Sprinklers R.LAlpert And Cfast Zone Modeling - Report Example

CFAST AND SPRINKLER ZONING MODEL Cfast and Sprinkler systems Customer Inserts His/Her Name Customer Inserts Grade Course Customer Inserts Tutor’s Name 14, 04, 2012 R.I Alpert Alpert’s contribution to the field of science through his equations has helped a great deal in the calculation of certain parameter like the determination of the upper layer temperatures, ceiling jet temperatures, and actuation time for the sprinklers. Alpert was among those who developed the earliest models for determining various parameters related to the fire. For instance, his first model was based on the ceiling jet flow that was released by stable fires for an uncovered ceiling; a ceiling on which the gas layer would not easily form. Therefore the temperatures are assumed to be at constant at the value obtained for r/H=0.18 where r/H is less than 0.18. He applied the relationships for plumes that emanate from a point source in order standardize the ceiling jet variables. As such, the normalized velocity relationship is provided as: In addition, Alpert derived simpler correlations for quantifying maximum gas temperatures as well as velocities at particular positions in the in the ceiling jet flows generated by stable fires. Alpert proved that the maximum gas temperature (Tmax) closer to the ceiling at a particular radial position r (on condition that r is less than 0.18H) can be related to the rate of the heat equation release (QckW) by the following steady state equations: This equation is simplified when H is used to represent the full length scale and thus give a more common equation: Where H is taken to be more than or equal to 0.18H, that is, within the area where plume impinges on the ceiling: Where; In the above formula, T is in ⁰C whereas U(Velocity) is in m/s The experiments performed by Alpert pertaining to the various parameters and correlations were based on the assumption of ambient temperature of 293K, normal atmospheric pressure, and the convective energy release that is equivalent to the total energy release rate, Qc = Q. th results obtained from the experiment are as shown with the dotted lines on the graph below Most of the model simulations and the experimental data calculations are based on the equations developed by Alpert. CFAST Zoning Model Zoning models and systems are useful in giving descriptions of the development of fires and smoke movements. These are mostly applied in the advanced models for the design and fire safety assessments in building and other engineering structures. There has been significant effort to link experiments with the fire models in the recent past. However, less has so far been done in improving the current fire codes. The incorporation of performance based building codes has led to the need for mathematical models. It is therefore imperative that those involved in design of building to have a better understanding of the useful codes as well as the limits within which to operate. CFAST It is a simulation model developed from the package HAZARD 1 designed and developed by NIST. This model is capable of estimating the compartments of temperature and the species concentration in an appliance meant for developing the input deck through a chain of options. Cplot is a data FAST and the CCFM.VENTST. CFAST is a zoning model for solving both mass and energy conservation equation that are very useful in the planning and dealing with fires. In this context, the growth aspect of the model is not included because of its redundancy in fire application. The particular fire-related physics that incorporates hot gas layers, fire plums, jetting systems for doors, transferring radioactive heats and the ignition of secondary fires have been included in this simulation model. There is also a provision for the ventilation system, which includes ducts, vent and the fans. Application limitations of CFAST model Generation of the chemical terms by the model are less and therefore unreliable. Its transport abilities are also minimal with no ability of modelling radiological sources of energy. Strengths The model can handle fires from many sources simultaneously It has a provision for the installation of ventilation systems It has a large database for entering details for the materials The database for the materials can be customized to suit a specific scenario, especially simulations for the HCL depositions Limitations Needs priori specifications relating to the specific characteristics of a fire The model is incapable of carrying out operations pertaining to the radiological sources and terminologies Because it utilizes the conservation equations in its operations, it is thus limited by the assumptions applied in these equations used in the zoning model. It does not take care of simulations related to the mitigation operations It is not reliable because it cannot manipulate data to give all the required variables It does not incorporate momentum in the simulations using the energy conservation equations There is restricted access to the simulated parameters when applying the model Input data requirements The model has a CFA ST input deck which is smaller and about three pages large hence easier to produce. To generate a deck, an ASCII text editor or Cedit, which are menu driven front end utilities. Several variables need to be fed into the system to produce the desired output. These include; the geometrical appearance of the room, the behavioural growth of fire as a function of time, the geometry of the ventilation system and the temporary behavioural patterns of fires, constraints related to fires like oxygen limits and species ratios. In addition, there is need for entry of information regarding the characteristics materials used for walls, floors, and the ceilings and the flammable materials. Output CFAST is capable of generating tubular graphics using the ASCII application and HPGL graphic outputs. Several outputs can be generated by this model including layer interface heights, higher and lower temperatures, temperatures on the wall surfaces, wall surfaces, ceilings, and floor surfaces. In addition, other outputs include pressures, locations of fires, optical densities, several flow rates, concentrations of species. The functioning of the CFAST system CFAST system can be run on a local machine or connected via the internet, thus representing local simulations and remote simulations respectively. Remote simulation is usually faster and reliable because the computational load of the fire simulation, user tracking, and rendering the virtual environment are dispensed among many machines connected to the same network. It is more advantageous to use a remote computer when applying the CFAST model than using a personal machine; usually a remote super computer. This therefore gives more computational advantages to the users. When using the local simulation system, there is need for running the CFAST server version of the software on the local machine to make it a default a program. However, this server will remain in the inactive mode until prompted to carry out a simulation operation by the local machine. When using the simulation server communicates automatically as long as the server is left running. After the CFAST is run on the local machine when using the local simulation model or start up communication with the server located on the remote machine. A full CFAST data file can be activated under certain conditions. There is need for the virtual environment which includes the building model and the data for mapping. This is important to allow for identification, numbering, and measurement of the rooms as part of the CFAST entries and facilitate the conversion of the portals into CFAST HVENT data statements. The objects for detailing and mapping within the room allows for all objects to be defined in terms of the CFAST objects. The information provided from the fire scenario is then used to complete the necessary data required for the running of CFAST. The transfer of data is done in such a manner that there is no intervention from the part of the user. There is a special data simulator protocol for visualizing data results that are channelled from the simulator into the user’s terminal within a shorter time. To control the tendency by CFAST from displaying busty results, the simulator has been designed in such a way that there is an allowance of approximately 10 minutes. The information displayed to the user is a typical representative of what the person inside the burning structure would see within this time limit (McGrattan, Baum, Rehm, Hamins, Forney, Floyd, Hostikka, & Prasad 2002, 6783). Comparison between the CFAST model and the experiments There has been significant efforts to bring harmony between the results obtained using the simulations from the CFAST model and those obtained from the real experiments. As such, it helps in determining the accuracy of the model and the range of the applicability of the model. Laboratory experiments conducted in the recent time have shown that there is a closer relationship between the results obtained from the CFAST simulations and those obtained from these experiments. In one of the experiments conducted by Miler and Rocket, the HAVARD V was applied in the modelling of eight room fires that was supplied with enough tools. Another experiment was conducted by Levine and Nelson sought to draw comparisons between the CFAST simulations and the experimental data obtained from the various experiments based on the HAVARD VI multi room model. The results obtained accurately envisaged the remote room CO accumulation and the pre-flashover temperatures. However, the post-flashover temperatures were not properly envisaged. The consistency of the results obtained was appreciable and within the allowable deviations. The data obtained from this experiment was take and simulated using the different models including CCFM, FIRST, FPETOOL, and CFAST. The results obtained from the comparison showed significant similarities and there was consistency in the results obtained (Lattimer 2002, p. 262). Another experiment was conducted using the aforementioned models in an aircraft hangar. There was high accuracy in the results obtained for all the models to fires less than 4 MW, but however, there was inconsistency for all the models when applied to the 36 MW fires. In another set up, Beard used three documented experimental data meant to give an evaluation of four fire simulations including ASSET, CFAST, FIRST, and JASMINE. The evaluation involved various variables both qualitative and quantitative. The results obtained showed a high level of accuracy and consistency. In another set up to compare the CFAST model to real experimental set up results, the predictions from the model were ranging from a lower percentage to a factor of two or three depending on the nature of the experiment. The disparities in the results obtained were attributed to the inconsistencies of errors from both the CFAST simulations and the experimental data. The time taken to reach the peak values and the temperatures for the layers showed greater consistency as compared to the gas species that showed the highest inconsistency. From the results obtained in this experiment, the inconsistencies noted were attributed to the errors resulting from the simulations by the CFAST model and the experimental constraints. A comparison made between the values from the simulations by the CFAST model and the data obtained from a post-flashover ship compartment fires. This also included the transfer of heat within the compartment as a variable for comparison. There was accuracy from the results obtained related to the fire compartments, decks, and also the compartments located above the deck. However, there was notable exaggeration in the prediction by CFAST model concerning the compartment5s located directly above the fire compartment. Dempsey, Pagni, and Williamson came up with a sequence of comparisons between the CFAST model and the experimental results obtained in one room gas burner fire experiment. The results showed that the temperatures of the hot gas layers were approximately 800⁰C. it was also surfaced from the simulations that CFAST estimations of the room hot gas layer were exaggerated and ranged from 150⁰C to 260⁰C (Hoover & Tatem, 2000 p. 63). An actual experiment was conducted in a compartment measuring 3.4m in width, 3.3m in depth, and 3.05m in height. The tests conducted in this experiment were done in two series. Series 1 utilized the natural ventilation system whereas series 2 made use of the artificial ventilation system. The various fuels utilized in this experiment included diesel pan fires, wood cribs, and the polyurethane slabs. Natural ventilation test They produced two layer systems that made it possible to determine the interface heights. The resultant temperatures from the thermocouple in the arrays placed over the interface pointy were moderated for the hot layer temperatures as well as under the interface point for the cold layer temperatures (Hoover 2008, p. 129). Forced ventilation system It was done in a similar way as the natural ventilation system except for the bidirectional probes, which were placed at the central point between the supply duct and the exhaust vent. Thermocouples were also located on this point. However, this experiment never produced the two layer system as it was the case with the natural ventilation system. As such, calculations of the upper layer temperatures were done by getting the average of the temperatures in the whole compartment height. In addition, the interior ceiling temperatures were arrived at by using the average of the four points that had been used in the measurement of this parameter (McGrattan et al, 2002). Simulation of the data quantities by the zone model for variables like the hot layer temperature, boundary temperatures, interface heights, the rate of heat release, among other variables require more than one instrument to measure. Therefore, there is need for processing of the various quantities measured by the different instruments to come up with expected results. Other zoning quantities are calculated by getting the averages of the multiple quantities measured for these variables such as the thermocouple temperatures. This implies that the results obtained from such simulations are prone to errors and hence have inaccuracies that are mainly dependent on the number of instruments used in the simulation of the data. CFAST encountered a problem in displaying the interface locations for the polyurethane foam as the decaying process proceeded. This was also seen for all the other fires as the fuels were approaching depletion. There was a limitation in the supply of oxygen and therefore affecting the rate of combustion of fuels (Peacock, Jones, Reneke, & Forney 2005, p. 143). Experiments involving the use of polyurethane and the wood cribs showed a greater matching from the temperature predictions by the CFAST model simulations. However, the other variable had higher inconsistencies, which are largely attributed to the difficulties related to the noisy data. In addition, therefore were consistent predictions for the diesel pan fires. Natural ventilation recorded the highest prediction as compared to the forced ventilation. There was over-prediction in the ceiling temperatures whereas the compartment heat loss was underestimated by the simulation model. There was also over prediction of the hot gas layer temperatures. On the overall, the upper layer temperatures were more consistent in the forced ventilation compared to the natural ventilation. There was an inconsistency in the trend followed by the ceiling temperature estimates for the upper layer thus implying that the boundary settings for the side that was not exposed had little or no heat loss at all (McGrattan & Bouldin 2004, p. 1001). SPRINKLERS The building code requires that sprinklers be installed on the buildings as part of the safety measures against fires. Sprinkler systems offer active fire protection to a building or any engineering structure. It comprises of water systems that offers the required pressure within the distribution piping system. The sprinklers are usually fitted into the piping system. Their uses vary a great deal from household purposes to industrial purposes as well huge commercial structures. Installation of the sprinklers is not only regulated by the building codes but also the insurance requirements as a necessary strategy for lowering the risks that may emerge as a result of a fire outbreak. The use of different materials in buildings, both residential and commercial call for associated safety standards to be incorporated because most these material are highly flammable Types of sprinkler systems L 1 Sprinkler systems In this system, all the areas of the building are fitted with the sprinklers. However, it is a common practice not to fit sprinklers in such areas as the lavatories including the water closets. Empty spaces with a height of less than 800mm are also not fitted with the sprinkler because this would limit proper functioning of the sprinklers. Lavatories containing electric hand driers are usually fitted with the sprinklers because they stand a higher risk of fire (Hoover 2008, p. 127). L 2 Sprinkler System In this system, sprinklers are only fitted in those areas of a building that are deemed vulnerable to fire outbreaks. Such areas include the sleeping areas, accommodation areas, store rooms, kitchens, and plant rooms. This system is also suitable for the areas or building occupied by old people or where occupants are not familiar enough with the building (Hu, Li, Huo, & Wang, 2005, p. 49). L3 Sprinkler systems In this system, emphasis is given to the escape routes like the corridors, staircases, and the lift shafts. P1 sprinkler system It functions like the L1 system and therefore covers the entire building with the exception of the lavatories, water closets, and the open spaces with a height less than 800mm (McGrattan 2004, p. 1001). P2 sprinkler system In this system, sprinklers are selectively fitted in the areas with a higher risk of vulnerability. Experiment using sprinklers A 12 ×7×3 m 3 airtight room was used for the experiment that sought to measure the buoyancy and sprinkler activation time. It involved the use of a newspaper as the material for burning. The inlet and the outlet ducts were designed in such a way as to allow for the entry of fresh air from the environment as well as exit of smoke respectively. Thermocouple trees were put in the test compartment to measure the smoke temperatures inside the compartment in the locations T1, T2, T3, and T3. A flow meter was incorporated in the design to measure the flow rate and the pressure of water in the pipes. To emphasize the understanding of the fire model and be able to make comprehensive comparisons from the experimental data obtained when using the sprinklers and the data obtained when using the sprinklers. Results The experiment involved the use of automatic sprinkler system in all tests conducted in this experiment. The temperature measured by the sprinkler system for the test 1, 2, and 3 indicated 74⁰C whereas the Response Time Index (RTI) for the respective tests was measured to be 138(ms)1/2. The temperature for the Test 4 indicated 68⁰C whereas the RTI for the same showed 28 (ms)1/2. However, Test 5 temperature results were similar to that obtained in the first three tests whereas the RTI value correspondent that for Test 4. Tests 1 and 3 utilized nominal 13.5 orifice sprinklers, which supplied water at a discharge density of 12 mm per minute. Test 2 applied nominal 12.3 orifice sprinklers with a discharge rate of 6mm per minute. Test 4, applied nominal 16.3 mm orifice Quick Response sprinklers that supplied water at the rate of 18mm per minute. Test 5 used nominal 25.4 Quick response Orifice sprinklers supplying water at the rate of 18 mm per minute. The spacing applied from one sprinkler to the other was same for the first four tests and included a distance of 3m by 3m whereas Test 5 applied a distance of 6.1m by 6.1m. In this experiment, a fire simulation model was used to predict results and help draw comparisons between the experimental data obtained when using the sprinklers and that obtained when using the simulation model. Two FMRC standard complete igniters made of cellucotton rolling and each wetted 237ml gasoline and covered by a plastic were used as the primary ignition sources. The ignition positioning for each sprinkler was centrally placed under each sprinkler (Hoover 2008, p. 126). Discussion The discrepancies between the experimental data and the one obtained using the simulation model ranged from 3% to 7%. The simulation model used in this case was FDS and the predictions from the model were compared to the results from the experiment conducted when using the sprinklers. These comparisons are as shown in the table below: Conclusions From the comparisons made between the CFAST simulation model and the experimental data. An assessment into these blind estimates linked with the data from experiment is a sure implication that CFAST is reliable and therefore able to produce better estimates for gas layer temperatures, interface heights, and boundary temperatures. The comparisons made in this paper have illustrated the importance of heat release rate predictions and the proficient decision when selecting the input data and the assessment outcomes from the CFAST model. It is also imperative to state that there is need for more advanced approaches for the conversion of the test data obtained from the experiments into the model entries and the zone model parameters. It is also revealed from these comparisons that the CFAST simulations consistently overestimated the upper layer temperatures than the experimental values obtained although the deviation never exceeded 50⁰C. This inconsistency can be attributed partly to the approach used in calculating the heat losses via the compartment boundaries. The observations and predictions portrayed in this study are consistent with the other publications on the test data comparisons with the CFAST model simulations (Hoover, & Tatem 2000, p. 279). Comparisons made between the experimental results and the simulations from the models indicate that sprinklers are effective and they give reliable data that is consistent with the simulation models. However, there is need for better methods that will ensure errors are limited and give more accurate results. The correlations that were developed by Alpert generated predictions that agreed with the measurements where the hot layers could not form. On formation of the hot layers, these correlations significantly underestimated the plume centreline temperature and ceiling jet temperatures. In actual compartment fires with less ventilation, the models generated basing on these correlations is likely to lead to inaccurate results. The radial reliance of the ceiling jet temperature can be simulated using the Alpert’s correlations, which over predicts the ceiling jet temperature decrease with the distance from the plume canter when a hot layer is presently there. References Hoover, J.B. 2008. Application of CFAST Zone Model to ships – Fire specification Parameters, Journal of Fire Protection Engineering, 18(199), 123-132 Hu, L.H., Li, Y.Z., Huo, R. And Wang, H.B. 2005. ‘‘Smoke Filling Simulation in a Boarding-Arrival Passage of an Airport Terminal Using Multicell Concept,’’ Journal of Fire Science, Vol. 23, pp. 31–53. Hoover, J.B. & Tatem, P.A. 2000. ‘‘Application of CFAST to Shipboard Fire Modeling Development of the Fire Specification,’’ Memo Report NRL/MR/6180–00-8466, Naval Research Laboratory, Washington, DC, USA, Peacock, R.D., Jones, W.W., Reneke, P.A. & Forney, G.P. 2005. ‘‘CFAST — Consolidated Model of Fire Growth and Smoke Transport (Version 6) User’s Guide, Special Publication 1041, National Institute of Standards and Technology,’’ Gaithersburg, MD, USA, Jones, W.W., Peacock, R.D., Forney, G.P. And Reneke, P. 2005. ‘‘CFAST — Consolidated Model of Fire Growth and Smoke Transport (Version 6) Technical Reference Guide, Special Publication 1026, National Institute of Standards and Technology,’’ Gaithersburg, MD, USA Lattimer, B.Y. 2002. “Heat Fluxes from Fire to Surfaces”, SFPE Handbook of Fire Protection Engineering, 3rd Ed. p. 2-269. McGrattan, K.B. and Bouldin, C. 2004. “Simulating the Fires in the World Trade Center” Conference Proceedings of the 10th Interflam, Vol. 2, pp. 999-1008, Interscience Communications Ltd, UK McGrattan, K.B., Baum, H.R., Rehm, R.G., Hamins, A., Forney, G.P., Floyd, J.E., Hostikka, S. and Prasad, K. 2002. “Fire Dynamics Simulator (Version 3)–User’s Guide” National Institute of Standards and Technology Report NISTIR 6783 Read More