• Aucun résultat trouvé

Opportunities for research in water-based fire suppression at the National Fire Laboratory

N/A
N/A
Protected

Academic year: 2021

Partager "Opportunities for research in water-based fire suppression at the National Fire Laboratory"

Copied!
48
0
0

Texte intégral

(1)

Publisher’s version / Version de l'éditeur:

Vous avez des questions? Nous pouvons vous aider. Pour communiquer directement avec un auteur, consultez la première page de la revue dans laquelle son article a été publié afin de trouver ses coordonnées. Si vous n’arrivez pas à les repérer, communiquez avec nous à PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca.

Questions? Contact the NRC Publications Archive team at

PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca. If you wish to email the authors directly, please see the first page of the publication for their contact information.

https://publications-cnrc.canada.ca/fra/droits

L’accès à ce site Web et l’utilisation de son contenu sont assujettis aux conditions présentées dans le site

LISEZ CES CONDITIONS ATTENTIVEMENT AVANT D’UTILISER CE SITE WEB.

Internal Report (National Research Council of Canada. Institute for Research in Construction), 1991-08

READ THESE TERMS AND CONDITIONS CAREFULLY BEFORE USING THIS WEBSITE. https://nrc-publications.canada.ca/eng/copyright

NRC Publications Archive Record / Notice des Archives des publications du CNRC :

https://nrc-publications.canada.ca/eng/view/object/?id=7f748e1c-a0b1-4a30-8431-c3d597f9c52a https://publications-cnrc.canada.ca/fra/voir/objet/?id=7f748e1c-a0b1-4a30-8431-c3d597f9c52a

NRC Publications Archive

Archives des publications du CNRC

For the publisher’s version, please access the DOI link below./ Pour consulter la version de l’éditeur, utilisez le lien DOI ci-dessous.

https://doi.org/10.4224/20374634

Access and use of this website and the material on it are subject to the Terms and Conditions set forth at Opportunities for research in water-based fire suppression at the National Fire Laboratory

(2)

Internal Report No. 610

Date of issue: August 1991

ANALYZED

r

L I B R A R Y

l

SEP

28

1991

B I B L I U

u

w e Q U E I I I I R e f

,

I

National Research Conseil natlonai

R427

/I

Council Canada de recherche8 Canada

no. 610

1

B&BG I Institute for lnstitut de

,

1 Research in recherche en

Construction construction

IWCmCMc

Opporfunities for Research in wafer-'bed

Fire Suppression at the NafSonal Fire

Laboratory

J.R. Mawhinney

This is an internal report of the Institute for Research in Construction. It is for personal use only and is not to be cited as a reference in any publication.

Canad"a

(3)

OPPORTUNITIES FOR RESEARCH IN WATER-BASED FIRE SUPPRESSION AT THE NATIONAL FIRE' LABORATORY

EXECUTIVE SUMMARY

This literature review identifies potential research opportunities for the National Fire Laboratory (NIX) at the Institute for Research in Construction in the area of water based fire suppression systems, including sprinkler systems. The report reviews research currently being conducted in Europe, United Kingdom, United States, Canada, Japan and Hong Kong on the extinguishing capacity of water and the potential effectiveness of water- based fire suppression systems in controlling f i s in buildings.

Eight categories of research relating to sprinkler systems and water-based fire

suppression systems are discussed. All of the research that is taking place internationally can be fitted into at least one of the categories. The eight categories themselves encompass basic, applied and qualitative types of research. Development of computer models of fire

growth and spread, plume dynamics, heat transfer and plume-spray interactions represents

basic research. Projects to relate sprinkler application rates to fuel loads, hazard

classification, building type and fuel configuration are applied research. The evaluation of the impact of sprinklers on life safety and the economics to justify or discredit their use represent qualitative research. Basic research is considered an essential underpinning to both applied and qualitative research.

Research projects that could be conducted at the Nn in conjunction with industry or other partners, which would be of value to the Canadian construction industry and which would compliment what is being done in other countries, are discussed. A ranking system is used to allow comparisons to be made of the relative merit of one topic versus another. Projects that involvepurchase of critical equipment or investment in developing expertise that is prerequisite to performing other research are identified. The report concludes with a table &recommended research bpics to be considered over the next 3 to 5 years at the National Fire Laboratory.

(4)

I TABLE OF CONTENTS

Introduction

...

1

I

...

i Categories of Research 1 I

...

I

Ranking of Potential Research Topics 2 I

...

i 1.0 Physical Characteristics of Water Sprays 3

I

1.1 Overview

...

3

I

...

i

1.2 Research Opportunities for the National Fire Laboratory - Table 1 4 1 1.3 Recommendations

...

5

i 2.0 Additives to Enhance Suppression Characteristics

...

5

2.1 Overview

...

5

...

2.2 Research Opportunities for the National F i e Laboratory - Table 2 6 2.3 Recommendations

...

7

3.0 Plume-Spray Dynamics

...

7

3.1 Overview

...

7

...

3.2 Research Opportunities for the National Fire Laboratory -Table 3 8

...

3.3 Recommendations 9 4.0 Heat Transfer in Water Sprays

...

; 10

4.1 Overview

...

10

4.2 Research Opportunities for the National Fire Laboratory -Table 4

...

11

4.3 Recommendations

...

12

5.0 Fire Growth Rate and Sprinkler Performance

...

:.

...

12

5.1 Overview

...

12

...

5.2 Research Opportunities for the National Fire Laboratory - Table 5 13

...

5.3 Recommendations 14 6.0 Sprinkler Activation and Performance

...

14

...

6.1 Overview 14

...

6.2 Research Opportunities for the National Fire Laboratory - Table 6 16 6.3 Recommendatiods

...

17

7.0 Special Water Spray Applications

...

17

7.1 Overview

...

17

Aircraft . sprinkler systems

...

18

Extinguishing high velocity gas jets

...

18

Protecting HEPA (high efficiency particulilte air filters)

...

19

Sprinklers on glazing to prolong fire resistance

...

19

(5)

7.2 Research Opportunities for the National Fire Laboratory -Table 7

...

20 Aircraft sprinkler systems

...

.;

...

20

...

Water spray systems for flammable liquids 21

...

Water spray systems as replacement for halon 21

7.3 Recommendations

...

22

...

8.0 Evaluating the Effectiveness of Sprinkler Systems 22

...

8.1 Overview 22

8.2 Research Opportunities for the National Fire Laboratory -Table 8

...

23 8.3 Recommendations

...

23 Conclusions

...

25 Table 9 Compilation of potential research topics in water-based fwe

suppression

...

26 Table 10 Summary of research topics relating to water based suppression

...

systems recommended for study at the National Fire Laboratory 30

(6)

OPPORTUNITIES FOR RESEARCH IN WATER-BASED FIRE SUPPRESSION AT THE NATIONAL FIRE LABORATORY

ABSTRACT

This literature review identifies potential research opportunities for the National F i e Laboratory (Nn) at the Institute for Research in Construction in the area of water based Fie suppression systems, including sprinkler systems. The report reviews research currently being conducted in Europe, United Kingdom, United States, Canada, Japan and Hong ~ b n g onthe extinmishing &macity of water and the potential effectiveness of water- basea f i e

upp press ion

sfstems

X

c6ntroiiing fires in buildikgs. Research projects that could be conducted at the NFL in conjunction with industry or other partners, which would be of value to the Canadian construction industry and which would compliment what is being done in other countries, are discussed. A ranking system is used to allow

comparisons to be made of the relative merit of one topic versus another. Projects that invoive purchase of critical equipment a investment hi developing expertise that is

arereauisite to aerformine other research are identified. The report concludes with a table bf rewmmendd researc6topics to be considered over the next3 to 5 years at the National Fire Laboratory.

INTRODUCTION

The purpose of this literature review is to identify potential research opportunities for the National Fire Laboratory (NFL) at the Institute for Research in Construction in the area of sprinkler systems or water-based suppression systems. Research currently being

conducted in Europe, U.K.! U.S., Canada, Hong Kong and Japan is providing new understanding about the ext~nguishing capacity of water and about the potential

effectiveness of water-based fire suppression systems in controlling fues in buildings. This study is based on a review of published information on sprinkler research being done worldwide and on discussions with individuals involved in sprinkler research in the U.S. and Canada. Its objective is to identify research that could be conducted at the NFL in conjunction with industry or other partners, which would be of value to the Canadian construction industry and which would compliment what is being done in other countries.

CATEGORIES OF RESEARCH

A bibliography containing 128 entries of journal articles, technical reports and papers appears in Appendix A. The material was assembled over a ten-month period between January and October, 1990. New articles and papers continue to appear, but it is

considered that the material reviewed identifies rite major directions of sprinkler research at

-

the present time.

All 128 items in the bibliography relate in one way or another to the use of water sprays for f i e suppression. To facilitate discussion of the research, the bibliography has been classified into eight general categories of research, as follows:

Category 1. Physical characteristics of water sprays

Category 2. Additives to enhance suppression characteristics Category 3. Plume-spray dynamics

(7)

Category 5. Fire growth rate and sprinkler performance Category 6. Sprinkler sensitivity and performance Category 7. Special water spray applications Category 8. Evaluating sprinkler effectiveness

In the bibliography in Appendix A, articles are grouped according to the category (or categories) into which they fall. Articles that treat more than one aspect of the technical question appear under more than one category heading. Categories 3 , 4 and 5 particularly have strong cross-relevance.

The eight categories listed can be described generally as either basic, applied or

qualitative twes of research. Categories 1,3 and 4, involving the physics of suppression, computer maelling of fire growth-and spread, plume dynamics, hear-transfer

a d

plume- s ~ r a v interactions. revresent basic research. The research described in Categories 2 , 5 , 6 &d

7

to relate spnnkier application to fuel load, hazard classification, build& type and fuel configuration, could be described as applied research. Finally, research described as Category 8, on evaluating the impact of sprinklers on life safety and the economics to justify or discredit their use, represents qualitative research. The basic or theoretical work is an essential underpinning of both applied and qualitative research.

RANKING OF POTENTIAL RESEARCH TOPICS

A number of detailed research topics, or "problems" are identified for each general category. In order to make it easier to compare the value of conducting research in one subject area versus another, in terms of the resources and objectives of the NI;IL, a ranking system involving four ranking criteria is used:

a) -: the degree of technical challenge or complexity;

b)

m:

the likelihood of the NFL making a contribution to the field in 3 to 5 years;

C)

Utilization:

the degree of utilization of current resources (skills or equipment) to be able to do the research: if demee of utilization is low, it follows that d e m e of

-

investment in training or equipkent would be high;

d)

m:

.the likelihood of obtaining outside funding to support the research, either during the development or application stage.

Each criterion is assigned an integer value between 1 and 5 such that 1 ranks "low" and 5 ranks "high." No anempt is made to be more precise than whole integers, since the sole purpose of the assigned numbers is to enable comparisons to be made. A high ranking (4 or 5), in any individual criterion, generally indicates that it would be worthwhile to undertake research in that area, if that criterion was the only factor to be considered. ,

It is possible to sum the individual rankings to obtain an overall "score" for each research project. Comparing scores will provide a measure of relative merit for the recommendations. It is important, however, to retain all four rankings so that the

significance of certain pertinent factors can be appreciated in detail. For example, a subject that has a low degree of utilization of current resources, combined with a low likelihood of obtaining outside funding, would have

a

relatively low overall score, even if it had a high potential for contribution. But, the imvortanc~ of making an initial investment in that research may be very high because it is prerequisite to performing other research. In that case, the factors of technical challenge and potential contribution should be weighted more than the negative factors of utilization and funding. Such considerations could be lost if the

(8)

overall score alone was used to judge merit, because the ranking system does not weight one criterion more than another. Additional commentary is provided where the research is considered a prerequisite to future research.

The following discussions provide an overview of the main research problems in each category and identify key researchers or agencies involved in the work. Research topics or possbifities to advance the state-of-the-arin each category are then tabulated and ranked according to their relative merit as potential research areas for the NFL. Research problems and theirirrankings are presented inbbular form at the end of each discussion and again in Tables 9 and 10 at the end of the report.

1.0 PHYSICAL CHARACTERISTICS OF WATER SPRAYS

1.1 O v e ~ e u

This category of study includes research into the physics or mechanics of how water extinguishes or suppresses fire, hence could be described as "basic" research. Water has a high latent heat of vaporization, such that much heat is extracted from the surrounding environment when water droplets pass from the liquid to the gaseous state. The total surface area exposed to heat, hence the size of the water droplets, plays a major role in determining the rate at which heat exchange occurs. Thus, solid hose streams of water are much less effective in cooling hot gases than fme droplets. Droplets must have enough momentum (mass times velocity) to be able to penetrate a fire plume and wet the burning and unburned fuel, but still be small enough to extract a sufficient quantity of heat from the hot gases so that combustion cannot be sustained.

There are at least two reasons for wanting to know the droplet size distribution in a sprinkler spray. One is to make it easier to design sprinklers to penetrate fire plumes for different intensities of fues. A second is to be able to quantify the heat absorption rate of the spray for computer models that calculate the cooling of the hot gas layer. In the search for the most effective distribution of droplet diameters for specific conditions, it is

important to be able to measure the droplet size distribution in the water spray. Factory Mutual Research Corporation (FMRC) began work in this area in the 1960's. An early approach involved capturing spray droplets on an oil surface. Individual droplets "float" on the oil (due to surface tension) where their diameters can be measured using a

microscope. This method was most suitable for measuring very fine mists such as occur in fuel sprays in engine combustion chambers. Large diameter droplets, such as are present in sprinkler spray, could not be measured this way. A more successful method was investigated by Kroesser (1969) of FMRC, who used liquid nitrogen to freeze-capture sprinkler water droplets. The diameter of the frozen droplets could then be measured, and the sizes counted to estimate the size distribution.

Kroesser's technique had only limited usefulness for measuring sprinkler sprays, however. Droplets conglomerated, or were misshapen (not spherical), and there was no information about the velocity of droplets. Velocity of a droplet in a hot gas has a

significant effect on the rate of heat exchange between the hot gases and the droplet. In the early 1980's both FMRC and the National Institute for Science and Technology (NIST,

--

.

formerly the National Bureau of Standards) adapted a technology used by meteorologists for measuring cloud dro~let sizes (You. 1983: Lawson and Evans. 1988). This techniaue was based o ~ ~ h a d o w ~ & ~ h s which used very.bright light to cast shadows of droplets dn a video camera. The camera could then "sample" at different i.*cations within the spray. The diameter of the shadows could be measured. A computerized statistical procedure provided an estimate of the droplet size distribution (Goodfellow, 1985).

(9)

The shadowera~h techniaue is verv useful for measuring the size distribution of droplets& wateFs<rays. Thd effectiveness of a sprinkler iticontrolling fire depends on more than the droplet size distribution, however. The momentum of the droplets governs whether the spray can overcome the buoyancy and shear forces of the plume of hot gases. In order to address this, researchers at NIST (Lawson) are reportedly developing a droplet measurhg device that will be expressly suited to the size range of sprinkler sprays

(100 pm to 2000 pm), and which will be able to measure the velocity (magnitude and direction) of the droplets. The NIST work is still in progress.

No references to British, European or Japanese efforts in this area were located. 1.2 Research O~~ortunities for the National

F

i

r

e

Laboratory

Both NIST and FMRC are working on new technology for measuring droplet sizes and velocities in sprinkler water sprays. It would be difficult for the NFL to compete with those organizations on measuring those parameters. Nevertheless, being able to measure them would be an important addition to any research involving water-based suppression systems. In particular, experiments to determine mass flow rates of water spray entrained in fire plumes or the minimum flow rates of very fine water sprays to achieve

extinguishment, would depend on being able to quantify the spray. Kokkala (1989) noted that lack of information on water droplet size and velocity was the "biggest defect" in the results of his experiments using water sprays on pool fires. A potential client of the NFL

was recently disappointed to learn that we could not determine droplet size distribution - information that was critical to the heat transfer calculations that he wished to verify.

Two research opportunities in Category 1 for the NFL over the next five years are evident. The first is to obtain the equipment and develop the expertise necessary to be able to measure water droplet size distribution. The second is to purchase equipment to measure droplet velocity as well as size distribution.

The technical challenge of developing expertise in measuring droplet size distribution is relatively high (Challenge = 4). However, it would also have a high potential for

contributing to improved understanding of water sprays and for expanding the application of water sprays in fire suppression systems (Contribution = 4). Because of the capital

expenditures necessary to purchase equipment and software, hnd for learning how to use the equipment, the degree of utilization is relatively low (Utilization = 2). Because it is not

an end

h

itself, but a iechnology that is to being able to study spray-fie

interactions in general, it may not be easv to obtain funding from outside agencies for initial ~urchase of theeauivment.

he

~unding criterion would Sherefore a v ~ e s o be low. The cost of the inves&eht could be recover2 over time, however,as it is' &zed on other research projects for which funding is available. A rank of 3 is assigned to Funding.

The preceding research opportunities and rankings in Category 1, Physical Characteristics of Water Sprays, are sumnmized in Table 1.

Table I

Score

*

12

10

Research Opportunlty

1.1 Develop capability to measure droplet & distribution.

1.2 Purchase equipment to measure droplet y&&y

as well as size distribution, when available.

Chal- lenge 4 5 Contrl- hution 4 3 Utillz- ntion 2 1 Fund- ing 2 1

(10)

1.3 Recommend-

The capacity for the NFL to conduct sprinkler related research of international quality will be greatly enhanced if the ability to measure at least droplet size distribution is developed (Item 1.1). For this reason, it is recommended (indicated by the asterisk) that the Nn make the necessary investment in equipment and trainiig to develop this

capability. The capital costs (approximately $50 K) could be indirectly absorbed in research contracts that would be generated once the capability was in place.

The complexity of the technology to measure droplet velocity (magnitude and direction), as described in Item 1.2, is much higher than that for measuring size

distribution. Because of this, it is somewhat less likely that the NFL will be able to make a significant contribution in this area in the next 3 to 5 years. It will be preferable to await the development of a measurement technology at other institutions, such as NIST.

2.0 ADDITIVES TO ENHANCE SUPPRESSION CHARACTERISTICS

This category of applied research examines ways to enhance the iire suppression action of water by modifying its physical properties. It may then become possible to use water on fuels that are incompatible with water, on fuels that repel or prevent penetration of water, or to reduce the auantitv of water needed to extinmish a fm. Additives include anti-freezing agents, foami& ageits, wetting agents, thickening agents, and agents to reduce or increase o~acitv and reflectivitv. Additives are also available to modify flow characteristics to rkucd costiand to modify the electrical conductiviiy of water.

Water extinguishes f i e in four principal ways: by heat extraction, suffocation,

emulsification and dilution. Under certain circumstances, water may chemically inhibit fire by breaking the combustion chain reaction. Additives will work to enhance one or more of these basic mechanisms of extinguishment. Heat extraction ability can be enhanced by adding various chemicals to either "thicken" or "thin" the water. Wetting agents reduce the surface tension of water so that it can penetrate into porous fuels, such as cotton bales, or break up into f i e r sprays in spray nozzle systems. Thickening agents arc: used to form a gel that sticks to vertical surfaces, where it will continue to absorb heat and prevent ignition of unburned fuel. Plain water would run off the surfaces and drain away. Where water curtains are used to block radiant heat or to cool surfaces uniformly, additives may be used to reduce or increase opacity of the liquid, which will modify the rate at which radiant heat will be absorbed. Emulsification is the process by which non-water soluble liquid fuel droplets are dispersed in water, each droplet surrounded by water so that volatile vapours cannot escape. Certain chemicals may be used to enhance the ability of water to form and sustain an emulsion. Dilution of water soluble liquid fuels may be enhanced by addition of certain chemicals.

The suffocating effect of water on fire can be enhanced by adding foaming agents. Foam, applied in a thick blanket over the burning materials, cuts off the oxygen supply to .

- -

- - -

the fuel. Attention must be aid to the ~otential <hemical interaction betweirithe foam and liquid fuel which might cause the foam to break down. Fuels that are water soluble (so- called "polar" solvents) can cause rapid breakdown of the foam unless chemically

appropriate foaming agents are used. Since foams are so widely used for outdoor fires, for example, for fuel spills and tank farms and for aircraft fires, freezing paint depressants may be added to permit their use in cold weather. Consideration must be givm to potential corrosion problems where foam is used to flood cable chases or equipment rooms.

(11)

Much research into use of additives to enhance the effectiveness of water as a fire suppressant has already been done. Work was done in the U.S. in the 1960's (Aidun and Grove, 1961; Strata, 1969). Starting with Strata, work was done at FMRC (U.S.) in the

1970's to introduce "ablative" liquids (thickened) for use in sprinkler systems for high challenge storage fires. In the U.K., considerable work was done at the Fire Research Station during the 1970's (Nash, Corrie et. al.) to look at high expansion foam as a possible approach for high rack storage warehouse fire protection.

Currently, the National Fire Protection Research Foundation (NFPRF) in the U.S. is funding research into the use of low-density foam in combined foam-water sprinkler systems for warehouses storing flammable liquids in plastic containers. Sprinkler systems discharging water only have been found to be seriously challenged by the rapidly growing fires that occur in such fuels. Based on personal communication with NFPRF, the results of initial tests indicate that foam-water sprinkler systems can achieve control over fire in single pallets of plastic containers of flammable liquids in cardboard cartons. Further testing is needed to assess the ability of foam-sprinkler systems to control fire in other configurations, such as high piled storage.

2.2

r

Res ch

n

Opportunities for the NFL to investigate use of additives to enhance the performance of water as a fire suppressant are few, but not non-existent. The NFL recently completed a contract to test foaming agents and develop a Canadian Standard for the Department of National Defense (DND), through Underwriters' Laboratories of Canada. The proposed tests are intended to help the client determine minimum quality criteria for purchasing quantities of foaming agent. Additives such as wetting agents and thickeners are mostly of interest to forest fire fighting agencies. Use of ablative (thickened) liquids in warehouse

sprinkler systems has been researched and the idea shelved. The relative merits of particular foaming agents are usually well researched and documented by the

manufacturers. In addition, practical problems associated with use of additives (corrosion, maintenance of reservoirs of chemicals, complicated injection devices) limit the application of such technology. Nevertheless, there remain some aspects of the use of additives that have not yet been extensively researched. "

Chemical Inhibition -

As mentioned in Section 2.1, there is a possibility that water with certain additives may chemically inhibit fire by interfering with the combustion chain reaction at the molecular level. This is at least the claim of a Canadian distributor of a f i retardant compound. The distributor claims that addition of the compound to water, which is then discharged as a fine spray, enhances the effectiveness of water spray to extinguish fire and prevent re-ignition of unburned fuel. Whether the reported extinguishing ability is indeed better than plain water spray, and if so, whether the improvement is due to enhanced heat extraction resulting from change in viscosity and enhanced droplet uniformity, or due to a genuine "chemical inhibition," remains to be verifkd

As will be discussed later in this report, the fire protection communities in North

America and Europe are seeking a replacement for halon, which chemically inhibits f i but which is envir6nmentally damaging. Fine water sprays are being examined by several interested parties as one possible alternative. British Petroleum Ventures in the U.K., and ~ e r b e G s in ~ w i t z e r l h d are presently researching such systems. Perhaps the effectiveness of fine water suravs can be enhanced bv the addition of certain * .

chemicals, thus increasing the potential of fine water sp;ay systems as a replacement for halon. These questions would merit study as part of any research into fine water spray suppression systems, which will also be discussed later in this report.

(12)

Modified Opacity or Reflectivity -

The effectiveness of additives to thicken the water to form a better heat absorbing film, or to make the water film opaque so as to better block passage of radiant heat, is of theoretical interest. These possibilities could be investigated as part of the NFL's ongoing tests on the use of water spray to increase the fire resistance rating of glazing. Certain additives might enhance the effectiveness of fine water sprays as a protection of people or equipment against radiant heat, as in the case of an external fuel fire exposing an aircraft passenger compartment.

The preceding research opportunities in Category 2, Use of Additives, can be summarized and ranked as shown in Table 2.

Table 2

2.3 Recommendations

Research into Items 2.1 and 2.3 in Table 2 would only be justifiable if clients emerged to fund the work, which is not considered likely. Item 2.2 could be briefly investigated as part of a more comprehensive study into fine water spray systems, which is discussed under Category 7. None of the suggested research areas warrant investment by the

NFL

at this time.

3.0 PLUME-SPRAY DYNAMICS

3.1 Q v e r v i e ~

Research into the dynamics of fire plumes and how water sprays interact with plumes represents a third category of study. The need for basic research of this type arises from current efforts in the U.S. (FMRC and NIST), the U.K. and elsewhere, to develop mathematical field models of how fne grows and spreads in a building. It is important to be able to quantify the effect that sprinklers have on those phenomena (fire growth and spread). This category of sprinkler research is closely linked to Category 1, droplet size measurement, and Category 4, heat transfer.

The subject area involves numerical modelling of spray induced flow fields. Questions of droplet size, velocity and momentum become very important in analyzing how much water actually penetrates the hot gases or fire plume and reaches the fuel. The theoretical component of this type of study demands strong modelling skills and heavy commitment of computer resources. Existing modelling codes, such as PHOENIX, HAZARD I (NIST) and the TEACH-T code (FMRC), promise impressive results, but need further

development and especially validation with full-scale testing.

Leading American researchers in plume-spray interactions include Alpert, Heskestad, Delichatsios You, Kung and Han at FMRC; and Quintiere at NIST. Researchers in the U.K. include developers of the JASMINE field model at the Fire Research Station (no

Research Opportunity

2.1 Continue testing foaming agents for client@).

2.2 Conduct tests using additives in fine water sprays

to determine if "chemical" suppression occurs.

2.3 Investigate effect of modified opacity or

reflectivity on effectiveness of fine sprays or thin rims to block radiative heat transfer.

Utlllz- atlon 4 4 4 Chal- lenge 2 2 3 Fund- ing 2 2 1 Contrl- hutlon 1 2 2 Score 9 10 10

(13)

reference available); and Hoffman and Galea at Thames Polytechnic (with Markatos at the National Technical University in Athens, Greece). Chow and Fong at Hong Kong Polytechnic, and other researchers in Japan, Australia and New Zealand are working on modeling plume-spray interactions.

In Canada, J. E. S. Venart, A. C. M. Sousa, at the University of New Brunswick and

G . Hadjisophocleous with the

NFL

have made a significant contribution to the computer

modelling of convection flow patterns in aircraft cabins under post-crash f i exposure (Hadjisophocleous, et. al, 1989). This experience with modelling fire plume development in aircraft compartments could be applied to an examination of the use of fine sprinkler spray suppression systems in aircraft compartments.

The geometry of the space in which the fire occurs also influences the suppressibility of a fire. In very high-ceilinged spaces (atriums), layering of smoke and hot gases can occur, such that the heat never reaches the ceiling level where it could activate sprinklers. Sloping ceilings or ceilings with deep channels may accelerate a fire plume. Acceleration of the hot gases will increase the rate of heat transfer to combustible ceiling materials, and make it more difficult for sprinkler spray to reach the fuel. This phenomenon may have been a factor in the variable performance noted for sprinklers operating under deep solid-web wood trusses in f i e tests conducted by the wood truss industry in 1989 in the U.S. (Fleming, 1990). It is possible that channelization created by the parallel trusses had a detrimental effect on both the response time of the sprinkler and the sprinkler spray pattern. In another example, the King's Cross subway station f i e in London revealed some

unexpected behaviour of fire due to the U-shaped cross section of the upward-sloping escalator tunnel (Aresu de Seui, 1989). Fire spread with unexpected rapidity up the escalator and into the ticket foyer. The geometry of the space was a significant but

, previously unsuspected factor in fire spread in that case.

The question of how smoke venting of buildings affects the performance of the

sprinkler system has received a great deal of attention. Recent studies in Europe contradict ~ o r t h American attitudes that smoke venting is detrimental to the p e r f o m c e of the sprinklers due to the fact that the ventilation may increase the fire intensity. Computer modelling to predict the influence of ventilation on sprinkler operating times is being done by several researchers (Gustaffson, 1989; Cooper, 1988). The problem is complex and difficult to model

-

zone models appear to be limited, and even field models have

limitations due to the complexity of real building ventilation conditions. Experimental work is needed to a) determine whether gravitational venting of buildings can actually prevent smoke logging of the building; and b) determine whether smoke vents will, in fact, delay sprinkler actuation and increase fire intensity.

More research into the effects of building geometry on the fire plume-spray interactions is needed in order to design effective sprinkler systems. Experimental work must be done to validate the computer models of fire plume and water sprays and to learn how to design sprinkler systems to achieve suppression in terms of fuel type, fuel configuration and building geometry.

3.2 Research Op~ortunities for the National Fire Labomtoy

Despite the h a w international involvement in this challenging subject area, the NFL is in

, a good position to &e a contribution to both the theoretical &d experimental aspects of this

field of studv. On the theoretical side. the field modeling expertise of G. Hadiisophocleous and the potential for the Nn to workcooperatively wit6 the-university of N ~ W ~runswick

(UNB) are two advantages. The NFL has already invested in HAZARD I, a computer code that models fire growth in a compartment, which could be modified to investigate

fire

plumes in high ceilinged spaces with sprinklers. In addition, UNB has capability with PHOENIX, a very powerful computer field model, that could be applied to this research. The NFL has

(14)

also obtained the right to use PHOENIX for research purposes. The Burn Hall facility at Almonte is a suitable facility for conducting experiments involving strong fire plumes and measuring temperatures and velocities to validate computer predictions. Other research organizations (international) are working on validation of the computer models, but not necessarily on the validation of models of the interaction of sprinkler sprays with the fire plume. The NFL could play a unique role by concentrating on such interactions.

The NFL has a unique facility in the smoke tower at the Alrnonte Field Station. Studies on the effect that sprinklers have on the nature of smoke and smoke movement in buildings could be conducted in this tower. Data from such experiments could be used to improve the computer models of smoke movement and fue spread in buildings.

Current thinking on sprinklers is based on the concept that sprinklers must be located at ceiling level, such that water droplets must penetrate the f i plume against the direction of gas flow. It is logical to locate sprinklers at the ceiling, of course, because the hot gases rise to that location and will automatically trigger the suppression system early in the development of the fue. However, where ceilings are very high, as in atrium, the sprinklers may never operate because heat becomes layered and never reaches the ceiling. Furthermore, when the fire plume is too strong, the droplets may never reach the level of the burning fuel. For such scenarios, it may be more appropriate to inject fine water sprays at the elevation of the fuel in order to entrain the water droplets into the combustion air feeding the fire. If the mass flow rate of water mist entrained into the fire is high enough in relation to the mass burning rate of the fuel, enough heat should be extracted to extinguish the fue (Evans and Pfenning (1985)).

Improvement of f i e protection strategies for aaia is likely to be of interest to building officials, therefore the potential to obtain outside funding is ranked as relatively high (4). Utilization of current resources is also ranked high (4), due to the existence of the Alrnonte Burn Hall, which could be used to measure the dynamics of f i e plumes in high spaces and to assess the effectiveness of fine water sprays injected into the combustion air supply on all sides of the fire. The data would be useful for validating computer models and could lead to innovative but practical fire protection strategies for atria.

The preceding research opportunities in Category 3, Plume-Spray Dynamics, can be summarized and ranked as shown in Table 3.

Table 3

3.3 Pecommendation~

The NFL is already involved in developing expertise with HAZARD I and PHOENIX, which model plume development. Every opportunity to use those models to investigate the effectiveness of sprinklers in high-ceilinged spaces, such as atria or industrial buildings,

Utillz- ation 3 4 4 Fund- ing 4 2 2 Research Opportunlty

3.1 Examine effect of sprinklers on plume dynamics in ahia or other types of high spaces, such as

high industrial buildings.

3.2 Examine effect that smoke venting has on operation of sprinklers in buildings, and on computer models of smoke movement. 3.3 Compare plume dvnamk for sprinkler sprays

from ceiling versus fme sprays injected into the combustion air supply at low elevations.

Score 16 14 15 Cbal- lenge 4 4 4 Contrl- butlon 5 4 5

(15)

should be pursued. Collaboration with FMRC, NIST or other agencies, including

UNB

in Canada, in the verification of computer models of smoke spread should also be sought. Item 3.1 in Table 3 represents an area in which the NFL will be compelled to "keep up" with advances made in other fire research organizations if it is to retain any credibility in the field.

Item 3.3 represents a "new" approach to the use of water sprays in large spaces; existing plume-spray models operate on the basis of droplets penetrating the plume from above. Research into this would be unique in North America, and therefore provides an opportunity for the NFL to make a significant contribution. Outside funding may be difficult to obtain, however.

4.0 HEAT TRANSFER I N WATER SPRAYS

4.1

The primary objective of basic research involving heat transfer in sprinkler sprays is to

understand and optimize the cooling effect of the water spray. It is important to be able to predict the degree of heat extraction by the spray for the purpose of computer modeling of fire growth in buildings.

Analysis of heat transfer in sprinkler sprays requires knowledge of fire dynamics, as

well as the characteristics of water soravs. Sprav characteristics include thermal proverties

- -

of water and &let size and velociiv disaibhti6ns. Understanding fire dynamics reuuires information abo& gas temperatures,;elocities and masstenergy flow rat&. The

mathematics of convective and radiative heat transfer to droplets and thin films must be understood. The rate of heat transfer between hot gases and the heat sensitive element of the sprinkler itself influences the time to activation of the sprinkler (which will be discussed in Section 6.1). The computational skills and resources required to deal with heat transfer are essentially the same as those required for analyzing plume-spray interactions.

Major researchers working in this area include Alpert and Delichatsios (1986) at

FMRC; You, Kung and Han, also at FMRC; Chow (1989), and Chow and Fong (1990) of Hong Kong Polytechnic; and Ravigunuajan and Beltrav (1989) (Belmv and Associates Inc., New York). The Fire Research Station (FRS) in the U.K. is doing research in this area as well, as part of their development work on the "JASMINE field model for predicting fire growth and spread. However, no direct references to their work are discussed in this review.

Alpert and Delichatsios, in their paper "Calculated Interaction of Water Droplet Sprays with Fire Plumes in Compartments," analyze mathematically, through the use of computer solutions (using the TEACH-T Code), the complex interaction between water droplet sprays and the buoyancy-driven gas flows induced by a fire. They attempt to predict spray penetration of the fire plume and the degree of cooling, taking into account such parameters as fire intensity, spray characteristics and ceiling height. You, Kun and Han (1986) - -

conducted ex~erimeits to measure the heat tranifer &curring in a sprinkler spray in a room. including radiative heat losses through the room opening. Chow and Fong (1990) at ~ o n g ~ o n ~ ~ o i ~ t e c h p i c have performed eaaustive numirical Ealculations to m d e l a simp=ed spray-plume heat transfer. Ravigururajan and Belmv (1989) have looked closely at modeling attenuation of radiative heat transfer by water droplets.

These researchers are doing highly analytical work. Despite its mathematical

sophistication, however, it is cleararthat many simplifying assumptions have to be made to make the heat transfer eauations solvable. For example, the numerical modelling done by Chow and Fong does not take into account combustibn occuning within the sm6ke layer, nor the effect that the sprinkler spray may have on reducing the intensity of the fire. The

(16)

results of the time-consuming computer calculations represent only a single, unlikely scenario. All researchers indicate that full-scale experimentation is needed to validate the results of the models.

4.2 Research Opportunities for the National Fire Laboratory

Research opportunities for the NFL in the area of heat transfer between water sprays and fire plumes or hot gas layers are similar to those suggested for fire plume-spray

interactions. The computational skills and computer resources needed to make a significant contribution to the theory of heat transfer in water sprays are high. Nevertheless, it is important that

NFL

researchers working on computer modelling for fires in buildings keep up with the development of algorithms for analyzing heat transfer to water sprays under development at FMRC and NIST in the U.S., and at the FRS in the U.K.

The trend for fire protection engineering in Canada and the U.S. is unmistakably towards use of computer models to assess how fire impacts a building and which fire suppression or control strategies are most cost effective. Without appreciation for the difficulties of modelling the effects of water spray on cooling and fire control, NFL researchers will not be able to remain at the leading edge of this technology. NFL

researchers will be expected by the Canadian conshvction community (building officials) to provide guidance on the appropriate use of computer models, as more and more Canadian

fire protection engineers begin to use and, unfortunately, mis-use the computer models that are already available.

A request for the NFL to analyze heat transfer in sprinkler spravs has already been received. The potential client (a university/consulting engineer joint venture) wanted to conduct tests to measure the heat transfer between hot water sprays and cool air. The prospect was discouraged partly by the fact that the NFL could not measure droplet size distribution. The existence of even one potential client for this type of research suggests that the probability of obtaining funding is relatively high (3).

The Risk Cost Assessment Model project, under development at the NFL, is not at this time able to take into account the actual impact of sprinklers on a building fire. Instead, a probability of either failure or effective operation of the sprinkler system is estimated. Use of the HAZARD I model or of PHOENIX to evaluate the effects of a sprinkler system on a fire could result in quantified parameters that would improve the capability of the Risk Cost Model.

As previously mentioned, all research agencies involved in dc veloping computer codes note that experimentation is needed to validate the results of the calculations. It may be possible for the NFL to work in cooperation with one of those agencies, for example FMRC, to conduct experiments that will provide data needed for a specfic validation. Slight modification of client-driven experiments that involve water sprays and real fires may result in collecting data that is of value to the client and as validation of certain theoretical concepts in the computer models.

Experimentation into heat transfer in water sprays will require capability to measure droplet size distribution in the water spray, as discussed under Category 1.

The preceding research opportunities in Category 4, Heat Transfer in Water Sprays, can be summarized and ranked as shown in Table 4.

(17)

-

12-

Table 4

4.3

Recommendations

Item 4.1 represents an area in which the NFL will be compelled to "keep up" with advances made in other fire research organizations, if it is to retain any credibility in the field. It is prerequisite to being able to conduct credible research in water-based

suppression systems. Collaboration with FMRC, NIST or other agencies, including UNB in Canada, in the verification of computer models should be encouraged. In terms of the use of fine sprays in place of conventional spridder systems in buildings, Item 4.3 is an area of research that is not being investigated in any other research agencies. It represents

an opportunity for the NFL to conduct unique research, subject to being able to attract outside funding.

5.0 FIRE GROWTH RATE AND SPRINKLER PERFORMANCE

5.1 Overview

The objective of this category of applied sprinkler research is to define the performance criteria for sprinkler systems in terms of fuel type and configuration, fire growth rate, heat release rate and building geometry. Empirical studies to relate sprinkler system

requirements to fuel loads, full-scale fire testing of sprinklers on high-rack storage, and the use of foam-water sprinkler systems for flammable liquids in plastic containers are

examples of research in this category.

Criteria for design of sprinkler systems are specifled in NFPA 13, Standard for the Installation of Sprinkler Systems, in a form that can be applied easily by designers to

typical buildings. The design method is based on classifying occupancies according to the activity conducted in the space, the quantity and type of fuel present, the potential for fast development of fire and the height of storage. Occupancies are classified as "light", "ordinary" or "extra" hazard accordingly. A serious drawback of the NFPA 13 hazard classification system is that occupancies that might be considered "light hazard" because of the activity that takes place (offices for example) may in fact have combustible contents that produce intense, fast-growing fires similar to "extra hazard fuel loads. In the last ten years, research has been done at FMRC to replace arbitrary classifications of hazard by occupancy with a classification system based on the heat release rate of the fuels. The required rate of application of water to control or suppress fires at the expected heat release rate is then determined through experimentation. Concepts of "Required Delivered Density

Score

*

14

14

15 Research Opportunlty

4.1. Continue to develop expertise with HAZARD I,

PHOENIX and other computer models to evaluate

heat transfer in water sprays.

*

4.2. Promote cooperation with FMRC to contribute to vaUdarion of plume-spray-heat transfer models with full-scale experiments. Ability to measure spray characteristics prerequisite.

4.3. Compare- to very f i e sprays injected

into the combustion air supply, versus heat transfer of larger drops penetrating a Fie plume

from above. Contrl- butlon 3 3 5 CbaI- lenge 5 5 4 Utlllz- ntlon 3 3 3 Fund- ing 3 3 3

(18)

(RDD)" and "Actual Delivered Density (ADD)" arise from this avenue of study. The sprinkler system designer must then ensure that the Actual Delivered Density meets or exceeds the Required Delivered Density for the given fuel load.

Experimentation has been done at FMRC to measure heat release rates of different commodities, the rate of application of water to the sprinkler and the quantity of water that actually penetrates the fire plume to reach the base of the fire (ADD). Testing has been

done through the NFPRF at FMRC and at Underwriters' Laboratories Inc. (ULI), to

develop Early Suppression, Fast Response sprinklers (ESFR) for high challenge fires in industrial occupancies (Carey, 1988, 1989; Fleming 1989). This research is already

resulting in improvements to the technology for the design of sprinkler systems. However, it is als6demoistrating that, despite the appkaling logic of matching water application rate to heat release rate. a meat deaI more testing will be required before all of the complexities of Fire control by water sprays are understd. ~ffectiGe f i e control involves -

understanding droplet size distribution, droplet momentum, fire plume dynamics and heat transfer to water sprays.

A concept that is closely related to plumespray interactions is that of "suppressibility." The recent work at ULI on development of ESFR technology has revealed that the

geometq of the fuel package effects the likelihood that the sprinklers will be able to

suppress a fire (Fleming 1989). In other words, the fuel properties alone are not a reliable basis for determining the required rate of water application. The configuration of the fuel array must be taken into account, not only in terms of the degree of shielding it may present for the sprinklers, but because it greatly influences the dynamics of the fire plume, which in turn determines the droplet momentum required to reach the seat of the fire. Much more research will be required to properly identify all of the factors that determine

"suppressibility" of specific fuel configurations, before Required Delivered Densities (RDD) can be confidently related to actual fire hazards.

The current research on new design concepts for high challenge industrial fires is being led by the American fire research community. Its approach is based on the use of ceiling sprinklers and high water-application rates with high-momentum droplets to penetrate the

fire plume. The preference for ceiling sprinklers comes about because of industry concerns

that "in-rack sprinklers hamper the use of the rack system for storage. Ceiling systems are costly, however, because they have a high water demand, both in terms of flow rate and pressure. Alternative approaches to fire protection of high-rack storage were examined at the Fire Research Station in the U.K. in the 1970's (Bridge and Young, 1974). Open sprinklers installed strategically in the storage racks, individually supplied with water by means of a solenoid valve activated by specially located heat detectors, appeared to have resulted in control of the fire at an earlier stage with much lower water demand. Bridge and Young (1974) estimate their approach could suppress a fire with approximately 113 of the flow required in the American approach. The potential demonstrated in the British

experiments for a more economical sprinkler system appears to have been obscured by the recent American focus on ceiling sprinklers.

The National Fire Protection Research Foundation Foam-water sprinkler project that was previously mentioned (Section 2.1) represents another example of research aimed at modifying sprinkler system performance to handle high heat release rate fires.

5.2 fRes ch i q

It will be difficult for the NFL to move into the forefront in this area because of the established momentum of the American research program. Contribution to the international data base on heat release rates of various fuels would require a large "products of

combustion collector" similar to FMRC's. Measurement of Required and Actual Delivered Densities for specific fuel packages requires the participation of industries willing to donate

(19)

large quantities of materials to be tested. The final authority in North America for establishing design criteria for sprinkler systems is shared between NFPA and FMRC. Any research done by the NFL would have to be supported by one of those organizations to be recognized.

The concept of suppressibility should be kept in mind during any sprinkler testing done at the NFL. There is strong interest in the international community in understanding the factors that affect suppressibility, so that they can be accounted for in the sprinkler system design process. Thus, although both the challenge and potential to make a significant contribution are relatively high, considerable time would be required to penetrate this research field. Since most of the major client groups are located in the

U.S.,

it would be relatively difficult to obtain funding for this work from Canadian sources, although a joint research uroiect with either ~~~~ ~

NFPRF

or FMRC is ~ossible. The NFL may be able to

I * ; ~ ~ ~

contribute in this area as a result of research done ostensibly for other p;rposes. Work done cooperatively with other organizations, to validate computer predictions of fire plume dynamics, for e x h p l e , may have secondary value in this area.

The preceding research opportunities in Category 5, Fire Growth Rate and Sprinkler Performance, can be summarized and ranked as shown in Table 5.

Table 5

5.3 Recommendations

If the NFL is to involve itself at a l l in sprinkler-related research, effort will have to be expended to stay abreast of the new technology directions established by the American sprinkler community. Therefore, Item 5.1 in Table 5 is marked with an asterisk because it is a prerequisite investment of time and expertise in order to maintain a role as an

organization capable of credible sprinkler-related research. Continued participation on the NFPA 13 Committee is the best way to accomplish this.

Research in items 5.2 and 5.3 could wait for the emergence of an appropriate

Research Opportunity

5.1 Keep abreast of changing technology in NFPA 13, new design concepts for sprinkler systems based on ADD, RDD, heat release rates.

*

5.2 Design experiments to examine "suppressibility"

-relating fuel characteristics, fuel geometry, and building features.

5.3 Resolve problems with in-rack sprinklers so that lower water demand systems can be used as an alternative to ceiling sprinklers.

opponunity, such as an interested client or request for a research propod

from

one of the American research aeencies. The NFL could uarticiuate with NFPRF or FMRC on a Joint

Contrl- button 2 4 3 Cbai- Lenge 3 4 4

Research Project bas%.

6.0 SPRINKLER ACTIVATION AND PERFORMANCE

Utlllz- atlon 4 3 3 6.1 Overview

This category of applied sprinkler research seeks to assess sprinkler performance in terms of time to activation and the number of sprinklers likely to be opened during a

tire.

f i n d - ing 2 2 2 Score *11 13 12

(20)

The objectives are to provide input to fire models and to find limits to sprinkler effectiveness. Research issues included in this categoty are: thermal sensitivity and response time of sprinklers; residential sprinkler systems; effects of obstructions on sensitivity and discharge pattern; sprinklers in concealed spaces; number of sprinklers opened; and estimating the impact of sprinklers on life safety.

Prediction of the time to activation of a sprinkler (or any other type of fire detector) is important for computer modelling of fires and occupant response in buildings. The rate of heat transfer between hot gases and the heat sensitive element of the sprinkler itself

influences the time to activation of the sprinkler. The sensitivity of the element is measured as the "response time index" (RTI) for the sprinkler. There are several methods of

measuring the RTI of a sprinkler, as for example the "plunge test" used at Factory Mutual in the U.S. (Theobald, 1989; Heskestad, 1989), and the "ramp test" preferred in the U.K.

.

(Theobald, 1989; Beever, 1989). In either case, the time to activation of the sprinkler is

determined by the rate of convective heat transfer between the hot gas and the heat sensitive element. The auestion as to which avvroach is most "realistic" in providing a measure of the expected response time for spri&ers is not particularly impc?:iant. w h i t is important is that sprinkler sensitivity can be quantified for use in computer models of fire development in buildings.

Research into sprinkler sensitivity has resulted in the most significant development in sprinkler technology in the last 25 years: "fast response" (or "quick response") sprinklers. Reduced lag time in the operation of the sprinkler has changed the focus of sprinkler systems from strictly property protection to life safety. Developed originally to make sprinklers effective for life safety in small residential spaces, fast response sprinklers are now being installed in all sorts of occupancies, including commercial buildings, industrial properties and hotels. There is a perception among designers and building and fire authorities that the reduced response time of sprinklers will significantly improve the overall performance of sprinkler systems. The fire will be detected while much smaller, hence the hazard to occupants of the buildings will be lower, as will the demand on the water supply. However, building codes, which have traditionally given significant credit to sprinkler systems as a fire safety measure in buildings, base those credits on the performance of regular sprinklers. The question that now arises is: "To what extent should building codes change their credit allowances for sprinklers, in recognition of the improved perfmance and reduced risk-to-life?" Research and experimentation will be needed to answer this difficult question. New interest in rehabilitation of existing buildings (upgrading to current building code standards) is likely to be a driving force behind such research.

The thermal sensitivity of a sprinkler depends on more than the design of the sprinkler itself. The proximity of obstructions, special ceiling constructions, and ambient

temperature conditions all have an influence on the ability of the sprinkler to operate. The potential for sprinklers in narrow concealed spaces to operate effectively has already been examined at the ~ ~~ - - ~

NFL

~ ~~ ~ b u g h e e d and Richardson. 1989). It is difficult. however. to ~~~ ~

-

~ -

-: - ~ ~ ~~ ~

generalize the results dfexpenments done to exarhne specific obstruction problems. Nevertheless, modifications to standard installation details are often necessary for unique buildings or situations that are not addressed by generalized installation guidelines.

Specially designed sprinklers, such as extended coverage sidewall sprinklers, have been developed to make it less costly to sprinkler existing buildings. The research and development for such products is mostly done in the U.S., and is conducted by such agencies as

ULI

as part of their listing process. For extended coverage sidewall sprinklers, even with fast response links, it can be expected that the worst-case fire will grow to larger magnitude than would be the case with regularly spaced sprinklers. As a result, smoke eenerdtion will be sieniFi~:antlv in~~eased.

In

such cases, il is itxloorlarll lo cotxlbiile the use of smoke detectors to providekarly warning for occupants with ihe sprinkler system (Bill.

(21)

Kung, Brown & Hill, 1988). This illustrates a situation in which the sprinkler system must operate in concert with other detection and alarm elements in the building. The importance of the proper functioning of complementary systems is sometimes overlooked by building officials and building operators. Any re-examination of the credits offered in building codes for installation of sprinkler systems as a remedial life safety measure should be based on a full understanding of the limitations of new technology.

Other issues related to sprinkler performance in buildings are being examined by researchers worldwide. One example is the use of sprinklers around openings between interconnected floor spaces or rue compartments to prevent passage of smoke, heat and flame (O'Neil, 1978). The use of directed spray nozzles to prevent fire spread through conveyor openings has long been of interest in industrial settings. An issue that has been identified by researchers working with smoke control systems, but which has not yet been adequately investigated, is the effect that sprinklers have on smoke movement in high buildings.

6.2 Research Opportunities for the National Fire Laboratow

Building officials, and Building Code experts are interested in how the use of fast response sprinklers in lieu of standard sprinklers will improve the performance of sprinkler systems. If substantial improvement is evident, it may be possible to increase the credits given in the Building Code for sprinklers as a fire safety feature in buildings. The NFL should determine how to quantify the differences between systems using fast response sprinklers and regular sprinklers in terms of enhanced life safety. The Risk Cost Assessment Model. uresentlv under develo~ment - - - - ~ ~~~ at the NFL. is an ideal instrument for examining this ques60n. experimenktion to verify act;al times to activation of sprinklers for given fie growth rates and building geometries may be needed. The results of the experiments would be compared to calculated response times based on the Response Time Index of the sprinkler and measured velocities of hot gases.

The study of smoke control methods is already part of the Nn program. That program could be extended to investigate the effects of sprinkler spray on smoke

movement, using the 10-storey tower at the Almonte field station. One aspect to study is how much the sprinkler spray reduces the temperature of the smoke and slows its spread through the building. Another question that could be answered through computer modeling and experimentation involves evaluating the effectiveness of water curtains (continuous water spray) around escalator openings in interconnected floor spaces, and the effect they have on smoke movement in the interconnected spaces. Examination of these questions would lead to improved design parameters for smoke control systems in buildings.

Requests to provide experimental information on the performance of sprinklers under particular conditions are occasionallv received bv the NFL. For e x m l e . a recent studv hone at the NFL involved evaluating the effects on the sprinkler spraybattern of

obstructions created by web members of wood trusses. Similar tests could be done to test the effect of ceiling obstructions on the time to activation of sprinklers. The ability of sprinklers to control a f r e within combustible truss assemblies could be investigated as well. The possibility of obtaining funding from the wood truss industry for this type of study is relatively high, yet the investment in equipment and training is relatively low.

The preceding research opportunities in Category 6, Sprinkler Activation and Performance, can be summarized and ranked as shown in Table 6.

(22)

-

17- Table 6

6.3 Recommendation$

Because of the close link between the NFL and the National Building Code of Canada, the NFL is strategically well placed to make a sign*cant contribution to building

technology in Canada by studying the issue of how to credit sprinklers as fire safety features in buildings. Item 6.1 in Table 6, therefore, represents a research area in which the NFL should increase its activity, despite the fact that it may not be that easy to obtain outside funding. This item is closely related to the matter of evaluating equivalencies to Building Code requirements, which is discussed under Category 8.

Item 6.2 involves examination of the effect that sprinklers have on smoke control systems. Every effort should be made to capitalize on the unique facilities at Almonte to contribute to this field of study.

Item 6.3 includes research questions for which it will be necessary to solicit clients. The wood industry, particularly truss manufacturers, could be expected to support research in this area. Efforts should be made to cultivate industry support for research in this area

All three items have overall scores greater than 15, hence merit consideration as important research opportunities.

7.0 SPECIAL WATER SPRAY APPLICATIONS

7.1 Overview

Many of the papers on sprinklers and water-based fire suppression systems reviewed for this study deal with applications of water spray technology to unique fire protection problems. The studies report on experiments done to demonstrate that sprinklers will control a particular fire hazard and to obtain information to establish design criteria for such systems. The concepts of optimum droplet size, the physics of suppression, fire

dynamics, and heat transfer must be understood and applied in order to determine that water spray is a feasible suppression agent for the problem at hand. In some cases, the

f i n d - ing 3 5 4 Research Opportunlty

6.1 Evaluate incremental benefit to existing NBC credits for sprinkler systems in buildings of using fast response. sprinklers in lieu of regular sprinklers.

6.2 Investigate the effect that sprinklers have on smoke movement in buildings, with focus on design aspects of smoke control systems. 6.3 Continue to respond to client requests for testing

performance factors of sprinklers under specified conditions. For e.g.

-

study effect that ceiling channelization, such as caused by deep composite wood trusses, has on the sprinkler sensitivity and perfonnance.

experiments yield only empirical information to guide in designing systems. In other cases, the experiments provide information at " f i t principles" ievel and conaibute to the

Score 15 17 16 Contrl- bution 4 5 3 ChaI- lenge 4 4 5

advancemeni of theore6cal work in plume dynamics, heat transfer and suppressibiiity.

Utlllz-

atlon

4

3

Figure

Table  9  Compilation of potential research topics in water-based fwe
Table 9.  Compilation  of  research topics in sprinkler and water spray  fue  suppression
Table  9.  Compilation of  research  topics in sprinkler  and  water spray fire suppression
Table  10.  Summary  of  research  topics  in  water  based  suppression systems recommended

Références

Documents relatifs

61 Universit` a di Pisa, Dipartimento di Fisica, Scuola Normale Superiore and INFN, I-56127 Pisa, Italy 62 Prairie View A&amp;M University, Prairie View, Texas 77446, USA.. 63

sensitive to the alloy composition (or surface topology/sphericity) as shown in Figure 6b while the Hall flow appeared to be unaffected by the composition. Figures 11c and 11d

To test the functional role played by altered levels of acH3K14 in the NAc, groups of mice subjected to chronic (10 d) social defeat stress and nonstressed control mice were, on day

Uncertainties in the parton distribu- tion functions (PDFs) are estimated using the set of 2×20 errors provided for the CTEQ6M [23] PDF. The analysis consistency uncertainty

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des

Thus, absence of Rad50 protein can be expected to result in increased levels of chromosome and chromatid breaks, which would agree with the 49% of the anaphase chromosome

For the fastener test coupons, life improvement factors were also calculated on crack growth from initiation to coupon failure.. This was done to partly isolate the scatter

height from radiosonde soundings, For a range of initial radii and vertical velocities several runs of the model were made using as cloud base both the lifting