Permitted distances between separations are usually stated in square feet, not in number of dwelling units. It is up to the ingenuity of the designer to see how many units he can cram into one fire subdivision. This is a menace to all in the area, and in some cases a loss of "livability", if not direct fire damage to all units, is an almost predictable result of any fire which involves so much as one room. If this situation is to remain unchanged, perhaps the victims of official neglect should at least be warned.

1. Only a few fatalities occurred in the fires investigated and none could be directly attributed to any building defect; however, there were some injuries and "near misses".

2. Non self-closing doors permitted smoke or fire to escape into the stairway from the apartment of origin. This often forces other occupants to the refuge of the balconies. Rescue efforts slow down the control of the fire, and ladder rescues are always hazardous.

3. Fire involving the attic can soon cause roof overhang structures to fall on persons escaping from the building.

4. While nominally three stories high, garden apartments are often four stories high on one side due to rolling ground. When this fourth floor is in the rear it is inaccessible to truck mounted aerial ladders. The raising of a 40 or 50 foot ground ladder requires the efforts of a well drilled crew of several men which has worked together. This is not often delivered early in a fire by volunteers, and paid departments often do not have enough men in ladder companies.

5. Complicated street layouts and cul-de-sacs may improve the livability of garden apartments, but if the first due fire unit commits itself to the wrong driveway, the building can be lost. Fire departments expend great effort to get maps of all projects and to drill drivers, but direction signs do not take fire problems into consideration, and building numbers follow no standard pattern.

6. A fire hydrant is a useless obstruction until a fire pumper hooks into it, but hydrant locations are rarely coordinated with firefighting procedures.

7. In a number of cases substantial quantities of smoke were distributed through hot air ducts and the void spaces around them.

8. Current design developments feature economy of material, often by substituting geometry for mass.

Gusset plate roof trusses are now commonplace. Standard firefighting procedures, based on the reserve strength in conventional wood rafters, often call for men to get onto a roof to open it to release pent up heat ("ventilation" in fire parlance). To the best of the writer's knowledge, no firefighter has been injured carrying out this practice on the fast to collapse, truss roofs. The usual reason assigned is that fire is "through the roof" on arrival.

Recently we have seen floors made of wooden gusset plate trusses (see fig. 7). These provide a common unfirestopped void over (and under) the entire area of four apartments on each floor. If, as is probable, the voids are interconnected by vertical voids, the entire building can be involved in fire very rapidly (see fig. 8). The possibility of early structural collapse cannot be discounted.

"I" beams rolled of steel develop their characteristic shape from the flanges provided to resist the maximum compression (found in the top) and tension (found in the bottom). Until recently, there was no wooden "I" beam. Now such

Figure 7.

IN

These trussed floors permit a combustible void over and under the four apartments on a floor. Vertical channels may interconnect the several voids. In such a structure it is possible to envision collapse not long after fire penetrates the void.

beams are offered; 2" x 4" lumber forms the top and bottom flanges; the web is plywood. Such a beam may carry the same load as its sawn counterpart, but it appears self-evident that it could not make the same contribution to a fire resistive (combustible) assembly as a solid sawn beam.

In

9. Many gas meters are often located in one room for ease of reading. one case the twenty or so lines leading to apartments were grouped together overhead in the passage between storage cages. They were supported by horizontal pipes attached to the joists by pipe strapping. The failure of one nail dropped all the pipes on advancing firefighters and caused a massive gas leak. A serious tragedy was narrowly missed.

10. In one fire extensive areas of brick veneer fell off during the fire. This falling masonry is hazardous both to escaping occupants and firefighters, even though the building is not impaired structurally.

9. CONCLUSIONS

A meeting of builders is hardly complete without a speaker who raises the shibboleth, "Buildings would cost far less if we didn't have those terrible specification codes. New performance codes would let us use imagination and technology.' Here is a suggested "performance code":

"A combustible multiple dwelling should be so designed and
built that the "design basis fire" will be confined to the
area of origin for 30 minutes."

The design basis fire would be a typical garden apartment fire based on a study of fire loads. Such a requirement would lead to the elimination of many penetrations of the protective gypsum sheath. Where openings are necessary, there would be a demand for listed devices to provide openings equivalent to the protective sheath in fire resistance. Sociologically, the requirement should have great appeal. Since each of us is entitled to do "his own thing" in his apartment, the living conditions of others are equally entitled to be protected (see fig. 9). To have the tenants barely escape alive is no longer an adequate design criterion.

What can be done to improve existing buildings to deliver the right to "livability" to their occupants?

Work should be done to determine whether a few strategically placed high temperature sprinklers, fed from readily available domestic water, could not create a steam condition within vertical and horizontal voids which would act to control the free movement of fire through these areas.

A lightweight field foamable gypsum would provide a ready means of fire stopping voids, after the installation of all utilities. No such material is known to exist.

Retroactive1 legislation should require:

1.

The installation of supervised automatic sprinklers fed from the
domestic system in all laundry, storage, rubbish and similar rooms.
Fire officials are unanimous as to their efficacy.

2. The equipping of all apartment doors with door closers.

Renailing of poorly nailed gypsum, particularly ceilings.

1 "Retroactive" safety legislation is not "ex post facto" legislation and is constitutional.

Figure 9.

Xmas Tree. The tenants of this apartment lived 200 feet away from the apartment where the fire started. Should not the building "system" protect them from the other tenant? "Fire Prevention" will not help these innocent victims. Architectural awareness, adequate code requirements, and competent enforcement are required to assure the right to live in peace.

5.

6.

7.

Elimination or adequate firestopping of all voids common to more than one apartment.

Elimination of all exterior wood balconies, sheathing and soffits.

Painting of all combustible attics with fire retardant coating. Greater efforts should be made to educate fire departments in the deficiencies of these buildings. Adequate education, leading to realistic prefire planning, should contribute to cutting off the extension of a fire closer to the origin.

Questioning of tenants discloses that probably less than half have personal property fire insurance. Such insurance is usually quite cheap, so there is little incentive to anyone to "sell" it. If the owner was required to provide a personal property insurance policy up to some limit for each tenant, the insurer would now be examining a much larger risk, and there might be some pressure from this quarter for improvement. This is not a foregone conclusion; without a doubt, the buildings are insured.

In summary, since most of our population must be housed in multiple dwellings, the constitutional right of a citizen to be secure in his apartment home must be extended to protect him from a fire caused by his neighbor exercising his "constitutional right to be careless".

NATIONAL BUREAU OF STANDARDS SPECIAL PUBLICATION 411, Fire Safety Research, Proceedings of a Symposium Held at NBS, Gaithersburg, Md., August 22, 1973, (Issued November 1974)

THE CURRENT STATUS OF FIRE DETECTION

George Sinnott

National Bureau of Standards, Washington, D.C.

The status of residential fire detection systems is presented in terms of the major problems encountered in their use. This includes a discussion of risk benefit considerations, false alarms, and the effectiveness of alarms. The impact of these considerations on acceptance standards for residential fire detection systems is pointed out.

Key words: Alarm communications; false alarms; fire alarms; fire
detectors; risk benefit; smoke detectors; standards.

Common to much of what you have heard here today is the high expectation of the ability of technology to solve many social, human problems, even those created by contemporary technology itself. This is so much in us that we find ourselves surprised, for example, when what seemed like routine engineering problems years ago have yet to come to fruition. This is true not only of the so called "big" problems like breeder reactors or automated transportation systems but also of specialized ones like automatic fire detection. A number of articles and testimony before the National Commission on Fire Prevention and Control, and before Congress have described the potential benefits of modern fire detectors and have expressed disappointment that for some reason fire detectors have not become widely used. I want to describe here what might be called a "hangup analysis" of fire detection. That is, I want to explore the hangups, those untoward social and technological side effects, that have prevented the wide acceptance of residential fire detectors.

A good place to start is with an appraisal of the risks that one runs of fire at home. Of the 12,000 fire deaths in the United States each year, a little over half occur in homes. There are about 700,000 residential fires each year and since there are about 65,000,000 residences in the country this means that as an overall national average the mean time between fires per residence is about 100 years. This gives us the first hint of a problem. There

are very few devices commonly used by man with a mean time between failure of more than 100 years. We must then face the situation of installing in our home (or requiring others to do so in theirs) devices that have a much larger chance of failing than of being useful, or else of creating devices of exceptional reliability by any standard. What's more, these exceptionally reliable devices, as we shall see later, must be exceptionally cheap.

The two ways of looking at the statistics, then, is perhaps the first hangup. At meetings like this we discuss the 6,000 lives lost and perhaps 10 times that many injuries and $874,000,000 direct financial loss each year. Fire is nevertheless a rare occurrence for any one family. The average family perhaps intuitively senses the 10,000 years mean time between fire caused death per residence or the annual average $14 financial loss per residence and relegates fire well down on its hierarchy of fears. However, three major model building codes in the country as well as the Department of Housing and Urban Development's minimum property standards do now or will soon require fire detectors in residences. How alarmed should a person be at this hazard? Are there rational implications as far as public policy is concerned? That is, is there a statesmanlike position somewhere between being cynical and hysterical? Some rather nice insights to this problem have been presented in recent years in several articles by Chauncey Starr. [1,2] The argument here is put in terms of the risk of fatalities even though injuries are certainly an important consideration. The problem with discussing injuries is that the severity of injuries is difficult to measure and the statistics are haphazard.