lowdown on high-rise fires, The

The fire protection community has worried about fire risk in high-rise buildings for as long as there have been high-rise buildings. Our codes and standards address that concern with special requirements.

When you think of high-rise buildings, do you think first of the Empire State Building or the Sears Tower-buildings with at least 30 or 40 floors? Many people do. Many people also think of high-rises as hotels and office buildings.

So it may come as something of a surprise that, for the purposes of writing fire codes or classifying fires, any building at least seven stories high is defined as a high-rise, although such buildings may also be defined in terms of linear height rather than stories. For example, the 1997 edition of NFPA 101, Life Safety Code, defines a high-rise as a building measuring more than 75 feet from the lowest level of fire department vehicle access to the highest occupiable floor.

In terms of reported fires, four property classes, not just offices and hotels and motels, dominate the statistics. Apartment buildings and hospitals, as well as other facilities that care for the sick, are also represented. In 1995 alone, 10,000 structure fires were reported in U.S. high-rise buildings in these four property classes combined. Together, these fires killed 55 people, injured 688, and caused $44.5 million in direct property damage (see Table 1).1

This suggests that the total number of high-rise fires in the United States in 1995-not just those in the four main property classes-was actually 12,900 to 13,800, down from 22,500 to 25,000 in 1985. Associated civilian deaths were 55 to 60 in 1995, down from 70 to 80 in 1985. Civilian injuries were 730 to 950 in 1995, down from 780 to 1,360 in 1985. And direct property damage was estimated at $60 to 570 million in 1995, down from S70 to 5135 million in 1985.

Even expressed as ranges, these statistics show that the high-rise fire problem is declining, and damages would show a steeper decline if adjusted for inflation. The four property classes selected account for most of the estimated high-rise fires and associated losses. Other types of high-rises with substantial direct losses include agricultural storage facilities, metal or metal product manufacturing facilities, energy production facilities, and buildings that are vacant or under construction.

Note that storage, manufacturing, and industrial facilities may be as tall as high-rise buildings without having a lot of floors. Since we define "high-rise" in terms of stories, not height, tall buildings with few floors are supposed to be excluded. However, the large number of reported high-rise fires in agricultural storage properties, such as silos and grain elevators, suggests one of two things. Either these tall buildings typically have enough floors to qualify as high-rises, even if most of their volume consists of a tall room without floors, or some fire officers are coding on height equivalents, even if the scale calls for stories. This is worth keeping in mind when you interpret these statistics.

It's also worth noting that most high-rise fires occur in apartment buildings. This may seem surprising, but it shouldn't. Homes dominate the U.S. fire problem so completely that it's always a good bet that the largest share of any newly examined fire problem, unless it's one that can't possibly occur in a home, will be in homes.

We may not realize how important homes are in the high-rise fire problem because of the way newspapers cover such fires. When a fire occurs in a high-rise office building, the fact that it's a high-rise is usually cited prominently, often more prominently than the fact that it's an office building. For a high-rise hotel or motel, the high-rise reference is still likely to be prominent. For a high-rise apartment building or hospital, however, the fact that it's a high-rise tends to be downplayed or omitted altogether. A tall building may be a tall building no matter what's going on inside, but when people think of a "towering inferno," they typically think of an office building or a hotel.

In fact, nearly every identifiable occupancy type has, at one time or another, shown up in the national fire incident databases as the occupancy of origin in a high-rise fire-which greatly complicates estimating the total size of the fire problem in high-rise buildings. It's probably not unusual for fires to be coded in terms of the occupancy of origin, much as is done in a shopping mall. This would partly explain fires coded as having occurred in "high-rise restaurants," for example.

Many fires are coded in terms of both property use and the complex of which the property was a part. A one-time analysis of the code for complexes indicated that nearly half the "high-rise restaurants" were, in fact, restaurants in high-rise medical, residential, or office buildings. Unfortunately, there's no simple way to regroup fires like these into fewer categories based on the building's primary use, given the available codes. Even if there were, you'd still have buildings with whole floors of offices, whole floors of apartments, whole floors of stores, whole floors of hotel rooms-and no one property use that stands out enough to be called primary.

Roughly 1 in every 12 reported apartment building fires occurred in a high-rise. This was the lowest share for any of the four property classes and may help explain why people don't think of apartment buildings when they think of high-rises. One-sixth to one-fourth of the reported hotel and motel fires occurred in high-rises. Roughly 1 of every 10 reported office building fires occurred in a high-rise. And roughly one-third of reported fires in facilities that care for the sick have occurred in high-rise buildings (see Table 2).

Is the fire risk higher in a high-rise?

Statistics on the U.S. building inventory by height of building are scarce and not ideally suited to calculating relative fire risk, but some analysis is possible.

The best data is on apartment buildings. Of the 94,724,000 occupied housing units that existed in the United States in 1993, 15,497,000 to 24,776,000 were in apartment buildings. The range is necessary because the housing data combine housing units in duplexes, which the fire databases don't consider apartment buildings, with housing units in buildings that have three to four units, which the fire databases do consider apartment buildings. There were 2,294,000 housing units in high-rises, which means that, in 1993, 9.3 to 14.8 percent of apartments were in high-rise buildings.

The year 1993 is the middle year of the period between 1991 and 1995, during which 8.8 percent of apartment fires occurred in highrises. Since 8.8 percent lies below the range of 9.3 to 14.8 percent, the risk of fire is lower in a high-rise apartment building than in an apartment building that isn't a high-rise.

During the same period, only 9.4 percent of civilian injuries, 5.7 percent of civilian deaths, and 5.6 percent of direct property damage in apartment fires occurred in high-rise buildings. For all loss measures except possibly injuries, then, fire risk is lower in a high-rise apartment than in any other type of apartment.

The figure of 2,294,000 housing units in high-rises has implications for how many high-rise apartment buildings there are, although you can't calculate the number exactly without knowing the average number of floors with apartments per building and the average number of apartments per floor. Considering reasonable possibilities, however, high-rise apartment buildings must number in the thousands and could conceivably number in the low tens of thousands.

For the other three property classes, other data must be used. In 1992, there were 21,000 office properties, 5,000 health-care properties, and an undetermined number of lodging properties, each with more than 100,000 square feet of space. Although it's extremely unlikely that any high-rise would have less than 100,000 square feet, it's quite possible that a building that isn't a high-rise could have more than 100,000 square feet. For example, NFPA's headquarters is only five stories high, including the basement, so it isn't a high-rise, but it does contain roughly 150,000 square feet of floor space. Therefore, the number of high-rise buildings in each category is some fraction of the number of buildings with at least 100,000 square feet of floor space.

Of the buildings with at least 100,000 square feet of floor space, those with more than 100,000 square feet contained 42.4 percent of the total square footage for office buildings, 72.9 percent of the footage for health-care facilities, and 26.2 percent of the footage for hotels and motels in 1992. Between 1990 and 1994,12.4 percent of office building fires, 33.8 percent of health-care facility fires, and 20.4 percent of hotel and motel fires occurred in high-rises.

The 1992 U.S. census data also show that the average number of square feet per building for buildings with more than 100,000 square feet was around 250,000 for office buildings and health-care facilities, and considerably higher, though indeterminate, for hotels and motels.

These statistics combined still won't tell us clearly whether the highrise share of fires, by property class, is higher or lower than the high-rise share of square footage in use, which is a reasonable measure of exposure. There's a range of possibilities consistent with these statistics for the share of square feet in high-rises, most of which indicate that the fire risk of office buildings is probably lower in high-rise buildings. The other two property classes, in which facilities are more likely to have large square footages in fewer than seven stories, are too uncertain to analyze.

When we talk about risk in this context, we mean only the risk of having a reported fire. When we talk about associated losses, however, we can make a few additional points. Fires in high-rise hotels and motels are much less likely to involve a death than fires in hotels and motels that are not high-rises. This means that, whether or not the risk of fire is somewhat higher in high-rise hotels and motels, the risk of fire death is probably much lower in high-rise hotels and motels. In office buildings, the risk of property damage resulting from a fire in high-rise office buildings between 1991 and 1995 was a small fraction of the risk resulting from a fire in low-rise office buildings, although the calculation is very sensitive to large-loss fires and can vary quite a bit from year to year.

Fire protection in high-rise buildings

Like its predecessors, the 1997 edition of the Life Safety Code has provisions for existing and new high-rises in each of the four property classes. Existing high-rise hotels must be protected throughout by an approved supervised automatic sprinkler system, unless every guest room or suite has exterior exit access in accordance with the Code. Existing high-rise health-care occupancies, as well as those three to six stories high, must be of fire-resistive construction. Other construction types are permitted only when automatic sprinkler protection is provided. Existing high-rise apartment buildings must have an approved supervised automatic sprinkler system installed throughout. They're exempted only if every living unit has exterior exit access or if the building has an "engineered life safety system" that's been approved by the authority having jurisdiction. The latter exception is also the only exception to the requirement for a complete, approved automatic sprinkler system in existing high-rise office buildings.

Despite these requirements, most of which have been in place for years, more than two-thirds of the high-rise apartment fires of 1995 occurred in buildings without sprinklers. Even in high-rise hotels and motels and high-rise facilities that care for the sick, one-fourth to three-tenths of the fires that occurred in 1995 occurred in unsprinklered buildings (see Tables 3 through 6).

Given the special requirements for high-rises that most building and fire codes contain, it's not surprising that automatic fire protection equipment and fire-resistive construction are much more common in high-rise buildings that have fires than in other buildings of the same use that have fires.

In 1995, the percent of buildings that experienced fires and had sprinklers was sometimes fairly similar for both high-rises and other buildings. In facilities that care for the sick, where sprinklers tend to be required in buildings of all heights, for example, 77.7 percent of highrises experiencing fires had sprinklers, while 66.6 percent of other buildings experiencing fires had sprinklers. More often, however, the percents were far greater for high-rises. For example, the presence of sprinklers was reported in 71.9 percent of high-rise hotel and motel fires, but their presence was reported in only 23.4 percent of other hotel and motel fires. The differences in the chances that a system will be present in high-rises and in other buildings are similar but less dramatic for detectors and fire-resistive construction.

The use of both sprinkers and detectors appears to have grown between 1985 to 1995, in some cases very rapidly, in both high-rises and other buildings in all four property classes, although again, the trend in recent years has been uneven. On the other hand, the use of fire-resistive construction in both high-rises and other buildings, particularly in facilities that care for the sick, appears to have declined.

The value of some of these fire protection features may be seen clearly in a statistical analysis of the loss-per-fire averages for the period between 1986 and 1995. For high-rise buildings, automatic suppression equipment is associated with a reduction of at least 71 percent in the rate of deaths per 1,000 fires for each property class and at least 42 percent in the average dollar loss per fire for each property class. Fire-resistive construction is associated with a reduction of 53 percent in average dollar loss per fire in apartment buildings. Because high-rise buildings often use all three systems, it's difficult to separate their effects on loss rates, and many rates are sensitive to deaths or large dollar losses in individual incidents.

Automatic suppression and detection equipment and fire-resistive construction also contribute to fire protection by helping to keep fires small. Suppression and construction do this directly, while detection does so by providing early warning, which often leads to earlier manual suppression (see Tables 7 and 8). In fact, fire and smoke are confined to the room or floor of origin in a much larger fraction of high-rise fires than low-rise building fires. In each of the four property classes, the probability that a high-rise fire will spread beyond the room of origin is roughly half the probability that fire will spread beyond the floor of origin in a building not tall enough to be a high-rise.

Finally, the power of these fire protection systems and their widespread use in high-rise buildings mean that people who die in high-rise residential fires are much more likely to be close to the fire, where those systems would have less time to act. In low-rise apartment buildings, 48 percent of those killed in fires between 1986 and 1995 were in the room in which the fire ignited; for high-rise apartment buildings, 81 percent of those who died were that close to the fire at the outset.

Other patterns of high-rise fires

Just because a fire occurs in a high-rise doesn't mean that it began on the upper floors (see Table 9). Indeed, most high-rise fires begin on floors no more than 70 feet abovegrade. Only in apartment buildings and office buildings do as many as 25 to 30 percent of high-rise fires start above 70 feet, in the part of the building that makes it a high-rise.

One special concern is fires that begin in exits. Because internal escape routes tend to be much longer in high-rise buildings, the consequences of a fire in the means of egress could be dire. In fact, more of the fires that begin in high-rise buildings do begin in means of egress, particularly halls or corridors, than fires in other buildings (see Table 10). Except for hotels and motels, however, the differences are small.

In any particular occupancy, high-rise buildings that have fires are much more likely to have sprinklers, detectors, and fire-resistive construction than shorter buildings of the same occupancy that have fires. Considering the extensive requirements of the Life Safety Code, however, it seems clear that there are still major gaps, particularly in the adoption and enforcement of provisions that require sprinkler and other life safety system retrofits in existing high-rise buildings.

This has implications for public officials and ordinary citizens in any city. Public officials should make sure that the latest edition of the Life Safety Code is in place and that the codes they've adopted are supported by effective code enforcement provisions, including plans reviews and inspection processes, both for new construction and for continued supervision of code compliance. The public can take responsibility for its own safety by insisting that its public officials take these steps.

As in so many areas of fire safety, we know what to do, we just need to do it. We mustn't forget that, when high-rise fires become large, they can overwhelm the resources of even a large, well-trained fire department.

1. To estimate the total high-rise fire problem in the United States in all kinds of properties, not just the principal four, we use hotel and apartment statistics to adjust for reported residential fires in buildings whose height is unreported and office and care-of-sick facility fires to adJust for reported nonresidential fires in buildings whose height is unreported. Hotels and apartments provide two different estimates of the best way to handle other fires, while offices and facilities that care for the sick provide two diferent estimates of the best way to handle other nonresidential fires. This produces four estimates for total high-rise fires, which are expressed as ranges.

TM of National Fire Protection Association

Copyright National Fire Protection Association Nov/Dec 1997
Provided by ProQuest Information and Learning Company. All rights Reserved