The Model of Success

Fire modeling applications are gaining in sophistication and importance thanks to a growing focus on performance-based design

By Craif Hofmeister, P.E., Senior Consulting Engineer, The RJA Group, Raleigh, N.C. October 1, 2001

When the term fire modeling is discussed, most people immediately assume it’s referring to a computer fire-model program. However, the development of a fire model primarily includes a detailed analysis of the proposed facility configuration and potential fire scenarios, as well as research of performance criteria against which the model results will be evaluated. If utilized as part of a performance-based design, much of the analysis portion of the model will be a continuation of the overall established goals and performance criteria.

Computer models are often the calculation basis for a fire model and can be used to predict many time-based quantities such as fire growth rates, temperature, smoke movement, species (air) concentrations and heat transfer. It can also be employed for related events such as sprinkler operation, smoke detector operation, glass breakage and occupant egress. Computer models offer an easier method of performing complex calculations and storing data, although in many cases, hand calculations may be warranted to obtain a specific piece of data or to confirm model results.

Computer models

The most general use of computer modeling is to predict the impact of a fire on the surrounding environment. Computer fire models are generally divided into two categories- field models and zone models -dependent upon the calculation methodology on which they are based.

Both model types are based upon equations that calculate the conservation of mass, momentum and energy. However, zone models typically apply the conservation relationships to three zones within a room including the fire plume, the heated gas upper layer and the cool gas lower layer. The results of the equations are therefore limited to average values for each zone and the model assumes a firm demarcation between the upper layer and lower layer-a rather broad presumption, albeit within the bounds of a reasonable estimation for many applications.

For example, zone models and the associated basic correlation are often used for performance-based atrium smoke management system design and analysis. The model and calculations can provide time-dependent data including the theoretical smoke layer decent, and average values for upper layer temperature, visibility and toxicity concentrations. The model results and methodology have proven reliable for installed systems for most situations.

Field models apply the conservation of mass, momentum and energy equations based upon much smaller, user-defined control volumes or cells and therefore can provide a much higher level of detail for a space. Field models are also known as computational fluid dynamics (CFD) models and typically require significant computer resources as well as extremely detailed data for input.

The good news is that the National Institute of Standards and Technology (NIST) has recently developed a CFD-based fire model-Fire Dynamics Simulator (FDS)-for use on a personal computer. The development and release of FDS has significantly advanced the accuracy, and therefore the extent of data available for use in an alternative design assessment. As an example, FDS allows the user to ‘test’ different scenarios with a relatively high degree of accuracy, such as the impact of heat-detector placement or beam depth on detector response in a warehouse setting. However, the use of FDS is dependent upon the data available for input and still requires significant computer time and resources for a single simulation.

Modeling applications

A common situation that arises during the design process is the desire by an owner or architect to include unique design elements that conflict with applicable code requirements. A total performance-based design approach is-or will be-an option to allow flexibility throughout the facility. More often an alternative approach can be developed specifically for the non-compliant condition.

One recent case study illustrates how modeling can be utilized to tackle unique performance-based design situations. (also see ‘A Unique Tunnel Operation,’ page 56).

In this case, a four-story, historically significant building previously used as an office building was to be renovated into a hotel. The existing structure contained two ornate, open convenience stairwells that connected all four stories of the facility, and did not meet applicable floor-to-floor separation and shaft enclosure requirements. The stairwells were not required egress components of the building and a complete automatic-sprinkler installation was scheduled as part of the renovation.

The first step was to develop assessment goals. In consideration of the historic nature of the building, the interested parties agreed that the analysis would be limited to an assessment of the hazard posed by the open stairwells on non-fire-floor exit-access corridors.

To accomplish this analysis, the designers utilized the FDS model to predict the effects of a credible design fire occurring in one of the lower level guestrooms and estimating the quantity, concentration and therefore the impact of combustion products on occupants of other levels. Because of the open stairwells, analysis of the floor of origin was not included as occupants of the fire floor were deemed at no greater hazard than if the stairwells were nonexistent or protected by shaft enclosures.

A fire occurring in the guestroom located directly across from the open stairwell on the first floor level was chosen as the design fire, as the scenario contains the highest concentration of combustible fuel packages that could credibly expose occupants of the building. A variety of fuel characteristics, including effective heat of combustion and toxicity species yields, were input specifically for cellulosic and textile materials for assessment of toxic conditions throughout the simulation. The room of origin’s door to the hallway was assumed to remain open to allow smoke and hot gas movement. It should be noted that an automatic door closer would likely result in a barrier to issuing smoke into the hallway.

The design fire was allowed to grow until sprinkler operation, at which point the fire’s energy release rate was assumed to remain steady, with no decay, for the remainder of the simulation. The FDS model contains the calculation algorithms to simulate sprinkler operation and the effect on the fire. The input data required to properly model the specific sprinkler operation, spray and characteristics, however, is generally not available until further sprinkler type and model testing is conducted. Regardless, assuming fire control at sprinkler operation with no subsequent decay is a generally accepted conservative assumption, and provided a sufficient level of detail for this simulation. Occupant tenability criteria was separately researched and developed with appropriate safety factors to result in simple pass/fail criteria for temperature and toxicity concentrations.

As expected, the smoke, heat and products of combustion from the fire scenario generally flowed from the fire room into the adjacent corridor and open stairwells, and began filling the upper floor corridors starting with the 4thlevel. Therefore, the highest temperatures and carbon monoxide concentrations of the upper floors occurred on the 4thfloor. The analysis of the developed fire scenario and calculated data from the FDS indicated that the existing open stairwells should not transfer hazardous temperatures and quantities of carbon monoxide to adjacent floor level corridors, thereby allowing the corridors to continue to serve as exit access paths.

This analysis recognized that exiting occupants from adjacent floors may be exposed to elevated temperatures and products of combustion to some extent, but the installed automatic sprinkler system, fire-alarm system and relatively short anticipated exposure time should result in a sufficient level of safety for the occupants.

The next generation

As the performance-based design options and the development of the next generation of fire model computer programs continue to evolve, the opportunity to utilize flexible design approaches will also continue to increase. These tools will likely be utilized to help evaluate and implement fire protection and life-safety strategies for all types and sizes of buildings, in addition to the current equivalency based and unique facility applications.

Additionally, the development of powerful fire modeling tools allow the fire-protection community the opportunity for further insight into the theoretical behavior of fire and smoke and its effect on the building environment. At the same time, the use of the fire modeling tools-especially the advanced models such as FDS-require a thorough understanding of the fire dynamics, fluid flow principles and calculation methods on which they are based.

Fire Modeling and Performance-Based Design

Performance-based fire-safety design methods are becoming recognized as a viable alternative to the traditional prescriptive building code based design, and the use of performance-based methods will likely increase in the future as both designers and authorities having jurisdiction (AHJ) become more experienced with the design and review process. In addition, most building and life-safety codes currently have, or are in the process of developing, a formalized performance-based design option.

An important tool within this performance-based design approach is fire modeling , which is often used to assess the effects of a fire on a facility and its occupants, and subsequently justify the performance of the performance-based design.

Fire modeling has also been used for several years within the prescriptive-code-based design process to evaluate the performance of specific building components and aid in post-fire investigation and reconstruction.

Traditional fire model development has typically provided a performance aspect for designing within the context of the prescriptive requirements, with a goal of providing a level of safety equivalent to the prescriptive requirements-hence the term equivalency. Current performance-based design methods attempt to split from most of the prescriptive requirements and provide a framework in which the building design is evaluated to meet performance goals established specifically for the project. The use of fire modeling, whether as part of a formal performance-based design or as an assessment tool for a specific building component or system, allows the architect, designer or owner flexibility in such areas as configuration, materials and operation.

A Unique Tunnel Operation

Fire protection measures during construction are attracting increasing attention due to several substantial fires throughout the country. However, the loss in most cases is limited to property and equipment. In the recent case of a tunnel boring operation, the life-safety risk to workers was significant due to the small enclosed space, the potential length of the exit travel distance, the limited accessibility for fire-fighting operations and the fact that it takes several years to complete the work. Fire modeling applications were well suited for the fire-safety analysis of this tunnel, which was approximately 670 feet below grade, approximately 5 miles in length near completion and included a single access shaft.

Due to the unique nature of the construction process and the fact that the tunnel would eventually be used as a water viaduct, there were limited applicable fire-protection code requirements. Therefore, a performance-based approach was the only option available. The project took place prior to many of the structured performance-based guidelines, although the process utilized was essentially the same.

Once again, the first step was to determine the goals to which the assessment would be compared. These goals were developed in conjunction with the AHJ, and were limited to life safety such that ‘occupants of the tunnel, who are not intimate with the initial fire development, will survive a fire in the tunnel or shaft and will not be overcome by products of combustion from a fire in the tunnel or shaft.’

The second step was to conduct a qualitative review of the tunnel operations and materials to determine potential severe case fire scenarios for review. The study also needed to research human exposure studies to determine the maximum safe level for toxic products of combustion and heat.

Due to the small confined areas of the tunnel, it was recognized that workers would be exposed to a certain amount of products of combustion in a fire scenario and therefore, two levels of tenability criteria were developed based upon the time and level of exposure. These were defined as ‘hazardous’ and ‘critical’ conditions, for which exposure to hazardous conditions for a short duration of time was acceptable while exposure to critical conditions for any time period was unacceptable.

The design fire scenarios were developed for each area of the tunnel and shaft and included a tunnel boring machine fire, diesel locomotive fires, conveyor-belt fires and general combustible (trash) fires. The fire scenarios were input into the computer model MFIRE, which provided time dependant locations of temperature and concentrations of carbon monoxide and carbon dioxide.

MFIRE was originally developed by the Bureau of Mines for ventilation in complicated mine shafts, and the program is essentially a modified zone model that allows the user to input outside influences such as a growing fire. Although not developed as a fire model, the program provides a good estimation of the development and movement of heat and toxic conditions.

The figure to the left shows a summary of the results of one of the fire scenarios. In this scenario, a fire involving a diesel locomotive with a simulated fuel system suppression system was reviewed. The results indicated that although workers in the immediate fire area could be exposed to hazardous conditions, the fire growth would be limited and the workers should be able to exit the area without being exposed to critical life-safety conditions.

Analysis of each of the fire scenarios resulted in recommendations-such as the inclusion of fuel-suppression systems for the diesel locomotives and water-spray systems for specific locations on the conveyor belts-to meet the life-safety objectives.

It should be noted that an important, and sometimes difficult portion of the analysis is clearly depicting the results of the analysis for review with the AHJ and interested parties. Most zone-type computer models are limited in calculation output, which is typically in the form of tabular data.