j002_001r00_investigacion_bibliografica.pdf

Discharge of Fire Suppression Agents to

Protected Areas

Bibliography Research and & the State

of the Art

F.Javier Garcıa Garcıa

31st January 2011

1

2

Abstract

This paper aims to show the present State of the Art regarding the useof clean agents, with special emphasis on HFC-227ea, for the suppressionof fires within enclosed areas. This method of suppression is based on thefast discharge of the N2-pressurized clean agent contained in industrialcylinders, through a net of pipes, into de protected room.

To this end, the existing literature covering the subject of fire ex-tinguishing systems based on clean agents is surveyed and a synopticevaluation presented. Special emphasis is put on papers covering the hy-drodynamics of clean agent discharge, as this will be the subject of acomplementary work to be presented by the author.

Initially a brief account of the historical reasons leading to the search ofclean agents is offered, namely the Earth’s ozone layer depletion; then theproblems of high Global Warming Potential (GWP) of HFC’s is discussed.Next, the main general principles on which this fire suppression techniqueis based are discussed.

The literature is reviewed, covering sequentially the subjects of hy-drodynamics of agent’s discharge, clean agent’s extinguishing efficiency,friction models and thermophysical properties of HFC-227ea. Finally,some conclusions are drawn regarding the necessity of developing mathe-matical models for agent discharge, what justifies the elaboration of onesuch model in a separated research work.

Keywords: fire suppression, clean agent discharge, transient fluiddynamics, HFC-227ea, Halon replacement

CONTENTS 3

Contents

List of Figures 4

1 Introduction 5

2 Description of the Technique 9

3 Literature Review 183.1 Hydrodynamics of Discharge . . . . . . . . . . . . . . . . . . . . 183.2 Extinguishing Efficiency . . . . . . . . . . . . . . . . . . . . . . . 293.3 Considerations About Friction . . . . . . . . . . . . . . . . . . . . 313.4 Thermophysical Properties of HFC-227ea . . . . . . . . . . . . . 32

4 Conclusions 34

Bibliography 35

Appendices 42

A Properties of HFC-227ea 42

LIST OF FIGURES 4

List of Figures

1 The largest Antarctic ozone hole recorded as of September 2006. 62 Thermophysical properties of most used extinction agents. . . . . 83 Chemical formula of HFC-227ea molecule. . . . . . . . . . . . . . 104 3-D image of HFC-227ea molecule. . . . . . . . . . . . . . . . . . 115 Typical battery of HFC-227ea cylinders for fire extinction appli-

cations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 A nozzle discharging clean agent to atmosphere. . . . . . . . . . 127 Typical small installation of HFC-227ea fire extinction. . . . . . . 138 A Typical installation to protect an engine’s room. . . . . . . . . 149 Vapour pressure curve of HFC-227ea. . . . . . . . . . . . . . . . . 1510 Discharge of HFC-227ea agent in a protected room. . . . . . . . . 1611 Isometric diagram for HFC-227ea superpressurized with N2 to

2500 kPa at 21�. . . . . . . . . . . . . . . . . . . . . . . . . . . . 1712 Isometric diagram for HFC-227ea superpressurized with N2 to

4150 kPa at 21�. . . . . . . . . . . . . . . . . . . . . . . . . . . . 1713 Experimental arrangement in Pitts et al. ([PYG+94]) experiments. 1914 Temporal variations of internal pressure during downward dis-

charges of HFC-227ea. . . . . . . . . . . . . . . . . . . . . . . . . 2015 Comparison of measured pressure-time history of a HFC-227ea

discharge with prediction. . . . . . . . . . . . . . . . . . . . . . . 2116 Two time traces of near-field dynamic pressure for releases of

HFC-227ea. Trace B is offset by 100 kPa. Conditions: A, 476 g,release pressure 4.45 MPa; B, 489 g, release pressure 4.41 MPa. . 22

17 Dynamic pressure measured 1.3 m from the vessel for HFC-227ea.For clarity, trace B is offset by 100 kPa. Conditions: A, 489 g,release pressure 4.41 MPa; B, 476 g, release pressure 4.45 MPa. . 23

18 Pressure variation inside the cylinder and at nozzle for HFC-227ea discharge. . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

19 Pressure variation inside the cylinder and at nozzle for clean agentdischarge. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

20 Actual and estimated pressure-time behaviour in a tight 34 m3

room upon HFC-227ea discharge to yield 5.9%. . . . . . . . . . . 2721 Development of room pressure during the discharge of 30.9% v/v

IG541. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2822 Development of room pressure during the discharge of 6.1% v/v

HFC-227ea. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2823 Predicted and experimental enclosure pressures for a 8 s, 9.8% of

HFC-227ea discharge. . . . . . . . . . . . . . . . . . . . . . . . . 2924 Temperature comparison for HFC-227ea extinguished fires with

and without WSCS. . . . . . . . . . . . . . . . . . . . . . . . . . 3125 Extinction system with very small length of discharge’s pipe . . . 34

1 INTRODUCTION 5

1 Introduction

This report aims to summarise the current state of the art regarding the hydro-dynamical modelling of processes in which a liquid is kept highly pressurisedby a seemingly ideal gas, both contained within a vessel provided with a fastvalve, which is suddenly opened and the liquid released to atmosphere at highspeed through a length of discharge’s pipe. A special emphasis is made on thetheoretical and mathematical models which attempt to explain such phenomena.

In the realm of fire protection engineering it is not infrequent the use of highlypressurized extinguishing agents that are suddenly discharged to the protectedareas in case of fire. This very effective method of fire extinction is carriedout with agents that normally are in gaseous state at standard conditions ofpressure and temperature, although within the pressurized vessels can very wellbe in liquid state, and thus modelled as incompressible fluids.

A number of years ago Halons were the main (if not the only) extinction agentsused to prevent the initiation of flames within small-to-medium rooms in whichfires were deemed specially catastrophic, such as computer rooms, critical docu-mentation vaults, museum’s restoration halls, process control rooms, etc. Theywere so effective in stopping the combustion reactions, that little discussionexisted about the possibility of finding some other kinds of extinction agents.

As it was established the role played by BFC’s and CFC’s in ozone layer deple-tion, and it became evident that these compounds were doomed for any indus-trial application as a result of 1987 Montreal’s Protocol, a vast scientific andengineering endeavour was originated with purpose of finding new substancesthat could substitute Halons in such important applications. Thus, there wasestablished the Halon Options Technical Working Conferences (HOTWC), orthe International CFC and Halon Alternatives Conferences and Exhibitions,and similar scientific and engineering meetings. Probably, the most ambitiousprogram was the Next Generation Fire Suppression Technology Program (NGP)of the National Institute of Standards and Technology (USA), which ran from1991 to 2006 under the supervision of renowned scientists like Richard G. Gannand others (see [Gan95], [Gan98], [Gan03], [Gan05])

1 INTRODUCTION 6

Figure 1: The largest Antarctic ozone hole recorded as of September 2006.

As a result of such standing and extensive research, a number of agents werefound that fulfilled acceptably the role of Halons for fire extinction applications,and the literature is full of articles, papers and proceedings describing them (see,for instance, [Coo94], [GGP94], [HGG+94], [HTSC94], [BGF+95], [MDK98],[NBHT98], [GHCP00], [Sas01], [RRC01], [Lin06], [PYB+06] and [BBFY08])

These extinction compounds are also called Clean Agents, because they haveno noticeable effect on Earth’s ozone layer. Some of these agents actually showeddesign concentrations to quench test fires not too different from those corre-sponding to Halons. This last feature is quite desirable, because it means that,in most cases, the existing installations (pipes, valves, manifolds, etc.) could bepreserved and only the actual agent’s cylinders had to be replaced, thus makingmore affordable the initiative of banning Halons and having them replaced byless environmentally hazardous compounds. Notably, one of the most suitableagents (with an acceptable price tag) was found to be HFC-227ea, which canreadily be used as a replacement agent in existing installations of Halon 1301(see Senecal 2001, [Sen01]).

But once the problem of Earth’s ozone layer depletion was given a solution,the world became conscious of a new threat to the planet: the global warming

1 INTRODUCTION 7

of atmosphere and the climate change it might engender. A great numberof studies were conducted in order to find the greenhouse effect attached toany gaseous substance that might be released to atmosphere. Nowadays everyindustrial compound exhibits in its Material Safety Data Sheet the degree ofglobal warming potential it carries.

The most effective clean agents, from extinction power’s point of view, normallyexhibit a very high degree of global warming potential (GWP), several thousandtimes higher than CO2 (see figure 2). With raising worldwide concern aboutclimate change, it is only natural to try to find agents as effective as those,but much less hazardous for the environment. While this research is being con-ducted, with varying degree of success, growing concern arises in rationalizingthe use of existing agents, with the aim that just the strictly sufficient amountto extinguish a potential fire is released to the atmosphere, and nothing else.There exists founded suspicions that, probably, present regulations and stan-dards concerning such fire protection techniques, do not adequately address thenecessity of reducing the amount of agent to be released to environment. to theactual minimum which guarantees fire extinction.

Thus, in the long term, those Clean Agents were not so clean, after all, and anew endeavour is again necessary to replace them with other agents that couldundertake the task. In the meantime, the emphasis is set on techniques thatmight improve the efficiency of existing agents, allowing for smaller amounts ofthem to guarantee the same extinction performance. This could very well beobtained with improved methods of discharge, that take into account the specifichydrodynamic features of this process. This justifies the necessity of developingnew models that could be used in the design of better systems, which fullyexploit the intrinsic capabilities of agents to quench fires.

1 INTRODUCTION 8

Figure 2: Thermophysical properties of most used extinction agents.

Although it there exist many empirical research regarding the effectivenessof these agents to extinguish various kinds of fire (see, for instance, [Coo94],[HGG+94], [HTSC94], [KSMKK96], [MDK98], [KOYS01], [RRC01], [Sas01],[Ska02], [Fri03], [SAA+03], [SAAM04], [Ben06], [Lin06], [PYB+06] and [YK06].among others), to our knowledge, very few studies have been done regarding thehydrodynamics aspects of the phenomenon, specially from the transient pointof view. It is our opinion that a deeper understanding of the dynamical processof agent release, will help in reducing and optimizing the amounts of agent thatpresently are regulated and established as suitable for extinction applications.

The information summarised in this State of the Art report is gathered with theintention of producing a work devoted to such endeavour: to provide a physico-mathematical model, and the associated dynamical equations, which will permit

2 DESCRIPTION OF THE TECHNIQUE 9

the description, comprehension and predictions of the free discharge process ofa highly pressurized liquid agent into atmosphere. It is also expected that thismodel will open the possibility of incorporating the results with it obtained,into the actual regulation and standards which have relation with this field offire protection engineering. The actual model will be presented in a separatepaper, while this work is devoted to investigate the findings that exist in theliterature regarding this phenomenon.

Although the model could be used for a number of extinction agents, it willbe particularized for HFC-227ea, as it is of widespread use in industry and itsGWP is moderately high1. With minor changes of values (viscosity, density,etc.) it could be applied to other agents.

Although this problem has been approached before (see, for instance, Pittset al. 1994, [PYG+94], or Cooper 1993, [Coo93]) the treatment found in theliterature has always been from a hydrodynamic steady state perspective. To ourknowledge, there is not yet a model for the problem’s transient hydrodynamicbehaviour. It is our opinion (and the results might confirm it) that such a fastprocess as this agent’s discharge (which normally takes less than 10 s, with avery fast evolution in the first second) could not be accurately described withthe equations of steady state hydrodynamics. Instead, full transient regimeequations must be used, which will allow for new and interesting hydrodynamicbehaviour to emerge.

Although simple, the proposed model still has taken into account most aspectsthat are relevant for this application:

1. Friction.

2. The compression of an ideal gas as the driving force which sets the liquidinto motion.

3. The liquid’s finite volume within the vessel.

4. The phenomenon’s very fast nature and the transient treatment of it.

A very recommendable handbook dealing with the general principles of transientfluid dynamics is Durst 2008, [Dur08].

2 Description of the Technique

A very well tested technique for fire-fighting, used throughout the world, consistsof discharging an agent very rapidly inside the room which is meant to beprotected. It is specially adequate for the very initial stages of fire development,

1HFC-227ea has a GWP of 2900, this is, 1 kg of HFC-227ea has the equivalent globalwarming effect of 2.9 tons of CO2. Other extinction agents have a much higher GWP: forinstance, HFC-23 has 11700 (14800 for a 100 years period). See thermophysical properties ofHFC-227ea in appendix A.


According to the nature and/or amount of agent released to the ambient, thefire is quenched by one or more of the following mechanisms2:

Heat removal: The combustion reaction is halted due to sudden decrease oftemperature, below flammability limit. The agent concentration does notneed to be very high, as long as it exhibits enough cooling properties.

Oxygen removal: The chemical reaction is halted due to the scarcity of abasic reactive. The agent’s concentration needs to be very high, as itmust displace oxygen from the room below the suffocation limit.

Reaction hindrance: The combustion reaction is hindered, and eventuallyhalted, due to new chemical compounds that interfere with the radicalscreated in the reaction chain. The agent concentration does not needto be very high, as long as it has the property of forming intermediatecompounds with these radicals.

Carbon-based clean agents relies primarily on the first mechanism, being theothers of secondary importance. Inert-gas clean agents use mostly the secondmechanism. The third mechanism is typical of Halons.

The agent primarily concerned with in this work is HFC-227ea, also known asR227, Heptafluoropropane, HFP, C3HF7, CF3 − CFH − CF3, etc. It is alsocommercialized under the trade registered names of FM-200, FE-227, Solka-flam 227, MH-227, NAF S 227, etc. It is a colourless and odourless gaseoushalocarbon.

Agent HFC-227ea adopts the chemical formula depicted on figure 3. It hasa boiling point of -16.4 � at atmospheric pressure, and it is a gas at roomtemperature. It is slightly soluble in water (260 mg/L).

Figure 3: Chemical formula of HFC-227ea molecule.

2A fourth obvious mechanism, removing combustibles from flame, is not mentioned becauseautomatic fire extinction systems seldom use it.


The molecule CF3 − CFH − CF3 is a reagent to generate sources of radicalCF3 − CF− − CF3 by deprotonation. This ability is clearly understood byconsidering its 3D molecular model, figure 4

The basic features of automatic fire extinction systems based on HFC-227eacould be found in general reference handbooks such as SFPE’s [DiN02] andNFPA’s [DiN03]. Some standards that regulate the design, installation, useand maintenance of such systems are: NFPA-2001 ([NFP08]), ISO-14520:2006([ISO06]), CEA-4045 ([CEA05]), VdS-2381 ([VdS09]), etc. In Spain the regula-tions are UNE-EN 15004-5:2009 ([UE09a], equivalent to ISO-14520) and UNE-EN 15004-5:2009 ([UE09b]).

Figure 4: 3-D image of HFC-227ea molecule.

Commercially, the agent HFC-227ea is supplied in cylinders of different stan-dardized volumes, normally at 24 bar of relative pressure at normal room tem-perature. The pressure is maintained with compressed gas N2, being bothHFC-227ea and N2 within the cylinder. Should the application demand morethan one cylinder, several of them could be deployed together forming a battery,as shown on figure 5. In such cases, normally the pressure of one cylinder isused to trigger the next one in the battery.

The cylinders are connected, via rigid pipe (normally with high pressure rat-ing), to a number of nozzles which warrant the adequate distribution of agentwithin the protected room (see figure 6). The trigger which releases the agent isnormally an automatic fire detection system, usually of the very-early-warningtype.


Figure 5: Typical battery of HFC-227ea cylinders for fire extinction applications.

Figure 6: A nozzle discharging clean agent to atmosphere.

Figure 7 shows a typical small installation that demands just one cylinder toguarantee fire extinction. The elements which constitute an automatic fire sup-pression system based on clean agent are:

� Agent cylinder. Clean agent and N2 are both contained inside the cylinder,both compressed to the design pressure. The cylinder is endowed with amanometer.


� Valve. This fast opening valve releases the agent to the ambient. It couldbe electrically or pneumatically activated. Usually the cylinder has amanual release lever which acts directly on this valve.

� Piping. It conducts the agent to the discharge nozzles.

� Discharge nozzles. Allow for the agent to vaporize, and distribute it uni-formly enough inside the protected room.

� Fire detectors. It could detect the presence of smoke, rising temperature,flame, or any other property associated with fire. It sends an alarm signalto the fire detection panel.

� Detection panel. Called Releasing Panel in figure 7, it is the electronicunit which receives the early-warning alarm signal from fire detectors,and triggers the agent’s release, according to whatever sequence has beenprogrammed.

� Horn/strobe. An audible/visible warning signal for the people inside theroom, announcing that agent’s release is coming. People must leave theroom and lock the door upon hearing this horn.

� Manual release. A push-button which sends a signal to the DetectionPanel, causing the release of agent to the room.

Figure 7: Typical small installation of HFC-227ea fire extinction.

A similar installation with a battery of cylinders is shown in figure 8, in whichthe same basic elements described above are present.


Figure 8: A Typical installation to protect an engine’s room.

Fire suppression with agent HFC-227ea is mainly based in its ability to vaporizeand rapidly remove heat from its surroundings. This causes a sudden temper-ature decrease in the protected room, which plunges below the flammabilitylimit, thus quenching any attempted fire that might have initiated.

Figure 9 shows the vapour pressure of HFC-227ea as a function of temperature.For a normal room temperature (some 21 �) this pressure is nearly 5 bar, whatexplains the flashing behaviour of HFC-227ea as it exits the discharge nozzles.In the excellent photograph of figure 10 it can be seen how the agent rapidlyvaporizes as it displaces some 1 m away from the nozzle.


Figure 9: Vapour pressure curve of HFC-227ea.

The calculation of heat removal from ambient as HFC-227ea vaporizes duringthe discharge process, is suitable described in general handbooks such as SFPE’s[DiN02] and NFPA’s [DiN03]. This cooling effect is not the only suppressionmechanism taking place in agent’s discharge: the high speed fluid also removespartially some oxygen from the fire, and disperses and mixes with pyrolysisproducts, thus interfering with combustion reaction.


Figure 10: Discharge of HFC-227ea agent in a protected room.

It is worth mentioning the interesting experimental work related with fire sup-pression, Blanchat et al. 2008, [BBFY08], which shows with detail the successivestages which takes place in a typical fire extinction process. The same issue,from a CFD simulation perspective, is also approached by Hewson et al. 2003,[HTSD03]. In this work the suppression modelling is based on the ratio be-tween the fluid mixing time scale and the flame chemical time scale. Flames areextinguished when the fluid mixing time is short relative to the chemical timerequired for combustion to occur. This justifies the importance of designingdischarge systems that confer a high velocity to the agent.

The main mechanism to achieve the desired discharge velocity is to N2-pressurizecorrectly the agent in the cylinders. Thus, a certain amount of N2 must be addedto obtain the right pressurization, but not an excessive one, as the pressurewithin the cylinders would raise rapidly as room temperature grows, perhapsabove the pressure rating of cylinders. The relation between N2 and HFC-227eainside the cylinder is called filling density (kg/m3).

In order to understand the importance of this pressurization, figure 11 showsthe pressure behaviour of a typical cylinder as a function of temperature, takinginto account several filling densities, This diagram has been drawn for a cylinderwith 2500 kPa at 21�. Alternatively, the agent could be initially N2-pressurizedto 4150 kPa at 21 �, and figure 12 shows the corresponding isometric diagram.

The importance of an isometric diagram is that it determines the maximum fill-ing density for an agent in a cylinder with a fixed pressure rating. The basic rule


is that the cylinder must not become liquid full at 54 � (for U.S. DOT) and/orthe pressure developed at 54 �must not exceed 5/4 cylinder design pressure.The pressure developed is a function of the agent, the superpressurization level,and the temperature. For 2500 kPa (at 21 �) cylinders, pressure is limited to5/4 of the design working pressure. which is 3445 kPa at 54 �. For HFC-227ea,as seen in figure 11, this yields a maximum filling density of 1153 kg/m3. The2500 kPa pressurization level is not based on characteristics of halon replace-ment halocarbons, but is a vestige of standard Halon 1301 systems.

Figure 11: Isometric diagram for HFC-227ea superpressurized with N2 to 2500kPa at 21�.

Figure 12: Isometric diagram for HFC-227ea superpressurized with N2 to 4150kPa at 21�.

3 LITERATURE REVIEW 18

A secondary mechanism to achieve a high discharge velocity, complementary tothe former, is an optimized piping design. It is thus necessary to have good dy-namical models of piping flow which take into account the transient behaviourof the discharge process. Also, it is important that those models describe ac-curately the dependence of the transient friction coefficient on flow parameters,as friction is the most important effect opposing agent’s motion.

Thus, it makes sense to develop such dynamical models of transient flow throughpipes, as they will be necessary in any effort to optimize current design ofautomatic fire suppression systems with clean agents. The following section isdevoted to assess the current State of the Art of most aspects of these extinctionsystems. A new mathematical model will be introduced in an additional researchwork, which derives the transient dynamical equations of agent’s flow directlyfrom the First Principles.

3 Literature Review

Although there exist a large number of works devoted to study and/or documentthe effectiveness of HFC-227ea (and other extinction agents) for various kindsof test fires, few articles can be found conducting extensive research on thehydrodynamics behaviour of the discharge process of such agents to protectedareas.

3.1 Hydrodynamics of Discharge

Grosshanler et al. 1994, [GGP94], is probably the most comprehensive workever written on the subject of agent’s discharge over fires in protected areas. Itcovers most areas of research which have a relationship with this issue: thermo-dynamical properties of agents, hydrodynamical modelling of discharge, hydro-dynamical experimentation of these phenomena, flame suppression effectivenessof such agents, flame inhibition chemistry, agent stability under storage anddischarge residue, corrosion of metals, elastomer seal compatibility, and humanexposure and environmental impact. It contains interesting developments fromthe theoretical point of view, viable numerical models, ingenious experimentalset-ups, and valuable results and findings.

The chapter devoted to hydrodynamics of agent discharge, written by Pitts etal. 1994, [PYG+94], is exhaustive and covers most aspects of this subject. Itdefines a mathematical model for an experimental set-up in which the processcan be studied and monitored, and the model’s predictions compared with themeasurement results. It is complete, full and detailed, and the model is quiteanalytical, but it is based on the equations found in Shapiro 1953, [Sha53], whichare developed under the assumptions of isentropic steady regime flow, and itdoes not take into account the transient flow equations which are deduced inShapiro 1954, [Sha54].


The work describes an experimental apparatus, shown on figure 13, and theassociated mathematical model that explains the agent’s behaviour as it is dis-charged through it. This approach to the problem considers the agent as acompressible fluid and makes use of the equations for steady isentropic com-pressible flow. The model takes into account the possibility of agent flashing.

Figure 13: Experimental arrangement in Pitts et al. ([PYG+94]) experiments.

It considers also a discharge flow coefficient for the orifice (nozzle) throughwhich the agent discharges to ambient, according to [Sha53]. The dimensionlessequations are then presented and solved numerically, for a set of initial valuestypical of standard extinction installations. The results are depicted in graphsand then compared to actual measurements. There is a comprehensive set ofcurves which shows most of the situations that can be subjected to experimentalcontrol.


The experiments are conducted with a number of different extinction agents,such as HFC-227ea, FC-218, FC-236, FC-31-10, HFC-32, HFC-125, Halon 1301,HCFC-22, HCFC-124, HFC-236fa, FC-318, FC-116 and HFC-134a. Here onlythe results concerning HFC-227ea will be highlighted.

One of such curves is shown on figure 14, which represent the results for thetemporal variation of the internal pressure during downward discharges of HFC-227ea. Pressures reported in the figure are gauge pressures. The pressure isnondimensionalized by the actual burst pressure, Pi, which is taken to be thepressure at t = 0 s. The curves show two distinct regions, separated by an in-flection point. The first region corresponds to the time interval during which theliquid agent is being propelled from the vessel. The second region correspondsto the period when the remaining vapour (mostly N2) is being vented from thevessel. The inflection point corresponds to the time at which the liquid agenthas just been completely expelled from the vessel.

Figure 14: Temporal variations of internal pressure during downward dischargesof HFC-227ea.

In the work there are abundant pictures taken from high speed cameras. Thesepictures show with a great level of detail the development of flow in the dischargeprocess.

Figure 15 shows the comparison of the actual measured values of dischargepressures for HFC-227ea with the model’s predicted results. Despite not usingtransient flow dynamical equations, the agreement is excellent.


Figure 15: Comparison of measured pressure-time history of a HFC-227ea dis-charge with prediction.

Perhaps the most interesting results, from the hydrodynamical point of view, areshown in figures 16 and 17. They correspond to the dynamical pressure curvesas measured by two pressure transducers: one located near the exit orifice (some13 mm downstream, figure 16), and another located further away (some 1.3 mdownstream of the vessel’s exit, figure 17). Although air entrainment alters thejet, far-field curves are not too different from near-field ones.


Figure 16: Two time traces of near-field dynamic pressure for releases of HFC-227ea. Trace B is offset by 100 kPa. Conditions: A, 476 g, release pressure 4.45MPa; B, 489 g, release pressure 4.41 MPa.


Figure 17: Dynamic pressure measured 1.3 m from the vessel for HFC-227ea.For clarity, trace B is offset by 100 kPa. Conditions: A, 489 g, release pressure4.41 MPa; B, 476 g, release pressure 4.45 MPa.

A research work very much in the line of [PYG+94] is Cooper 1993, [Coo93],although it studies a different agent, R22, with different properties than HFC-227ea. As remarked above, the discharge treatment is done under the assump-tion of steady compressible flow.

Also Pitts et al. 1993, [PYBG93], develops this same issue from an experimentalpoint of view, but it is less exhaustive than [PYG+94].

Wysocki 1996,[Wys96], studies also the discharge process of clean agents, in-cluding HFC-227ea. It uses Bernoulli equation in the Hesson’s adaptation form,and does not consider explicitly the transient nature of dynamics, although mostof the work is experimental rather than theoretical. Most dynamical curves areplotted with percentage of agent discharged in abscissa, rather than time.

There is just one curve with time in abscissa, namely the time variation of thepressure, both in cylinder and in nozzle, corresponding to the discharge test ofa cylinder of HFC-227ea with filling density of 1121 kg/m3, see figure 18. This


pressure versus time tracing taken during a HFC-227ea discharge shows thecontinual variation in cylinder pressure and a lesser variation in pressure at thenozzle during the course of discharge. The peak in the cylinder pressure at 2.5s indicates the beginning of flashing, in which agent’s flow changes from single-phase to two-phase. The peak in the nozzle pressure trace at 9.6 s indicates achange from two-phase flow to vapour flow.

Figure 18: Pressure variation inside the cylinder and at nozzle for HFC-227eadischarge.

DiNenno et al. 1994,[DFF+94], develops an interesting study of the clean agentdischarge’s process, based on the model developed for Halon 1301 by Elliot etal. 1984, [EGK+84]. It offers a mathematical model for the phenomenon, butthis model does not take into account the transient dynamical equations, and itrather considers some quasi-steady flow. The agent is assumed to flow in liquid,two-phase and vapour modes. It shows very interesting experimental pressuredata for the first few seconds of process, as it can be seen in figure 19. In thisfigure the discharge flow has been divided into five distinctive sections.


Figure 19: Pressure variation inside the cylinder and at nozzle for clean agentdischarge.

Zalosh-Wang 1996, [ZW96], develops a mathematical model based on mass andenergy conservation, in which no transient dynamical equation is considered.The article shows curves of HFC-227ea concentration versus time, both obtainedfrom analytical calculations and from actual test results. The agreement is goodin a long time lapse, but no comparison is made between experimental andtheoretical results in the first few seconds. Actually, the discharge concentrationcurves are modelled as straight lines. A theoretical pressure curve is offered,but it corresponds to room pressure, not to pipe flow pressure. Thus, scarceinformation is offered about the phenomenon’s dynamics.

Bird et al. 1994, [BGH+94], measures the discharge process of HFC-227eain an experimental set-up, and compares the results with a computer model,named Transient Flow Analysis (TFA), which is not described in the paper.TFA is presented and explained in Kim et al. 1994, [KLGS94], which describesthe computer model used to calculate the discharge process, but the programand/or numerical results are not shown on the paper. TFA is the result of amodification to an already existing computer program for modelling Halon 1301discharge, described in Coward et al. 1992, [CHM92].

TFA is further explored in Bird et al. 1995, [BGF+95], which studies the dis-charge of HFC-227ea in large piping distribution systems, with several cylinders(4 to 14) installed in battery. The discharge pressure is monitored in up to9 nozzles, and also inside the pipe. TFA’s predicted results are compared toactual measurements. No mathematical model is presented in this work.


The flow within pipes during a discharge process is studied in Cleary et al. 1994,[CGY94], which characterizes the flow as transient and two-phase. The workincludes an analysis of the discharge’s first fractions of a second, but shows onlythe experimental curves with no theoretical result nor computer program. Pipepressure traces, corresponding to a transducer located some 2 m away fromdischarge point, show an interesting transient spike phenomenon at around 700ms, probably related with N2 flowing alone in the pipe.

Using the same experimental set-up as in [CGY94], a new paper was publishedby Cleary et al. 1995, [CYK+95], in which the discharge process was studiedafter the agent exits the pipe and is released into test engine’s nacelles. Thepressure transient spike in the pipe transducers is also present, surely due to thepresence of N2 alone in the pipe as the liquid agent runs out of the vessel. Inthis paper the time range under study is longer than that in the previous work,but again no equations nor model was offered. It is based on the model forHalon 1301 discharge developed in the paper from Elliot et al. 1984, [EGK+84].

A computer model is profusely described in Tuzla et al. 2000, [TPC+00]. Theprogram is named FSP and it represents an improvement of the computer modelpresented in [EGK+84]. It covers transient phenomena and two-phase flow.The computer model is tested against the same experimental apparatus used in[CGY94] and [CYK+95], providing acceptable results when the predicted curvesare compared with the actual experimental data.

Also transient experimental phenomena are studied in Wysocki-Christensen1996, [WC96], although it treats inert agent IG5413, not HFC-227ea. It ismentioned here for its study on pure gas discharge, without consideration forflashing or two-phase flow. Some experimental curves are offered which couldbe compared with those corresponding to HFC-227ea.

A different approach is offered in Senecal-Prescott 1995, [SP95]. It shows theresults of experiments in which it is measured the pressure variation within aone-way tight room that undergoes the discharge of agent HFC-227ea. The pres-sure is monitored for relatively long times (some 30 s), compared to the actualdischarge’s time. An interesting outcome of this study is that pressure withinthe room turns negative (below atmospheric) for some time after discharge (seefigure 20). Also the agent dispersion throughout the room is observed andmonitored.

3IG541 is a blend of inert gases, Ar 40%, N2 52%, CO2 8%, and it acts suffocating theflames, this is, removing oxygen from fire. This agent is stored at very high pressures (normally200 or 300 bar) and is never in liquid state within the cylinders.


Figure 20: Actual and estimated pressure-time behaviour in a tight 34 m3 roomupon HFC-227ea discharge to yield 5.9%.

This same negative pressure phenomenon is reported on Robin et al. 2005,[RFS05]. This work offers a comparison in the pressure curves relative to dis-charge of IG541 (which is always in gaseous state) and HFC-227ea, which ex-hibits a flashing behaviour and vaporizes entirely. Figure 21 shows a positivepressure within the room at every instant after IG541 discharge, while figure22 shows alternatively positive and negative room pressure after HFC-227eadischarge.


Figure 21: Development of room pressure during the discharge of 30.9% v/vIG541.

Figure 22: Development of room pressure during the discharge of 6.1% v/vHFC-227ea.


The work also compares the actual test results with the predictions of a com-puter model, of which no details are given (see figure 23).

Figure 23: Predicted and experimental enclosure pressures for a 8 s, 9.8% ofHFC-227ea discharge.

Of a different nature are other works which deal with the transient hydrody-namics phenomena from another point of view, namely, these extracted fromthe rupture of a highly pressurized pipe. Although those processes are differentfrom the one considered here, some of the equations and the general treatmentof the problem might offer clues on how to approach the discharge phenomenastudied in this work.

Denton 2009, [Den09], is a PhD Thesis which focuses very much on transientflow, although it is not initially addressed to the Fire Protection EngineeringCommunity. It develops a CFD program to simulate highly transient flows, andit offers the results of such simulations for a number of interesting cases, notablythe discharge of pressurized fluids through orifices.

Other works which deal with highly transient flow in pipes, in a similar fash-ion than [Den09], are: Kim-E 1981, [KE81], Valero 1998, [Val98], Oke et al.2003, [OMER03], Mahgerefteh et al. 2006, [MOR06], and Nouri et al. 2010,[NBZR10].

3.2 Extinguishing Efficiency

A good general introduction to this subject is offered in Friedman 2003, [Fri03],to be found within the NFPA’s handbook [C+03]. The extensive work of


Grosshandler et al. 1994, [GGP94], still constitutes the most comprehensivestudy on every aspect of fire quenching with clean agents.

Hamins et al. 1994, [HGG+94], to be found within the book [GGP94], is anextensive work which analyses carefully most fire extinguishing mechanisms.It includes chemical models of reaction inhibition, comprehensive experimentalfindings, and a great number of curves which could be quoted to validate futuremodels in this field.

Bennett 2006, [Ben06], is a very advisable primer for readers who wish to take afirst contact with the subject. It clearly establishes the principles that determinewhat makes an agent good for fire extinguishing. Those principles are supportedon experimental evidence which is offered on the work.

Linteris 2006, [Lin06], offers a complete account of chemical flame inhibitionmechanisms, rather than fire suppression. It outlines the conditions for whicha chemically active agent can be effective, and when it is most effective. Thesegeneral principles are demonstrated with numerical and experimental data fordifferent kinds of fire.

Moghtaderi et al. 1998, [MDK98], studies the ignition delay caused in firesby agent HFC-227ea. It presents a number of experimental findings regardingflame inhibition, and characterizes some chemical reactions occurring in theinteraction of HFC-227ea with methane inside a flame.

Saso 2001, [Sas01], also studies the influence of inhibitors in fires, with specialemphasis on catalysis. When the concentrations of the fuel and the oxidizer ina combustible mixture are kept constant, the global reaction rate is controlledby the flame temperature, the global frequency factor, and the global activationenergy. These factors are changed and the results are plotted in numerousgraphs.

Kim et al. 1996, [KSMKK96], offers the results of a real extinction test withclean agents, namely HFC-227ea and HCFC Blend A4. conducted in a realinstallation under total flooding technique. The acid gases generated duringfire suppression are also measured and analysed.

Robin et al. 2001, [RRC01], presents a comparison among the extinction powerof different clean agents, for various classes of test fires. Results are givenregarding the concentration of agent necessary to warrant a fire quenching. Thetests under comparison include also inert gases like IG541. HFC-227ea resultsin one of the clean agents which demands less concentration, thus proving it isone of the most efficient agents.

Sheinson et al. 2003, [SAA+03], and Sheinson et al. 2004, [SAAM04], introducea new technique for fire suppression, namely the simultaneous discharge of HFC-227ea (called HFP in these work) and a very fine water mist (called WSCS),

4HCFC Blend A, also known as NAF S-III, is composed of 82% of HCFH-22 (CHClF2),9.5% of HCFC-124 (CHClFCF3), 4.75% of HCFC-123 (CHCl2CF3), and 3.75% ofisopropenyl-1-methyl-cyclohexane (C10H16)


which helps in cooling the ambient in which the fire develops. Figure 24 showsthe improvement on extinction efficiency obtained with this method: the twocurves correspond to the extinguishing of a 1900 kW fire of methyl alcoholwith and without WSCS (measurements taken at 4.0 m height). Figure 24demonstrates that WSCS reduced both the peak overhead temperature and thetemperature several minutes after HFC-227ea discharge. This reduction protectsthe compartment by helping decreasing the compartment’s temperature belowthe flashpoint of the fuel.

Figure 24: Temperature comparison for HFC-227ea extinguished fires with andwithout WSCS.

Yang-Keyser 2006, [YK06], makes an interesting study about the times neces-sary for extinguishing, with different agents and test fires, for aircraft enginenacelles.

3.3 Considerations About Friction

One key aspect that deserves to be taken into account in the hydrodynamicalstudy of the clean agents’ discharge process, is the calculation of friction withinthe pipe. Most authors use correlations that are suitably tested for steady flows,but very little (if anything) is found on the suitable friction factor correlationto be used in case of highly transient flows.

The most common empiric correlation regarding friction factor f , namely theColebrook-White formula, relates f with Reynolds number Re, which in turn


relates to velocity. This makes the friction factor f time dependent. The samecould be said about the various correlations that have been proposed in theliterature for the friction factor (see, for instance, [Bra09], [Cla09], [Den09],[ED98], [GS08], [KM86], [KY78], [RK06], [SG06], [TCK05],...). But all theseempirical formulae have been devised for situations of steady state flow. Toour knowledge, no such correlation has ever been introduced for transient flows,what constitutes no surprise, given the unmanageable amount of variables whichtake part on even the simplest instances of transient flows (such as the oneconsidered here: clean agent’s discharge through pipes). What seems fairly wellestablished is that friction stress in transient phenomena is higher (some wouldsay much higher) than their steady flow counterparts.

A detailed study of the friction factor correlations available in the literature canbe found in the excellent handbook of Rohsenow et al. 1998, [RHC98]. Thetables 5.8 and 5.9 within chapter 5, written by Ebadian-Dong, [ED98], offer acomprehensive collection of such correlations. The case treated here, dischargeof clean agents through pipes, correspond to table 5.8, smooth circular ducts.

Bratland 2009, [Bra09], presents a very extensive discussion on the friction factorfor transient flows, although he proposes no correlation (see chapter 2). He alsoincludes a table with a number of the correlations found on [ED98].

Denton 2009, [Den09], uses the Fanning friction factor in the equations, althoughthe processes considered there are highly transient. Nevertheless the authorrecommends the usage of Chen correlation (rough pipes, page 84), rather thanColebrook-White. For smooth pipes (the case considered here) he recommendsthe Techo et al. correlation.

Taylor et al. 2005, [TCK05], presents a extensive discussion of the effect ofsurface roughness on friction within pipes. It proposes new methods for deter-mining such roughness.

Kurokawa-Morikawa 1986, [KM86], introduces the concept of transient frictioncoefficient, and it even offers some clues as to how could it be estimated, but nocorrelation is presented. The work introduces explicitly the fluid’s accelerationin the expression leading to estimate the difference between transient and quasi-steady friction coefficients.

3.4 Thermophysical Properties of HFC-227ea

Aside from the data supplied by Dupont (manufacturer of HFC-227ea, not manyreferences could be found offering new findings regarding the thermophysicalproperties of agent HFC-227ea.

Yang et al. 1995, [YHB95] uses a model based on the extended correspondencestates principle, to find out the solubilities of N2 in several clean agents, whichuse this gas as a pressuring agent. It also calculates the pressure-temperaturerelationship for agent/N2 mixtures. The model claims to be accurate within10% of actual experimental measurements.


Baginskii-Stankus 2002, [BS03], studies some thermodynamic and transportproperties of HFC-227ea, namely heat capacity and thermal conductivity. Itoffers correlations to calculate these values within 2% of actual experimentalmeasurements. It centres on liquid HFC-227ea.

Benedetto et al. 2000, [BGS+00], proposes a equation of state for gaseous HFC-227ea, obtained directly from speed of sound measurements. This determinationof speed of sound is also valuable in those cases in which the vapour phase ofHFC-227ea could choke during discharge process.

Chen et al. 2003, [CHC03] offers generalized estimation equations for the ther-mophysical properties of saturated HFC-227ea. The estimated quantities areenthalpy, entropy, density of saturated liquid and gas, enthalpy of vaporization,and saturated vapour pressure.

4 CONCLUSIONS 34

4 Conclusions

As a conclusion, there are a number of works, articles, papers, technical reports,etc concerning experimental and empirical aspects of agent discharge, but little(if any) study from a pure theoretical point of view. Some purely theoreticaland mathematical work seems to be missing, so that a solid framework existsinto which insert the experimental findings and the empirical conclusions.

The literature review performed in this work has shown that no explicitly tran-sient dynamical-analytical model could be found for the clean agent dischargephenomenon. Instead, there are CFD modelling and steady flow equations todescribe this process.

Thus it is pertinent that such a model be developed and proposed to the FireProtection Engineering Community, as a valuable tool for improving the calcu-lations and design of actual clean agent fire suppression installations.

Such a work will be presented in a paper to complement this one here. It willcontain most of the issues arisen in this Bibliography Research and State ofthe Art investigation for this discharge process. Perhaps this dynamical modelwill prevent that some Engineer ever designs a fire suppression system like theone shown in figure 25, in which the agent’s discharge is carried out with nosignificant length of pipe, thus making it possible that liquid at very high speedcould impinge directly on people standing by.

Figure 25: Extinction system with very small length of discharge’s pipe

REFERENCES 35

References

[BBFY08] T.K. Blanchat, A.L. Brown, V. Figueroa, and S.S. Yoon. Bench-mark enclosure fire suppression experiments and modeling. InSuppression and Detection and Applications-A Technical WorkingConference (SUPDET 2008), March 2008. 6, 16

[Ben06] J.M. Bennett. Verification of NGP Fire Suppression Principles.In R.G. Gann, S.R. Burgess, K.C. Whisner, and P.A. Reneke, ed-itors, Halon Options Technical Working Conferences (HOTWC)-2006, volume NIST Special Publication. National Fire ProtectionAssociation, May 2006. 8, 30

[BGF+95] E.B. Bird, H.D. Giesecke, T.J. Friderichs, A. Maranghides, andR.S. Sheinson. Results of Benchmark Comparisons of Calculatedand Measured Flow Parameters for Discharges of Halon Replace-ment Chemicals. In Halon Options Technical Working Conferences(HOTWC)-1995. National Institute of Standards and Technology,May 1995. 6, 25

[BGH+94] E.B. Bird, H.D. Giesecke, J.A. Hillaert, T.J. Friderichs, and R.S.Sheinson. Development of a Computer Model to Predict theTransient Discharge Characteristics of Halon Alternatives. InHalon Options Technical Working Conferences (HOTWC)-1994.National Institute of Standards and Technology, May 1994. 25

[BGS+00] G. Benedetto, R.M. Gavioso, R. Spagnolo, M. Grigiante, andG. Scalabrin. Vapor Phase Helmholtz Equation for HFC-227eafrom Speed of Sound Measurements. In Fourteenth Symposium onThermophysical Properties, June 2000. 33

[Bra09] O. Bratland. Pipe Flow 1: Single-phase Flow Assurance. e-bookwww.drbratland.com, 2009. 32

[BS03] A.V. Baginskii and S.V. Stankus. Thermodynamic and TransportProperties of Liquid HFC-227ea. International Journal of Ther-mophysics, 24(4):953–961, July 2003. 33

[C+03] A.E. Cote et al., editors. Fire Protection Handbook, vols. I & II.National Fire Protection Association, 19th edition, 2003. 29

[CEA05] CEA4045. Fire extinguishing systems using liquified ’halocarbon’gases. Planning and installation. Comite Europeen des Assurances,April 2005. 11

[CGY94] T.G. Cleary, W.L. Grosshandler, and J.C. Yang. Flow of Alter-native Agents in Piping. In Halon Options Technical WorkingConferences (HOTWC)-1994. National Institute of Standards andTechnology, May 1994. 26

[CHC03] Z.S. Chen, P. Hu, and W.L. Cheng. Generalized Estimation Equa-tions for Thermophysics Properties of Saturated Fluids and theSaturated Properties Table of HFC-227ea. In Fifteenth Sympo-sium on Thermophysical Properties, June 2003. 33

REFERENCES 36

[CHM92] R.N. Coward, J.A. Hillaert, and D.M. McCrory. Analytical meth-ods for modeling discharge characteristics of Halon 1301 fireprotection systems. In Halon Alternatives Technical WorkingCoference-1992. National Institute of Standards and Technology,May 1992. 25

[Cla09] D. Clamond. Efficient resolution of the Colebrook equation.Ind.Eng.Chem.Res., 48(7):3665–3671, February 2009. 32

[Coo93] L.Y. Cooper. Dischage of Fire Suppression Agents from a Pres-surized Vessel: A Mathematical Model and its Application to Ex-perimental Design. NISTIR, (5181), May 1993. 9, 23

[Coo94] L.Y. Cooper. Dispersion of Fire Suppression Agents DischargedFrom High Pressure Vessels: Establishing Initial/Boundary Con-ditions for the Flow Outside the Vessel. In Liquid Atomization andSpray Systems, 6th International Conference Proceedings., volumeICLASS, July 1994. 6, 8

[CYK+95] T.G. Cleary, J.C. Yang, M.D. King, C.I. Boyer, and W.L.Grosshandler. Pipe flow characteristics of alternative agents forengine nacelle fire protection. In Proccedings on Halon OptionsTechnical Working Conference-1995. National Institute of Stan-dards and Technology, May 1995. 26

[Den09] G.S. Denton. CFD Simulation of Highly Transient Flows. PhDthesis, University College London, July 2009. 29, 32

[DFF+94] P.J. DiNenno, E.W. Forssell, M.J. Ferreira, C.P. Hanauska, andB.A. Johnson. Modeling of the Flow Properties and Dischargeof Halon Replacement Agents. In Proccedings on Halon OptionsTechnical Working Conference-1994. National Institute of Stan-dards and Technology, 1994. 24

[DiN02] P.J. DiNenno. Halon Replacement Clean Agent Total FloodingSystems, volume SFPE Handbook of Fire Protection Engineering,chapter 4-7. National Fire Protection Association, 3rd edition,2002. 11, 15

[DiN03] P.J. DiNenno. Direct Halon Replacement Agents and Systems,volume NFPA Fire Protection Handbook of NFPA Fire ProtectionHandbook, vols. I & II, chapter 11-2. National Fire ProtectionAssociation, 19th edition, 2003. 11, 15

[Dur08] F. Durst. Fluid Mechanics. An Introduction to the Theory of FluidFlows. Springer-Verlag Berlin Heidelberg, 2008. 9

[ED98] M.A. Ebadian and Z.F. Dong. Forced Convection, Internal Flowin Ducts, chapter 5. Handbook of Heat Transfer. McGraw-Hill,3rd edition, 1998. 32

[EGK+84] D.G. Elliot, P.W. Garrison, G.A. Klein, K.M. Moran, and M.P.Zydowicz. Flow of nitrogen-pressurized Halon 1301 in Fire Extin-guishing Systems. Jet Propulsion Laboratory, JPL Publication(84-62), November 1984. 24, 26

REFERENCES 37

[Fri03] R. Friedman. Theory of Fire Extinguishment, chapter 2-5, pages2–83/2–96. NFPA Fire Protection Handbook, vols. I & II. NationalFire Protection Association, 19th edition, 2003. 8, 29

[Gan95] R.G. Gann. Next Generation Fire Suppression System Technology:A National Research Plan. In Halon Options Technical WorkingConferences (HOTWC)-1995. National Institute of Standards andTechnology, May 1995. 5

[Gan98] R.G. Gann. Next-Generation Fire Suppression TechnologyProgram. In Halon Options Technical Working Conferences(HOTWC)-1998. National Institute of Standards and Technology,May 1998. 5

[Gan03] R.G. Gann. Next Generation Fire Suppression Technology Pro-gram: FY2003 Progress. In Halon Options Technical WorkingConferences (HOTWC)-2003. National Institute of Standards andTechnology, May 2003. 5

[Gan05] R.G. Gann. Next Generation Fire Suppression Technology Pro-gram: FY2005 Progress. In Halon Options Technical WorkingConferences (HOTWC)-2005. National Institute of Standards andTechnology, May 2005. 5

[GGP94] W.L. Grosshandler, R.G. Gann, and W.M. Pitts, editors. Eval-uation of Alternative In-Flight Fire Suppressants for Full-ScaleTesting in Simulated Aircraft Engine Nacelles and Dry Bays, vol-ume SP. National Institute of Standards and Technology, April1994. 6, 18, 30

[GHCP00] W.L. Grosshandler, A. Hamins, S.R. Charagundla, and C. Presser.Suppression Effectiveness Screening for Impulsively DischargedAgents. Halon Options Technical Working Conference, May 2000.6

[GS08] C.T. Goudar and J.R. Sonnad. Comparison of the iterative ap-proximations of the Colebrook-White equation. Hidrocarbon Pro-cessing, 87(8):79–84, 2008. 32

[HGG+94] A. Hamins, G. Gmurczyk, W.L. Grosshandler, R.G. Rehwoldt,I. Vazquez, T.G. Cleary, C. Presser, and K. Seshadri. FlameSuppression Effectiveness, volume NIST-SP 861 of Evaluation ofAlternative In-Flight Fire Suppressants for Full-Scale Testing inSimulated Aircraft Engine Nacelles and Dry Bays, chapter 4. Na-tional Institute of Standards and Technology, April 1994. 6, 8,30

[HTSC94] A. Hamins, D. Trees, K. Seshadri, and H.K. Chelliah. Extinction ofNonpremixed Flames With Halogenated Fire Suppressants. Com-bustion and Flame, 99(2):221–230, August 1994. 6, 8

REFERENCES 38

[HTSD03] J.C. Hewson, S.R. Tieszen, W.D. Sundberg, and P.E. Desjardin.CFD modeling of fire suppression and its role in optimizing sup-pressant distribution. In Proccedings on Halon Options Techni-cal Working Conference-2003. National Institute of Standards andTechnology, 2003. 16

[ISO06] ISO-14520-1. Gaseous fire-extinguishing systems-Physical proper-ties and system design-Part 1: General requirements. InternationalStandard Organization, 2006. 11

[KE81] M.E. Kim-E. The possible consequences of rapidly depressurizinga fluid. Master’s thesis, Massachusetts Institute of Technology,January 1981. 29

[KLGS94] S.K. Kim, T.J. Lestina, H.D. Giesecke, and R.S. Sheinson. De-velopment of Computer Models for the Discharge of Halon Al-ternatives. In Halon Options Technical Working Conferences(HOTWC)-1994. National Institute of Standards and Technology,May 1994. 25

[KM86] J. Kurokawa and M. Morikawa. Accelerated and Decelerated Flowsin a Circular Pipe. Bulletin of JSME, 29(249):758–765, March1986. 32

[KOYS01] H. Kashiwagi, S. Oshikawa, J. Yui, and Y. Saso. Effect of FireSize on Suppression Characteristics of Halon Replacement Total-Flooding Systems. In Halon Options Technical Working Confer-ences (HOTWC)-2001. National Institute of Standards and Tech-nology, April 2001. 8

[KSMKK96] A.K. Kim, J.Z. Su, J.R. Mawhinney, and M. Kanabus-Kaminska.Full-Scale Fire Testing of HFC-227ea and HCFC Blend A. InHalon Options Technical Working Conferences (HOTWC)-1996.National Institute of Standards and Technology, May 1996. 8, 30

[KY78] C.A. Kot and C.K. Youngdahl. Transient Cavitation Effects inFluid Piping Systems. Nuclear Engineering and Design, 45:93–100, 1978. 32

[Lin06] G.T. Linteris. NGP Research on Fire Suppression Chemistry. InHalon Options Technical Working Conferences (HOTWC)-2006.National Institute of Standards and Technology, May 2006. 6, 8,30

[MDK98] B. Moghtaderi, B.Z. Dlugogorski, and E.M. Kennedy. DetailedChemical Kinetic Modelling Study on High Temperature Ignitionof Methane/Air Mixtures Doped With C3F7H. In Halon OptionsTechnical Working Conferences (HOTWC)-1998. National Insti-tute of Standards and Technology, May 1998. 6, 8, 30

[MOR06] H. Mahgerefteh, A.O. Oke, and Y. Rykov. Efficient numericalsolution for highly transient flows. Chemical Engineering Science,61(15):5049–5056, March 2006. 29

REFERENCES 39

[NBHT98] T. Noto, V. Babushok, A. Hamins, and W. Tsang. InhibitionEffectiveness of Halogenated Compounds. Combustion and Flame,112(1/2):147–160, January 1998. 6

[NBZR10] A. Nouri-Borujerdi and M. Ziaei-Rad. Numerical modeling of tran-sient turbulent gas flow in a pipe following a rupture. TransactionB: Mechanical Engineering, 17(2):108–120, April 2010. 29

[NFP08] NFPA-2001. NFPA-Standard on Clean Agent Fire ExtinguishingSystems. National Fire Protection Association, August 2008. 11

[OMER03] A.O. Oke, H. Mahgerefteh, I. Economou, and Y. Rykov. A tran-sient outflow model for pipeline puncture. Chemical EngineeringScience, 58(20):4591–4604, July 2003. 29

[PYB+06] W.M. Pitts, J.C. Yang, R.A. Bryant, L.G. Blevins, and M.L.Huber. Characterization and Identification of Super-EffectiveThermal Fire Extinguishing Agents. Final Report. NGP Project4C/1/890. Technical Report NIST Technical Note 1440, NationalInstitute of Standards and Technology, July 2006. 6, 8

[PYBG93] W.M. Pitts, J.C. Yang, B. Breuel, and G. Gmurczyk. Dynamics ofthe release of alternate Halon replacement agents from pressurizedbottles. In Halon Alternatives Technical Working Coference-1993.National Institute of Standards and Technology, May 1993. 23

[PYG+94] W.M. Pitts, J.C. Yang, G. Gmurczyk, L.Y. Cooper, W.L.Grosshandler, W.G. Cleveland, and C. Presser. Fluid Dynamicsof Agent Discharge, volume SP of Evaluation of Alternative In-Flight Fire Suppressants for Full-Scale Testing in Simulated Air-craft Engine Nacelles and Dry Bays, chapter 3. National Instituteof Standards and Technology, April 1994. 4, 9, 18, 19, 23

[RFS05] M.L. Robin, E.W. Forssell, and V. Sharma. Pressure Dynamicsof Clean Agent Discharges. In Halon Options Technical WorkingConferences (HOTWC)-2005. National Institute of Standards andTechnology, May 2005. 27

[RHC98] W.M. Rohsenow, J.P. Hartnett, and Y.I. Cho, editors. Handbookof Heat Transfer. McGraw-Hill, 3rd edition, 1998. 32

[RK06] A.R. Rao and B. Kumar. Friction factor for turbulent pipe flow.2006. 32

[RRC01] M.L. Robin, T.F. Rowland, and M.D. Cisneros. Fire SuppressionTesting: Extinguishment of Class A Fires With Clean Agents. InHalon Options Technical Working Conferences (HOTWC)-2001.National Institute of Standards and Technology, April 2001. 6, 8,30

[SAA+03] R.S. Sheinson, S. Ayers, R. Anleitner, D. Morse, D. Szwarc,L. Levenberry, and A. Maranghides. Combining a Water SprayCooling System With Heptafluoropropane for Total Flooding FireSuppression. In Halon Options Technical Working Conferences

REFERENCES 40

(HOTWC)-2003. National Institute of Standards and Technology,May 2003. 8, 30

[SAAM04] R.S. Sheinson, S. Ayers, R. Anleitner, and A. Maranghides. Hep-tafluoropropane With Water Spray Cooling System as a To-tal Flooding Halon 1301 Replacement: System ImplementationParameters. In Halon Options Technical Working Conferences(HOTWC)-2004. National Institute of Standards and Technology,May 2004. 8, 30

[Sas01] Y. Saso. Roles of Inhibitors in Global Gas-Phase Combus-tion Kinetics. In Halon Options Technical Working Conferences(HOTWC)-2001. National Institute of Standards and Technology,April 2001. 6, 8, 30

[Sen01] J.A. Senecal. New Technology Delivery System for FM-200Clean Agent. In Halon Options Technical Working Conferences(HOTWC)-2001. National Institute of Standards and Technology,April 2001. 6

[SG06] J.R. Sonnad and C.T. Goudar. Turbulent flow friction factor calcu-lation using a mathematically exact alternative to the Colebrook-White equation. J.Hydraulic Engineering, 132(8):863–867, August2006. 32

[Sha53] A.H. Shapiro. The Dynamics and Thermodynamics of Compress-ible Fluid Flow, volume I. The Ronald Press Company, 1953. 18,19

[Sha54] A.H. Shapiro. The Dynamics and Thermodynamics of Compress-ible Fluid Flow, volume II. The Ronald Press Company, 1954.18

[Ska02] R.R. Skaggs. Assessment of the Fire Suppression Mechanics forHFC-227ea Combined With NaHCO3. In Halon Options Tech-nical Working Conferences (HOTWC)-2002. National Institute ofStandards and Technology, May 2002. 8

[SP95] J.A. Senecal and R.C. Prescott. FM-200 Suppression Systems:A Conservative Discharge Test Method and In-Room PressureVariance Upon Discharge. In Halon Options Technical WorkingConferences (HOTWC)-1995. National Institute of Standards andTechnology, May 1995. 26

[TCK05] J.B. Taylor, A.L. Carrano, and S.G. Kandlikar. Characterizationof the effect of surface roughness and texture on fluid flow-Past,present and future. Proceedings of ICMM2005, 3rd InternationalConference on Microchannels and Minichannels, June 2005. 32

[TPC+00] K. Tuzla, T. Palmer, J.C. Chen, R.K. Sundaram, and W.S. Ye-ung. Development of computer program for fire suppressant fluidflow-Final technical report. Technical report, Lehigh University,September 2000. 26

REFERENCES 41

[UE09a] UNE-EN-15004-1. Sistemas fijos de lucha contra incendios. sis-temas de extincion mediante agentes gaseosos. parte 1: Diseno,instalacion y mantenimiento. AENOR, 2009. 11

[UE09b] UNE-EN-15004-5. Sistemas fijos de lucha contra incendios.Sistemas de extincion mediante agentes gaseosos. Parte 5:Propiedades fısicas y diseno de sistemas de extincion medianteagentes gaseosos con HFC 227 ea. AENOR, 2009. 11

[Val98] E. Valero Sanchez. Analisis del proceso de descarga de fluidossometidos a altas presiones y temperaturas. PhD thesis, E.T.S.Ingenieros Aeronauticos, Madrid, December 1998. 29

[VdS09] VdS2381. Fire Extinguishing Systems using Halocarbon Gases.VdS Schadenverhutung GmbH, June 2009. 11

[WC96] T. Wysocki and B. Christensen. Inert gas fire suppression sys-tems using IG541 (INERGEN). Solving the hydraulic calculationproblem. In Proccedings on Halon Options Technical WorkingConference-1996. National Institute of Standards and Technology,1996. 26

[Wys96] T. Wysocki. Single point flow calculations for liquefied compressedgas fire extinguishing agents. In Proccedings on Halon OptionsTechnical Working Conference-1996. National Institute of Stan-dards and Technology, May 1996. 23

[YHB95] J.C. Yang, M.L. Huber, and C.I. Boyer. A Model for Calculat-ing Alternative Agent/Nitrogen Thermodynamic Properties. InHalon Options Technical Working Conferences (HOTWC)-1995.National Institute of Standards and Technology, May 1995. 32

[YK06] J.C. Yang and D.R. Keyser. Fluid Dispensing and Dispersion. InHalon Options Technical Working Conferences (HOTWC)-2006.National Institute of Standards and Technology, May 2006. 8, 31

[ZW96] R. Zalosh and W. Wang. Mathematical modeling of FM-200 dis-charge ana leakage from an enclosure. In Proccedings on HalonOptions Technical Working Conference-1996. National Instituteof Standards and Technology, May 1996. 25

42

Appendices

A Properties of HFC-227ea

DuPont™ FM-200® (HFC-227ea)FIRE EXTINGUISHING AGENT

Properties, Uses, Storage, and Handling

DuPont™ FM-200® Fire Extinguishing AgentProperties, Uses, Storage, and Handling

Table of Contents Page

Introduction ............................................................................................................................................................................ 1

Chemical Properties of FM-200® .......................................................................................................................................... 1

Uses ......................................................................................................................................................................................... 1

Physical Properties ............................................................................................................................................................... 1

Materials Compatibility ........................................................................................................................................................ 5

Stability with Metals ................................................................................................................................................... 5

Compatibility with Elastomers .................................................................................................................................. 5

Compatibility with Plastics ........................................................................................................................................ 6

Safety ...................................................................................................................................................................................... 6

Inhalation Toxicity ....................................................................................................................................................... 6

Cardiac Sensitization ................................................................................................................................................... 6

Skin and Eye Contact ................................................................................................................................................... 6

Spills or Leaks ............................................................................................................................................................... 7

Storage and Handling ......................................................................................................................................................... 7

Shipping Information for the United States ........................................................................................................... 7

Containers ................................................................................................................................................................... 7

Bulk Storage Systems ................................................................................................................................................. 7

Transfer of FM-200® from Containers ........................................................................................................................ 8

Leak Detection .............................................................................................................................................................. 8

Handling Precautions for FM-200® Shipping Containers ....................................................................................... 8

Nitrogen Superpressurization of FM-200® ....................................................................................................................... 11

Henry’s Law Constants ....................................................................................................................................................... 11

Recovery, Reclamation, and Disposal .... .......................................................................................................................... 15

Recovery .................................................................................................................................................................... 15

Reclamation ................................................................................................................................................................ 15

Disposal ....................................................................................................................................................................... 15

IntroductionChlorofluorocarbons (CFCs) and bromine-containing compounds such as the Halons (Halon 1301, Halon 1211) have many unique properties. They are low in toxicity, nonflammable, noncorrosive, and compatible with other materials. In addition, they offer thermodynamic and physi-cal properties that make them ideal for a variety of uses. CFCs have been used as aerosol propellants, refrigerants, blowing agents for plastic foams, cleaning agents for metal and electronic components, and in many other applications. The Halons have been used as fire extinguishing agents and explosion suppressants for the protection of high-value equipment and assets, and have been employed in hand-held portable extinguishers, total flooding systems, and local application systems.

However, the atmospheric stability of these compounds, coupled with their bromine and/or chlorine content, has linked them to depletion of the earth’s protective ozone layer. As a result, DuPont has stopped production of CFCs and Halons and introduced environmentally acceptable alternatives, such as FM-200®. FM-200® contains no bro-mine or chlorine; as a result, it does not contribute to the destruction of stratospheric ozone, i.e., FM-200® has an ozone depletion potential (ODP) of zero.

Chemical Properties of FM-200®

Chemical Name 1,1,1,2,3,3,3-heptafluoropropane

Molecular Formula CF3CHFCF3

Molecular Weight 170.03

CAS Registry Number 431-89-0

ASHRAE Designation HFC-227ea

UsesFM-200® is listed as an acceptable replacement for Halon 1301 and Halon 1211 in the United States Environmental Protection Agency’s Significant New Alternatives Policy (SNAP) program. FM-200® is a suitable fire extinguishing agent for total flooding, portable, and local application sys-tems. FM-200® is noncorrosive, electrically nonconductive, free of residue, and characterized by low toxicity. It is ideally suited for protection of high-value assets such as those found in computer rooms, data control centers, telecommu-nication facilities, and museums.

The fire extinguishing concentrations of FM-200® allow it to be used as a total flooding agent in normally occupied spaces for the protection of Class A (solid), Class B (liquid and gas), and Class C (electrically energized) hazards.

FM-200® is also suitable for use as an inertion agent in explosion suppression applications.

Physical PropertiesPhysical properties of FM-200® are given in Tables 1 to 3 and Figures 1 and 2.

For complete thermodynamic properties, see DuPont Bul-letin T-FM-200.

1

2

Table 1 Physical Properties of HFC-227ea

Property

Chemical name 1,1,1,2,3,3,3-Heptafluoropropane

Chemical formula CF3CHFCF3

Molecular Wt. 170.03

Boiling Point, 1 atm, °C (°F) –16.34 (2.59)

Freezing Point, °C (°F) –131 (–204)

Critical Temperature, °C (°F) 101.75 (215.1)

Critical Pressure, kPa (psia) 2925.0 (424.24)

Critical Density, kg/m3 (lb/ft3) 594.25 (37.098)

Liquid Density @ 25°C (77°F), kg/m3 (lb/ft3) 1387.7 (86.63)

Density, Saturated Vapor at Boiling Point, kg/m3 (lb/ft3) 8.4860 (0.52979)

Vapor Density @ 25°C (77°F) and 1 atm 7.1461 (0.4461)

Specific Heat, Liquid (Cp) @ 25°C (77°F), kJ/kg–°C (Btu/lb°F) 1.1816 (0.28242)

Specific Heat, Vapor (Cp) @ 25°C (77°F), kJ/kg–°C (Btu/lb°F) 0.81327 (0.81327) and 1 atm

Vapor Pressure, Saturated @ 25°C (77°F), kPa (psia) 454.73 (65.9)

Heat of Vaporization @ B.P., kJ/kg (Btu/lb) 131.77 (56.7)

Thermal Conductivity, Liquid @ 25°C (77°F), W/m–°C (Btu/hr-ft°F) 0.060491 (0.034975)

Thermal Conductivity, Vapor @ 1 atm, W/m–°C (Btu/hr-ft°F) 0.013336 (0.0077103)

Viscosity, Liquid @ 25°C (77°F), cP ( lb/ft-hr) 0.23935 (0.57901)

Viscosity, Vapor @ 1 atm, cP ( lb/ft-hr) 0.011590 (0.028038)

Relative dielectric strength @1 atm, 25°C (N2=1) 2.00

Solubility of Water in HFC-227ea @ 20°C (68°F), ppm 600

Ozone Depletion Potential 0.0 (CFC-11 = 1)

Global Warming Potential, GWP 3220 (100 yr ITH. For CO2, GWP = 1)

Atmospheric Lifetime, years 34.2

TSCA Inventory Status Reported/Included

European List of New Chemical Substances EINECS, Listed (207-079-2)

SNAP Status Listed

Inhalation Exposure Limit (AELa) 1000 ppm 8 hr and 12 hr TWA a AEL (acceptable exposure limit) is an airborne exposure limit established by DuPont that specifies time-weighted average concentrations to which nearly all workers may be repeatedly exposed without adverse effects.

3

Table 2 Vapor Pressure and Density of FM-200® (SI units)

Vapor Liquid Saturated Vapor Vapor Temperature Pressure Density Density Density @ 1 atm °C (kPa) (kg/m3) (kg/m3) (kg/m3)

–15 107.33 1539.7 8.961 8.4325 –10 132.23 1522.1 10.921 8.2412 –5 161.41 1504.2 13.205 8.0603 0 195.36 1486.0 15.853 7.8889 5 234.58 1467.3 18.905 7.7260 10 279.57 1448.2 22.411 7.5709 15 330.89 1428.6 26.421 7.4229 20 389.08 1408.4 30.996 7.2815 25 454.73 1387.7 36.202 7.1461 30 528.42 1366.2 42.118 7.0163 35 610.79 1344.0 48.833 6.8918 40 702.45 1320.9 56.454 6.7720 45 804.09 1296.7 65.109 6.6568 50 916.39 1271.4 74.956 6.5459 55 1040.10 1244.8 86.189 6.4389 60 1175.90 1216.5 99.062 6.3356 65 1324.70 1186.2 113.900 6.2359 70 1487.40 1153.6 131.170 6.1395 75 1664.90 1117.9 151.500 6.0462 80 1858.30 1078.2 175.870 5.9559 85 2068.80 1032.8 205.840 5.8684 90 2298.10 978.6 244.310 5.7836 95 2547.90 907.8 298.000 5.7013 100 2821.60 786.8 397.240 5.6215

Table 3 Vapor Pressure and Density of FM-200® (English units)

Vapor Liquid Saturated Vapor Vapor Temperature Pressure Density Density Density @ 1 atm °F (psia) (lbm/ft3) (lbm/ft3) (lbm/ft3)

10 17.50 95.51 0.63 0.5197 20 21.93 94.28 0.77 0.5069 30 27.18 93.02 0.95 0.4948 40 33.35 91.73 1.16 0.4834 50 40.55 90.41 1.40 0.4726 60 48.88 89.05 1.68 0.4624 70 58.45 87.64 2.00 0.4527 80 69.38 86.19 2.38 0.4434 90 81.79 84.68 2.81 0.4345 100 95.80 83.11 3.31 0.4261 110 111.54 81.46 3.88 0.4179 120 129.15 79.73 4.54 0.4102 130 148.77 77.90 5.30 0.4027 140 170.55 75.94 6.18 0.3955 150 194.65 73.84 7.22 0.3886 160 221.26 71.54 8.45 0.3820 170 250.55 69.00 9.93 0.3756 180 282.77 66.10 11.76 0.3694 190 318.18 62.68 14.10 0.3634 200 357.11 58.32 17.33 0.3576

4

Figure 1. Vapor Pressure of DuPont™ FM-200® (English Units)

Figure 2. Vapor Pressure of DuPont™ FM-200® (SI Units)

5

Materials CompatibilityBecause FM-200® can be used in a variety of applications, it is important to review the materials of construction for compatibility when designing new equipment, retrofitting existing equipment, or preparing storage and handling facili-ties. The following are general test results. To determine the compatibility of the specific materials being considered for use in a particular system, additional tests should be conducted with these materials at the conditions of that system.

Stability with Metals

Most of the commonly encountered metals such as steel, cast iron, brass, copper, tin, lead, and aluminum can be employed with FM-200® under normal conditions of use. Testing to ASTM G31 at 130°F for 18 days indicates that the following metals are suitable for use with FM-200®:

Aluminum 1100 Nickel 200 Aluminum 2024 Copper CDA 110 Inconel 600 Cast Iron, grey Stainless Steel 316 Lead Stainless Steel 304 Carbon Steel 1020 Yellow Brass Silver 999+ fine

High temperature stability tests were conducted with FM-200® and commonly used metals at 175°C for two weeks, and FM-200® was found to be very stable. At tem-peratures above 175°C, depending on specific conditions,

some metals may act as catalysts for the breakdown of FM-200®. These conditions include: presence of moisture or other contaminants, type of metal, metal surface area, contact with liquid or vapor agent, as well as time and temperature of contact.

Halocarbons may react violently with highly reactive metals such as the alkali and alkaline earth metals, sodium, potas-sium, and barium, in their free metallic form. Materials become more reactive when finely ground or powdered, and in this state magnesium and aluminum may react, especially at higher temperatures. Highly reactive materi-als should not be brought into contact with FM-200® until a careful study is made and appropriate safety precautions are taken.

Compatibility with Elastomers

Compatibility tests were performed with several commonly used elastomers. Elastomer coupons were 50% immersed in liquid FM-200® for two weeks at room temperature. Additional tests were conducted per ASTM D471 at 100°C (212°F). All of the elastomers tested exhibited minimal swell with the exception of urethane and Viton® A. Results are summarized in Table 4.

Table 4 Elastomer Compatibility

Linear swell, Weight Gain, Hardness Change, Elastomer % % units

Exposure at Room temperature, 23°C (72°F) for 14 daysButyl 0 0.37 0Nordel® EPDM 0.20 1.44 1.6Neoprene W 0.05 0.66 0NBR 0 1.86 4.0Hypalon® CSM 0.19 1.41 2.4Viton® A 9.49 26.83 –44.0Epichlorohydrin homopolymer 0.15 0.08 5.5FA polysulfide 0.05 0.06 6.9Hytrel TPE 1.33 5.71 4.6 Exposure per ASTM D471 at 100°C (212°F)Buna N -3.1Butyl 3.6EPDM 1.0Hypalon® -2.0NR Rubber 1.7Neoprene G 0.8Neoprene W -3.6SBR -1.2Silicone 2.8Urethane >10Viton® A 8.4

6

Compatibility with Plastics

Compatibility tests were also performed with several com-monly used plastics. Results are summarized in Table 5.

SafetyUsers of FM-200® should read and understand the DuPont Material Safety Data Sheet (MSDS). Copies of the FM-200® MSDS can be obtained from DuPont Customer Service or International Offices (see last page of this docu-ment for locations, telephone numbers, and Web site).

Inhalation Toxicity

FM-200® poses no acute or chronic hazard when it is handled in accordance with DuPont recommendations and when the exposure is maintained below the recommended exposure limits. DuPont has established the Allowable Exposure Limit (AEL) for FM-200® at 1000 ppm, 8-hr and 12-hr TWA.

However, inhaling high concentrations of FM-200® vapor may cause temporary nervous system depression with anesthetic effects such as dizziness, headache, confu-sion, loss of coordination, and even loss of consciousness. Higher exposures to the vapors may cause temporary al-teration of the heart’s electrical activity with irregular pulse, palpitations, or inadequate circulation. Intentional misuse or deliberate inhalation may cause death without warning.

If a person is experiencing any of the initial symptoms, they should be moved to fresh air and kept calm. If not breath-ing, give artificial respiration. If breathing is difficult, give oxygen. Seek medical attention.

Cardiac Sensitization

If vapors are inhaled at a concentration of 105,000 ppm and higher, the heart may become sensitized to adrenaline, leading to cardiac irregularities and, possibly, cardiac arrest. Similar effects are observed with many hydrocarbons and halocarbons at high concentrations. The likelihood of these cardiac problems increases if the person is under physical or emotional stress.

Because of possible disturbances of cardiac rhythm, cat-echolamine drugs, such as epinephrine, should be consid-ered only as a last resort in life-threatening emergencies.

The threshold cardiac sensitization, lowest observed adverse effect level (LOAEL) for FM-200® is 105,000 ppm (10.5%) and the no observed adverse effect level (NOAEL) is 90,000 ppm (9%) as determined in epinephrine-chal-lenged dogs.

Skin and Eye Contact

At room temperature, FM-200® vapors have little or no effect on the skin or eyes. However, in the liquid form, FM-200® can freeze the skin or eyes on contact, causing frostbite. If contact with liquid does occur, soak the ex-posed area in lukewarm water, not cold or hot. In all cases, seek medical attention as soon as possible.

Always wear protective clothing when there is a risk of exposure to liquid FM-200®. Where splashing of FM-200® may occur, always wear eye protection and a face shield.

Table 5 Plastic Compatibility

Weight Gain, Surface Plastic % Condition

High-density polyethylene (HDPE) 0.11 No ChangePolystyrene (PS) –0.03 No ChangePolypropylene (PP) 0.06 No ChangeAcrylonitrile-butadiene-styrene (ABS) –0.03 No ChangePolycarbonate (PC) –0.10 No ChangeNylon 6/6 –0.17 No ChangePolytetrafluoroethylene (PTFE) 5.23 No ChangePolyimide (PI) -0.11 No ChangePolyethylene terephthalate (PET) -0.04 No ChangePolybutylene terephthalate (PBT) -0.06 No ChangeAcetyl -0.04 No ChangePolyvinyl chloride (PVC) -0.06 No ChangePolyphenylene oxide (PPO) -0.05 No ChangePolyphenylene sulfide (PPS) -0.38 No Change

7

Spills or Leaks

If a large release of vapors occurs, such as from a large leak or spill, the vapors may concentrate near the floor or in subfloor areas and displace the oxygen available for breathing, causing suffocation.

Evacuate everyone until the area has been well ventilated. Use blowers or fans to circulate the air at floor level. Do not re-enter the affected area without self-contained breathing apparatus or unless the area has been monitored to indicate that that the concentration of FM-200® vapors in the area is below the AEL of 1000 ppm.

Always use self-contained breathing apparatus or a sup-plied air mask when entering tanks or other enclosures where vapors might exist. Use the buddy system and a lifeline. Refer to the FM-200® MSDS for more information.

FM-200® vapors have virtually no odor. Therefore, frequent leak checks or the installation of area monitors are neces-sary in enclosed areas where leaks can occur.

To ensure safety when working with halocarbons in confined areas:

1. Route relief and purge vent piping (if used) outdoors, away from air intakes.

2. Make certain the area is well ventilated, using auxiliary ventilation, if necessary, to move vapors.

3. Make sure the area is clear of vapors prior to beginning work.

4. Utilize monitoring equipment to detect leaks.

Storage and HandlingShipping Information for the United States

FM-200® is a liquefied compressed gas. According to the U.S. Department of Transportation (DOT), a liquefied com-pressed gas is a gas, which when packaged under pres-sure is partially liquid at temperatures above -50° C (49CFR 173.115). The appropriate DOT designation for FM-200® is as follows:

Proper shipping name: Heptafluoropropane

Hazards class: 2.2

UN No.: 3296

DOT/IMO Labels: Nonflammable Liquefied Compressed Gas

Containers

Four types of containers are being used globally for shipping FM-200®. Specifications for the containers are provided in Table 6.

The 68 kg (150-lb) size cylinder of FM-200® is a freestand-ing, upright returnable cylinder, equipped with a nonrefill-able liquid/vapor valve. With this two-way valve, FM-200® can be removed from the cylinder as either vapor or liquid through the single or double outlet without inverting the cylinder. The outlet is designed for a CGA-660 connection. The handwheel for discharging liquid is on the side of the valve. A dip tube, which extends to the bottom of the cyl-inder, is attached to the valve’s liquid port. The handwheel for discharging vapor is located on the top of the valve. A diagram of this cylinder is shown in Figure 3. A diagram of the Ceodeux brand liquid/vapor valve used on the 68-kg size cylinder is shown in Figure 4. The 68 kg size cylinders are usually shipped on a pallet, stacked on their side.

The 544 kg (1,200-lb) size FM-200® cylinder is a freestand-ing, upright returnable cylinder equipped with a forklift lift-ing attachment incorporated into the foot ring or at the tope of the cylinder. These cylinders are fitted with Ceodeux or Superior brand valves designed for a CGA 660 connection. A diagram of this cylinder is shown in Figure 5.

ISO containers are used for export shipments of FM-200®. The overall ISO container dimensions shown in Table 6 represent the frame in which the container is shipped. The tank itself is approximately 19 feet (5.8 me-ters) long, and has an outside diameter of approximately 86 inches (2.2 meters). Individual valves for liquid and vapor discharge are provided. Acme 1-3/4” x 1” MNPT valves (part number A2003) are employed for vapor removal and Acme 3-1/4” x 2” MNPT valves (part number A2063) for liquid removal. These valves fittings are on the ISO con-tainer; to unload the ISO container, couplings and adaptor fittings are required. Vapor removal requires the use of an Acme 1-3/4 x 1” MNPT adaptor (part number A1131F). Liq-uid removal requires the use of an Acme 3-1/4” x 2” MNPT adaptor (part number A1157F). A diagram of a typical ISO tank is shown in Figure 6.

FM200® is also shipped in 5000 gallon (8927 L) tank trail-ers. These tank trailers are equipped with 1-1/4” mail Ever- Tite (vapor) and 2” male Ever-Tite valves (liquid). Required power is either 240 volt or 440 volt.

Bulk Storage Systems

DuPont sells bulk storage systems to its FM-200® custom-ers. The systems are prefabricated, tested, and ready to install onsite. The units are designed to optimize economy, efficiency, and safety in the storage and dispensing of FM-200®. The delivered systems include all components, such as storage tanks, pumps, piping, valves, motors, and gauges, as an integrated unit. All systems are equipped with the DuPont Fluorochemical Emissions Elimination Delivery (FEED) System to prevent emissions during de-liveries and with dual pumps to provide an installed spare. The units are skid-mounted and require only placement on a concrete pad and connection to electrical and process systems.

8

A typical bulk storage system is shown in Figure 7.

Your DuPont Marketing Representative can arrange for guidance on site selection, purchase, installation, startup, and maintenance.

Transfer of FM-200® From Containers

The preferred method for transfer of liquid FM-200® from the cylinder is to use a suitable pump. There are several industrial pumps suitable for the transfer of FM-200®. Con-tact an industrial pump manufacturer for the recommended pump.

The receiving container should be evacuated to eliminate contamination by air and to facilitate transfer of FM-200®.

If a pump is not available the chilled transfer line method will facilitate transfer of FM-200® to the receiving container. This method chills the FM-200® as it passes through the transfer line, reducing the pressure in the receiver to in-duce transfer by pressure differential. A coil of compatible metal tubing of sufficient pressure rating is positioned in the transfer line between the supply and the receiver. The coil is placed in a cold bath, such as water ice or dry ice.

Leak Detection

Whenever a system is assembled or serviced, it should be checked for leaks. There are many commercially available leak detectors. These devices are readily available through a refrigeration supply house.

A detailed discussion of leak detection, along with a list of manufacturers of leak detection equipment, is available in DuPont Bulletin ARTD-27 (H-31753-2).

Handling Precautions for FM-200® Shipping Containers

The following rules for handling FM-200® containers are strongly recommended:

• Usepersonalprotectiveequipment,suchasside- shield glasses, gloves, and safety shoes, when handling containers.

• AvoidskincontactwithliquidFM-200®; it can cause frostbite.

• Neverheatacontainertoatemperaturehigherthan 52°C (125°F).

• NeverrefillreturnablecylinderswithoutDuPontconsent.DOT regulations forbid transportation of returnable cylin-ders refilled without DuPont’s authorization.

• Neveruseamagnetorsling(ropeorchain)toliftcontain-ers. Lifting may be accomplished by the use of a safe cradle or platform basket that holds the container.

• Neverusecontainersasrollers,supports,orforanyotherpurpose than to contain FM-200®.

• Protectcontainersfromanyobjectsthatwillresultin a cut or other abrasion in the surface of the metal.

• Nevertamperwiththesafetydevicesinthevalves or container.

• Neverattempttorepairoraltercontainersorvalves.

• Neverforceconnectionsthatdonotfit.Makesurethethreads on the regulator or other auxiliary equipment are the same as those on the valve outlets.

• Keepvalvestightlyclosed,withvalvecapsandhoods in place when the container is not in use.

• Whenstoringcontainersoutside,storeunderaroofandprotect from weather extremes.

• UseavaporrecoverysystemtocollectFM-200® vapors from lines after unloading.

Table 6 Specifications for FM-200® Shipping Containers

DOT Net Weight, lb Net Weight, kg Type Dimensions Specification FM-200® FM-200®

68 kg (150 lb) 12” x 46” 4BW240 150 68

10” x 56” 4BW400 150 68

544 kg (1,200 lb) 30” x 53” 4BW260 1,200 544

30” x 56” 4BW240 1,200 544

ISO Container 8’ x 8’ x 20’ (frame 51 37,000 16,784

Tank Trailer 5,000 gallon MC-330 or -331 37,000 16,784

9

Figure 3. 68 kg (150 lb) Size Cylinder

Figure 4. Liquid/Vapor Valve

Figure 5. 544 kg (1,200 lb) Size Cylinder

10

Figure 6. Typical ISO Tank

Figure 7. Typical Bulk Storage System

11

Table 7 Weight of Nitrogen Required for Superpressurization of DuPont™ FM-200® (English Units)

Weight of nitrogen per lb of FM-200® at 70°F

Fill Density lb/ft3 360 psig 600 psig

oz oz 40 0.534 0.944 45 0.463 0.819 50 0.407 0.718 55 0.361 0.636 60 0.323 0.568 65 0.291 0.511 70 0.263 0.461 75 0.239 0.418

Table 8 Weight of Nitrogen Required for Superpressurization of DuPont™ FM-200® (SI Units)

Weight of nitrogen per kg of FM-200® at 21°C

Fill Density kg/m3 2500 kPa (gauge) 4150 kPa (gauge)

g g 600 36.4 63.8 700 30.3 53.1 800 25.7 45.1 900 22.2 38.8 1000 19.3 33.8 1100 17.0 29.7 1200 15.1 26.3

Nitrogen Superpressurization of DuPont™ FM-200®

FM-200® is shipped in cylinders that contain essentially pure FM-200®. These containers are evacuated before filling to remove air, and the FM-200® contains less than 1.5% (vol.) non-condensible gases (air, nitrogen, etc.) in the vapor space. The pressure in these cylinders is therefore due to the vapor pressure of FM-200® alone.

In fire suppression applications it is desirable to increase the available pressure above the vapor pressure of FM-200®. This is accomplished by adding nitrogen to the FM-200® either during or after transfer, and is termed “superpressurization”. Superpressurization increases the total pressure available for flow from the container through downstream piping, provides a pressure pad to keep the liquid compressed in the liquid phase during flow, and also serves to stabilize the container pressure over a wide temperature range.

To determine the amount of nitrogen required for super-pressurization of FM-200® at various fill densities, it is nec-essary to understand the solubility relationship of nitrogen and FM-200®. Extensive experimental work was conducted

by DuPont’s Central Research and Development group to develop this information. The Peng-Robinson Equation of State (PREOS) was then used to calculate the following:

• Weightofnitrogenrequiredforsuperpressurization

• Isometricdiagrams

• Henry’sLawConstants

Tables 7 and 8 provide the weight of nitrogen required to pressurize FM-200® to 360 psig (2500 kPa) and 600 psig (4150 kPa). Isometric diagrams for FM-200® are shown in figures 8 through 11.

Henry’s Law Constants

PREOS was also used to calculate the Henry’s Law Con-stants as shown in Figure 12 (English units) and Figure 13 (SI units).

12

Figure 8. Isometric Diagram – DuPont™ FM-200® Superpressurized with Nitrogen to 360 psig at 70°F

Figure 9. Isometric Diagram – DuPont™ FM-200® Superpressurized with Nitrogen to 600 psig at 70°F

13

Figure 11. Isometric Diagram for FM-200® Superpressurized with Nitrogen to 4150 kPa at 21°C

Figure 10. Isometric Diagram for FM-200® Superpressurized with Nitrogen to 2500 kPa at 21°C

14

Figure 13. Henry’s Law Constant for Nitrogen Solubility in DuPont™ FM-200® (SI Units)

Figure 12. Henry’s Law Constant for Nitrogen Solubility in DuPont™ FM-200® (English Units)

15

Recovery, Reclamation, and DisposalResponsible use of FM-200® requires that the product be recovered for reuse or disposal whenever possible.

Recovery

Recovery refers to the removal of FM-200® from equip-ment and collection in an appropriate container. Recovery does not involve processing or analytical testing. Recovery is normally performed when a system must undergo main-tenance and the FM-200® is then returned to the system after completion. There are a number of recovery devices on the market. These devices contain a compressor and an air-cooled condenser, and may be used for liquid and vapor recovery. Before purchasing a specific recovery unit, check with the manufacturer to be sure that it contains the elastomeric seals and compressor oil compatible with FM-200®.

Reclamation

Reclamation refers to the reprocessing of FM-200® recov-ered from a system to new product specifications. Quality of the reclaimed product is verified by chemical analysis. In the United States FM-200® is included in DuPont’s recla-mation program. Contact DuPont for further information.

Disposal

Disposal refers to the destruction of used FM-200®. Disposal may be necessary when FM-200® has become contaminated with other materials and no longer meets the acceptable specifications of DuPont or other reclaimer. DuPont does not presently accept severely contaminated FM-200® for disposal; licensed waste disposal firms are available. Be sure to check the qualifications of any firm before sending them used FM-200®.

K23261 (09/09)

Copyright © 2009 DuPont or its affiliates. All rights reserved. The DuPont Oval Logo, DuPont™, The miracles of science™, Nordel®, Hypalon®, Viton®, Hytrel®, FM-200®, and FE™, are registered trademarks or trademarks of E. I. du Pont de Nemours and Company or its affiliates.

NO PART OF THIS MATERIAL MAY BE REPRODUCED, STORED IN A RETRIEVAL SYSTEM OR TRANSMITTED IN ANY FORM OR BY ANY MEANS ELECTRONIC, MECHANICAL, PHOTOCOPYING, RECORDING OR OTHERWISE WITHOUT PERMISSION OF DUPONT.

The information set forth herein is furnished free of charge and is based on technical data that DuPont believes to be reliable. It is intended for use by persons having technical skill, at their own discretion and risk. The handling precaution information contained herein is given with the understanding that those using it will satisfy themselves that their particular conditions of use present no health or safety hazards. Because conditions of product use are outside our control, we make no warranties, express or implied, and assume no liability in connection with any use of this information. As with any material, evaluation of any compound under end-use conditions prior to specification is essential. Nothing herein is to be taken as a license to operate under or a recommendation to infringe any patents.

For further information regarding DuPont Fire Extinguishing Agents, contact: www.cleanagents.dupont.com

AmericasDuPont FluoroproductsChestnut Run Plaza 702-1274EP.O. Box 80702Wilmington, DE 19880Tel: (800) 473-7790

Europe, Middle East and AfricaDuPont de Nemours International S.A.2, Chemin du PavillonCH-1218 Le Grand-SaconnexGeneva, SwitzerlandTel: 41-22-717-5296

AsiaDuPont Taiwan, Limited13F, Hung Kuo Building167 Tun Hwa North RoadTaipei, Taiwan 105ROCTel: 886-2-25144488

j002_001r00_investigacion_bibliografica.pdf

Documents

discharge of hfc

ea agent

charges of hfc

ea molecule

clean agent discharge

typical battery of hfc

ea cylinders

vapour pressure curve