inputs generally include the fire heat release rate, room furnishings characteristics, the locations and sizes of room openings, and the building materials and room dimensions[NIS11].

Outputs of these models normally include combustion gas concentrations, the height of the surface forming a common boundary between the lower and upper layers, lower and upper temperatures, time to flashover, and prediction of fire alarm or sprinkler activation time[NFP11]. There are a number of zone fire models available commercially from several sources globally[Har14]. CFD models split an enclosure or a room into several small 3-Dimensional boxes referred to as cells. The enclosure may include so many cells ranging from centimeters to meters in size.

CFD fire models are established on the fundamental physical principles of momentum, mass, and energy conservation[Gor08]. The computer computes the movement of smoke and heat between the cells over a period of time. At given point in time, there is a possibility of finding the gas concentrations, velocity, and temperature in every cell. Just like the case with zone models, the conditions in every cell are presumed to be constant. However, the large number of cells makes it possible to predict the conditions in the enclosure in much greater detail.

CFD models have the capability of predicting the conditions in very small and very large spaces, in complex multi-room configurations, and in spaces with complex shapes that are not probable with zone models. Because of their complexity, CFD models need a higher level of skills and expertise to operate and are presently operated on costly computer equipment. General purpose CFD fire models are available commercially from several sources globally. Advanced CFD models, particularly designed for fire safety analysis, are currently being developed at the NIST (National Institute of Standards and Technology)[NIS11].

Detector response models primarily forecast the time to activate an initiating device.

Read More