What’s at Risk?

By Jeffrey Lasalle, P.E., Principal, LaSalle Engineering, LLC, Willow Grove, Pa. September 1, 2005

While it’s true that fire protection in educational occupancies is generally well-regulated by applicable building codes— the Life Safety Code and local fire codes—is this really good enough for parents, teachers and administrators? After all, it is a well-established fact that code requirements are a minimum and do not address the specific needs of a building owner or user.

Fortunately, qualitative risk-based approaches to fire safety can be a cost-effective way to address specific risks and improve life safety in educational occupancies beyond the minimum code requirements. However, such solutions require an understanding of the school fire record, knowledge of the population characteristics and the application of a holistic, engineering-based approach to fire safety.

All classrooms not equal

The 2003 International Building Code defines educational occupancies as those that house, for instructional purposes, at least six occupants ranging from 2ty classroom buildings are classified as business occupancies. A similar definition can be found in the 2003 Life Safety Code .

In any case, this classification includes within its scope buildings that have a broad range of occupant ages, capabilities, fire hazards and risks. Clearly a “one-size-fits-all” approach to fire safety is not appropriate.

Reading the numbers

Before beginning a discussion of risk-based design approaches, it’s important to take a look at the numbers. K-12 schools in the United States spend $20 billion annually on construction. Roughly $8 billion of that total goes toward renovations and additions, and 80% of those retrofits are earmarked for providing new space for students. It’s not surprising that budgets for “non-essentials” are shrinking and funding for fire protection is generally limited to meeting minimum code requirements.

But fire protection should be an essential budget item for schools. History shows that educational facilities have had their share of tragedies (see “Fires in Schools,” below). The leading causes of fires in educational occupancies in 2002 were arson and cooking. These were followed by heating equipment, open flames and other heat sources. Bathrooms and locker rooms were the leading areas of origin, with 23% of all structure fires starting in these spaces.

A more detailed review of National Fire Incident Reporting System data over a multi-year period shows the trends to be the same as the single-year data. The most significant and obvious trend is that incendiary fires are a major concern for educational occupancies, especially for middle and high schools. Conversely, the major source of fires in daycare centers is cooking equipment. In any case, this data can be used to inform decisions about the application of fire-safety systems.

In recent years, the fire-protection engineering community has increasingly emphasized performance-based approaches to building fire safety. Advances in the understanding of fire growth, smoke spread, human behavior, structural response to fire and improvements in technology and analytical techniques are factors that have increased the industry’s confidence in performance-based solutions. Building designs that challenge conventional code requirements, unique occupancy fire-safety challenges and the desire for efficient use of available resources all provide the demand for such approaches.

Comprehensive descriptions of performance-based options for fire-safety design can be found in the International Code Council’s Performance Code , Chapter 5 of the Life Safety Code and in the Society of Fire Protection Engineers’ Engineering Guide to Performance-Based Fire Protection . A detailed review is beyond the scope of this article, but a brief description of the process follows.

A classic performance-based approach involves establishing fire-safety goals, translating those goal statements into functional objectives and then developing specific criteria that can be measured and predicted through the use of fire modeling or other techniques. An example is taking a fire-safety goal such as, “Protect the occupants not intimate with initial fire development from injury and loss of life,” and translating that policy statement into a specific performance criterion: “Keep the smoke layer interface 6 ft. above the highest floor level of exit access for a period of 20 minutes or for twice the expected exit time, whichever is greater.”

This performance criterion, variations of which are currently found in building codes, can be used in an analysis to determine which types of anticipated fire events warrant special attention. Fire models can then be used to predict the rate of smoke production, and exit models to predict the time required for occupant egress. Next, appropriate factors need to be included to account for detection time, notification time and delays in occupant reaction time. A factor of safety will generally be applied to account for uncertainty in the model predictions with the purpose of this approach being to demonstrate that the available safe egress time is greater than the required egress time.

Characterizing risk

Obviously, not all fire scenarios can be addressed through design, as this would become cost-prohibitive. To help prioritize the allocation of fire-safety resources, a risk characterization, with characterizations of incident likelihood, can be undertaken to identify those events to be addressed in the design (see the example below).

Probabilities are combined with deterministic predictions of consequences to assign priority to choices of fire-protection solutions. A risk characterization matrix graphically depicts those scenarios requiring a design solution. Although it can take many forms, risk is often defined as the product of the probability of an event and the consequences of the event.

Using this definition, a particular focus would be put on events that have a high probability of occurring, as shown in Table 2 (p. 48), that can result in the greatest amount of damage or injury, i.e., those events with a “value” of risk that lies in the upper-right region of the matrix.

Although this approach may be somewhat involved for standardized use in design of fire safety for educational occupancies, it doesn’t mean that some basic principles of performance- or risk-based design can’t be used in regular design practice. Goals can be established, likely risks addressed in a qualitative way and tools such as the Fire Safety Concepts Tree (NFPA 550) used to provide a framework for decision-making on fire-safety approaches to complement applicable codes.

Practical examples

An examination of the code requirements for educational occupancies reveals that the requirements for certain building features are different (see this story online at www.csemag.com for a table breaking down these requirements). For example, the IBC provides an exception for fire-alarm systems in occupancies under 50, while the Life Safety Code does not. The IBC also has no special exit requirements, where the Life Safety Code has several. This begs the question about which code is more appropriate. But rather than engage in a debate about the propriety of specific code requirements, focusing on the fire risk and the desired performance of the building fire-safety system can inform the decision-making process in a way that will enhance the cost-benefit of fire-safety systems as well as achieve the desired goals.

In terms of the fire-safety concepts tree mentioned above, the primary purpose of a building fire-alarm system is to help safeguard exposed occupants by detecting the need for movement, signaling that need and providing instructions. It would seem reasonable to conclude that there would be no higher goal in an educational occupancy than the safety of the occupants. Is it then unreasonable to question why the base code requirement for fire-alarm systems in educational occupancies is for a manual system? Under the current codes, manual pull stations may be omitted when smoke detectors are provided throughout corridors and certain other areas are equipped with heat detectors.

In a manual fire-alarm system, alarm initiation requires action to be taken by a human in order to begin the evacuation process. In light of the lessons learned from the past, one obvious way to significantly reduce the risk to a school’s occupants is to provide automatic smoke detection. If the expense of complete building smoke detection is excessive, then selective detection in locations with a high risk of fire origin should be pursued. Based on the fire record, bathrooms, locker rooms and other spaces with high concentrations of combustible loading and ignition sources would be logical places to place automatic detection.

If there is a concern about vandalism, why provide manual pull stations as the default fire alarm initiating device? Certainly it is easier to surreptitiously activate a manual pull station than a smoke detector. Also, advances in detection technology are resulting in improvements in immunity against unwanted alarms, so placing detectors in such areas as locker rooms may not be inviting unnecessary evacuations, as feared by some.

Similar arguments could be made for the inclusion of sprinklers in areas other than those mandated by code. As the fire record shows, a base requirement for sprinklers only in areas located below the level of exit discharge may be missing the most common locations of fire origin in educational occupancies.

Significant improvements in the level of safety—without substantial capital investment in either extensive fire modeling or unnecessary equipment—could result from pursuing a discussion with the school board and other stakeholders. Of course, detailed fire modeling could demonstrate the benefits of additional protection equipment, but the point here is that a thought process that focuses on goals and risks can lead to similar conclusions in a shorter period of time.

Finally, applying performance-based design to fire safety should go beyond the fire-protection systems to include consideration of how the building is being used.

Thinking risk

In a nutshell, taking a holistic, risk-based approach to fire-safety performance means more than predicting fire scenarios and consequences. It means thinking in terms of the whole building fire-safety system—its components, their interactions and how they work collectively to protect occupants and the building itself. It means thinking about the practicality, maintainability and long-term reliability of these systems in the context of how the occupants use the building. Such attention to detail can go a long way toward improving the reliability of fire-safety approaches and reducing fires in our schools.

Table 1 — Prioritizing Risk for Resource Allocation

Incident Likelihood Description
Anticipated Incidents that may occur several times over the life of the building; similar events have occurred in the past.
Unlikely Incidents that are not expected to occur such as significant equipment failure and/or negligent action.
Extremely Unlikely Incidents that require more than a single equipment failure or resulting from another unlikely event.
Beyond Extremely Unlikely Incidents so remote that they don’t warrant serious consideration.

Table 2 – Risk Characterization Matrix

Probability of Event P (I) Low Damage or Loss Low to Moderate Damage or Loss Moderate Damage or Loss High Damage or Loss Very High Damage or Loss
Very High 4 3 1 1 1
High 4 3 2 1 1
Moderate 4 4 2 2 1
Low 5 4 3 3 3
Very Low 5 5 4 4 4

Fires in Schools

Unfortunately, the historical record shows that educational facilities have had their share of tragedies. In December of 1958, a fire originating in a basement trash area at Our Lady of the Angels School in Chicago caused the death of 90 students.

While not all school fires are so threatening, the record is full of examples of relatively minor fires causing significant property damage and injuries.

As for more recent incidents, a Dec. 2004 Federal Emergency Management Agency report indicated that the 2002 average dollar loss for a school fire was $15,956, as compared to $21,505 for other nonresidential fires. And a Nov. 2004 NFPA report, Structure Fires in Educational Occupancies , shows a 53% reduction in educational structure fires from 15,100 in 1980 to 7,100 in 2001.

These are encouraging numbers. What is not so encouraging, however, is the NFIRS 2002 finding that school structure fires have an injury rate of 22 per 1,000 fires, as compared to 14.4 per 1,000 fires for nonresidential structures.

An Integrated Approach

When applying a performance-based design process, one should also consider how the building is being used. For example, wear and tear due to a large number of occupants is an obvious issue. Floor coverings, walls and doors all absorb a tremendous amount of abuse over the course of a school year and throughout their useful lives. In this context, it is not uncommon to find doors blocked open to prevent such abuse. However, the presence of unprotected vertical openings is a major contributor in exposing occupants to noxious byproducts even when they are not near a fire.

A simple solution is the inclusion of magnetic door holders tied into the building fire-alarm system that release the doors upon loss of power or activation of the fire-alarm system. While this may seem obvious, it cannot be developed unless someone asks the question and understands the need for an integrated, holistic approach to life safety. This type of approach recognizes that building fire safety depends on the coordination of every building system as well as an understanding of the importance of people in fire safety practice.