The issue of what constitutes a balanced design approach to fire protection has been the subject of ongoing discussions within the processes used to develop prescriptive code requirements. Unfortunately, the discussion is often made more challenging by the fact that the concept of balanced design has not been well defined, and there is no consensus on how to evaluate when "balance" has been achieved. The purpose of this article is to propose a methodology that may be used when considering the concept of balanced design.

Definition
It is difficult to discuss a concept when all parties involved may not be using the same definition. Balance can be defined as a harmonious arrangement or relation of parts or elements within a whole. Fire protection engineers understand that a relationship exists between the various fire protection design features and that they impact the overall level of safety provided within a building. Therefore, the difficulty in achieving a consensus regarding a balanced design approach to fire protection rests solely on the concept of a harmonious arrangement. For some, a harmonious arrangement involves including all possible fire protection features and systems. Obviously, this can lead to unnecessary redundancies in protection.

Alternatively, the approach to defining balanced design should involve an arrangement or relation of the various fire protection features and systems to achieve an acceptable level of safety. The concept of considering multiple safeguards is contained in NFPA 101, Life Safety Code1, as follows:

4.5.1 Multiple Safeguards
The design of every building or structure intended for human occupancy shall be such that reliance for safety to life does not depend solely on any single safeguard. An additional safeguard(s) shall be provided for life safety in case any single safeguard is ineffective due to inappropriate human actions or system failure.

With this concept as a basis, various methodologies can be explored to evaluate whether a design or a code results in an arrangement or relation of the various fire protection features and systems to achieve an acceptable level of safety.

Fire Safety Concept Tree
NFPA 550, Guide to the Fire Safety Concepts Tree,2 can be used to analyze the impact of various fire safety concepts and to identify gaps and redundancies in fire protection strategies. Although NFPA 550 addresses fire prevention issues as well as managing the fire, this article will focus on strategies related to managing the fire impact (see Figure 1). As illustrated in Figure 2, the fire impact can be managed by either managing the fire or managing the exposed. The fact that the diagram contains an "or" gate indicates that one need only manage the fire or manage the exposed. However, the use of the "or" gate also assumes 100 percent effectiveness and reliability associated with the chosen strategy, assuming that the fire safety objectives are to be achieved in all reasonably credible fire scenarios.

Figure 3 indicates that the fire can be managed by controlling the combustion process, suppressing the fire, or controlling the fire by construction (often referred to as containment or compartmentation). It should be noted that compartmentation as a fire protection strategy also appears in the sections of the Fire Safety Concepts Tree addressing managing the exposed.

Much has been written regarding the reliability of automatic sprinkler systems.3,4,5,6 The reliability of other fire protection features and systems is not as well documented. The code developer or design professional must determine the extent to which the reliability of the fire protection feature or system must be evaluated. The ICC Performance Code for Buildings and Facilities indicates that the design needs to include reliability to ensure that the fire safety objectives are met.7

In a similar manner, NFPA 101and NFPA 50008 contain a required design scenario in which the failure of individual fire protection features and systems must be evaluated when doing a performance-based design. The use of design fire scenarios will be discussed later in this article and is mentioned here only to indicate that one needs to consider the reliability of all fire protection features and systems, and determine if the reliability is consistent with the stated fire safety objectives.

As such, the real issue related to using the Fire Safety Concepts Tree to evaluate balanced design concepts for fire protection is one of reliability. By evaluating the various "or" gates, the Fire Safety Concepts Tree can identify what redundant features will result in a higher probability of achieving the desired performance objectives. Unfortunately, lacking the reliability information, it cannot be used effectively in a cost benefit analysis.

For example, a fire protection strategy to achieve the performance objectives may be to provide automatic sprinkler protection throughout the building. If the automatic sprinkler system is designed to suppress the fire ( instead of control mode) and the suppression will occur within the performance limits necessary to achieve the fire safety objectives, the Fire Safety Concepts Tree would indicate that the sprinkler system is all that is required, assuming the reliability of the system is also within the desired fire safety objectives. However, if the suppression system is not designed for suppression, is not able to suppress the fire before other performance limits are reached, or the reliability of the system is not consistent with the stated fire safety objectives, some additional protection is necessary. The Fire Safety Concepts Tree indicates that using alternative protection strategies involving controlling the combustion process or ignition would assist in meeting the desired fire safety objectives by either preventing the fire or controlling the fire so that the system can suppress the fire before performance limits are achieved. Alternatively, the effects of the fire could be contained using effective compartmentation and providing structural stability, which would not only assist in managing the fire but would also assist in limiting the exposed and safeguarding those exposed. The two approaches are substantially different in nature and could be evaluated as trial designs.

Developers of code requirements can use the Fire Safety Concepts Tree to evaluate whether the redundant features often required by prescriptive codes truly work together to assist in achieving the desire fire safety objectives. Although the evaluation would be qualitative in nature without good reliability data, the Fire Safety Concepts Tree might also be used to determine if too many redundancies are being required. The Fire Safety Concepts Tree might be further used to determine what redundant features provide input into various fire protection strategies as illustrated in the previous example.

Using Design Fire Scenarios
Since sufficient data to quantify the reliability of various fire protection features and systems do not exist, an alternative approach would be to perform a "what-if" analysis. With the "what-if" analysis, various failure scenarios can be used to determine the consequence of the failure. One would not expect the level of protection to be the same as if the protection feature performed as expected. The purpose of this analysis would be to ensure that the outcome resulting from the failure would not be unacceptable.

This concept is supported in the various performance codes in the U.S. The ICC Performance Code for Buildings and Facilitiescontains a matrix indicating the maximum level of damage to be tolerated as a function of the performance group and the magnitude of the design event.7 One would expect the magnitude of the design event to be greater when assuming a failure event, and furthermore, one would hope that the probability of the event would be less than the probability of successful performance of the fire protection feature. As such, Table 303.3 (see Figure 4) permits a greater level of tolerable damage.

Likewise, the design fire scenarios that involve failure events contained in NFPA 5000will most likely not result in the same level of safety as one would expect to be achieved in the other design fire scenarios and as defined by the stakeholders' goals and objectives. The associated annex note indicates that acceptable performance should not be defined as meeting all of the stated fire safety goals and objectives.8 When determining if the resulting-performance is acceptable, the annex note further states that consideration can be given to the level of safety provided and the probability of the failure event occurring.

It should also be noted that a design fire scenario involving a failure need not be analyzed when the reliability of the fire protection feature or system is acceptable and when the level of safety provided, assuming the fire protection feature or system is not provided, is acceptable. In other words, if the other design fire scenarios assume that automatic sprinkler protection is not provided and the performance objectives are met, the design professional does not have to evaluate the level of safety provided should the automatic sprinkler system fail to control the fire even when one is provided.

Examples
The first example of the methodology proposed herein would involve an atrium smoke management system designed to maintain the smoke layer interface level in an atrium above a specified level for a specified period of time. Most of the design fire scenarios would typically assume that the smoke detection provided in the atrium will actuate the smoke management system. Using the "what-if" analysis combined with the Fire Safety Concepts Tree, a design fire scenario might include the failure of the fire alarm control panel to perform as expected. As such, not only would there be a delay in actuation of the smoke management system, there might also be a delay in notifying the occupants of the building to evacuate and in notifying the fire department of the fire condition. Using the Fire Safety Concepts Tree, one might also note that the failure of the fire alarm control panel could result in other failures, including automatic closing doors not closing and therefore not safeguarding the exposed as intended with the original design.

The analysis of this scenario might indicate that the required safe egress time would be longer due to the delay in occupant notification, that smoke may spread beyond the atrium, and that the smoke layer might descend below the intended level thereby reducing the available safe egress time. As such, the level of safety may not be as high as that provided if the fire alarm control panel functioned properly. The purpose of the analysis would be to determine that the outcome of the event is still acceptable given consideration to the probability that the fire alarm control panel might fail.

Another example of the methodology proposed herein would involve the failure of an automatic sprinkler system to operate during a fire in an office building. Based upon typical prescriptive code requirements in the U.S., corridor walls need not have a fire-resistance rating, travel distance may be increased, minimum performance of interior finish materials may be decreased, structural fire-resistance ratings may be decreased, egress capacity may be decreased, and a manual fire alarm system need not be provided. The analysis of this scenario might indicate that required safe egress time would be longer, and that since the fire will not be controlled by the sprinkler system, the production of heat and products of combustion and the spread thereof will be greater, thereby reducing the available safe egress time. Again, the purpose of the analysis would be to determine that the outcome of the event is still acceptable given consideration to the probability that the automatic sprinkler system might fail.

Although the two examples contained herein involve the failure of a fire alarm control panel and an automatic sprinkler system, it should be noted that all fire protection features and systems should be evaluated using the "what-if" analysis. This approach can be used to evaluate the reliability of any specific trial design as well as the adequacy of specific code provisions or proposed changes to prescriptive codes.

Concluding Remarks
Although this article has focused on the concept of balanced design as being discussed in the United States, the author is aware of similar discussions occurring throughout the world. The examples cited herein are based upon the author's experience with prescriptive and performance codes, and the code development processes within the United States. Again, while some of the specific details of the code requirements may vary from country to country, the issues and the methodologies contained herein should be valid throughout the world.

A true risk analysis would involve the use of accurate probability data including the reliability of the various fire protection features and systems. Although attempts have been made to quantify the reliability of various fire protection features and systems, validated reliability data are not available for most fire protection features and systems. Until the time when valid reliability data become available for all fire protection features and systems, the methodologies contained herein may be the best methods available to deal with the concept of a balanced design approach to fire protection.

Willian E. Koffel, is with Koffel Associates

References

  1. NFPA 101, Life Safety Code, National Fire Protection Association, Quincy, MA, 2005.
  2. NFPA 550, Guide to the Fire Safety Concepts Tree, National Fire Protection Association, Quincy, MA, 2002.
  3. Budnick, E., P.E., "Automatic Sprinkler System Reliability," Fire Protection Engineering, Winter 2001.
  4. Kelly, K., "Trade Ups," Sprinkler Quarterly, Summer 2003.
  5. Koffel, W., Reliability of Automatic Sprinkler Systems, posted at www.fcia.org.
  6. Rohr, K., and Hall, J. Jr., "U.S. Experience With Sprinklers and Other Fire Extinguishing Equipment," National Fire Protection Association, August 2005.
  7. ICC Performance Code for Buildings and Facilities, Country Club Hills, IL, International Code Council, 2003.
  8. NFPA 5000, Building Construction and Safety Code, National Fire Protection Association, Quincy, MA, 2006.

* Reprinted with permission from NFPA 550-2002, Fire Safety Concepts Tree, Copyright 2002, National Fire Protection Association, Quincy, MA 02269. This reprinted material is not the complete and official position of the NFPA on the referenced subject, which is represented only by the standard in its entirety.