Discharge of Fire Suppression Agents to Protected Areas Bibliography Research and & the State of the Art

F.Javier Garc a Garc a 31st January 2011

Abstract This paper aims to show the present State of the Art regarding the use of clean agents, with special emphasis on HFC-227ea, for the suppression of res within enclosed areas. This method of suppression is based on the fast discharge of the N2 -pressurized clean agent contained in industrial cylinders, through a net of pipes, into de protected room. To this end, the existing literature covering the subject of re extinguishing systems based on clean agents is surveyed and a synoptic evaluation presented. Special emphasis is put on papers covering the hydrodynamics of clean agent discharge, as this will be the subject of a complementary work to be presented by the author. Initially a brief account of the historical reasons leading to the search of clean agents is oered, namely the Earths ozone layer depletion; then the problems of high Global Warming Potential (GWP) of HFCs is discussed. Next, the main general principles on which this re suppression technique is based are discussed. The literature is reviewed, covering sequentially the subjects of hydrodynamics of agents discharge, clean agents extinguishing eciency, friction models and thermophysical properties of HFC-227ea. Finally, some conclusions are drawn regarding the necessity of developing mathematical models for agent discharge, what justies the elaboration of one such model in a separated research work. Keywords: re suppression, clean agent discharge, transient uid dynamics, HFC-227ea, Halon replacement

CONTENTS

Contents
List of Figures 1 Introduction 2 Description of the Technique 3 Literature Review 3.1 Hydrodynamics of Discharge . . . . . . . 3.2 Extinguishing Eciency . . . . . . . . . . 3.3 Considerations About Friction . . . . . . . 3.4 Thermophysical Properties of HFC-227ea 4 Conclusions Bibliography Appendices A Properties of HFC-227ea 4 5 9 18 18 29 31 32 34 35 42 42

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

LIST OF FIGURES

List of Figures
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 The largest Antarctic ozone hole recorded as of September 2006. Thermophysical properties of most used extinction agents. . . . . Chemical formula of HFC-227ea molecule. . . . . . . . . . . . . . 3-D image of HFC-227ea molecule. . . . . . . . . . . . . . . . . . Typical battery of HFC-227ea cylinders for re extinction applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A nozzle discharging clean agent to atmosphere. . . . . . . . . . Typical small installation of HFC-227ea re extinction. . . . . . . A Typical installation to protect an engines room. . . . . . . . . Vapour pressure curve of HFC-227ea. . . . . . . . . . . . . . . . . Discharge of HFC-227ea agent in a protected room. . . . . . . . . Isometric diagram for HFC-227ea superpressurized with N2 to 2500 kPa at 21. . . . . . . . . . . . . . . . . . . . . . . . . . . . Isometric diagram for HFC-227ea superpressurized with N2 to 4150 kPa at 21. . . . . . . . . . . . . . . . . . . . . . . . . . . . Experimental arrangement in Pitts et al. ([PYG+ 94]) experiments. Temporal variations of internal pressure during downward discharges of HFC-227ea. . . . . . . . . . . . . . . . . . . . . . . . . Comparison of measured pressure-time history of a HFC-227ea discharge with prediction. . . . . . . . . . . . . . . . . . . . . . . Two time traces of near-eld dynamic pressure for releases of HFC-227ea. Trace B is oset by 100 kPa. Conditions: A, 476 g, release pressure 4.45 MPa; B, 489 g, release pressure 4.41 MPa. . Dynamic pressure measured 1.3 m from the vessel for HFC-227ea. For clarity, trace B is oset by 100 kPa. Conditions: A, 489 g, release pressure 4.41 MPa; B, 476 g, release pressure 4.45 MPa. . Pressure variation inside the cylinder and at nozzle for HFC227ea discharge. . . . . . . . . . . . . . . . . . . . . . . . . . . . Pressure variation inside the cylinder and at nozzle for clean agent discharge. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Actual and estimated pressure-time behaviour in a tight 34 m3 room upon HFC-227ea discharge to yield 5.9%. . . . . . . . . . . Development of room pressure during the discharge of 30.9% v/v IG541. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Development of room pressure during the discharge of 6.1% v/v HFC-227ea. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Predicted and experimental enclosure pressures for a 8 s, 9.8% of HFC-227ea discharge. . . . . . . . . . . . . . . . . . . . . . . . . Temperature comparison for HFC-227ea extinguished res with and without WSCS. . . . . . . . . . . . . . . . . . . . . . . . . . Extinction system with very small length of discharges pipe . . . 6 8 10 11 12 12 13 14 15 16 17 17 19 20 21

22

17

23 24 25 27 28 28 29 31 34

18 19 20 21 22 23 24 25

INTRODUCTION

Introduction

This report aims to summarise the current state of the art regarding the hydrodynamical modelling of processes in which a liquid is kept highly pressurised by a seemingly ideal gas, both contained within a vessel provided with a fast valve, which is suddenly opened and the liquid released to atmosphere at high speed through a length of discharges pipe. A special emphasis is made on the theoretical and mathematical models which attempt to explain such phenomena. In the realm of re protection engineering it is not infrequent the use of highly pressurized extinguishing agents that are suddenly discharged to the protected areas in case of re. This very eective method of re extinction is carried out with agents that normally are in gaseous state at standard conditions of pressure and temperature, although within the pressurized vessels can very well be in liquid state, and thus modelled as incompressible uids. A number of years ago Halons were the main (if not the only) extinction agents used to prevent the initiation of ames within small-to-medium rooms in which res were deemed specially catastrophic, such as computer rooms, critical documentation vaults, museums restoration halls, process control rooms, etc. They were so eective in stopping the combustion reactions, that little discussion existed about the possibility of nding some other kinds of extinction agents. As it was established the role played by BFCs and CFCs in ozone layer depletion, and it became evident that these compounds were doomed for any industrial application as a result of 1987 Montreals Protocol, a vast scientic and engineering endeavour was originated with purpose of nding new substances that could substitute Halons in such important applications. Thus, there was established the Halon Options Technical Working Conferences (HOTWC), or the International CFC and Halon Alternatives Conferences and Exhibitions, and similar scientic and engineering meetings. Probably, the most ambitious program was the Next Generation Fire Suppression Technology Program (NGP) of the National Institute of Standards and Technology (USA), which ran from 1991 to 2006 under the supervision of renowned scientists like Richard G. Gann and others (see [Gan95], [Gan98], [Gan03], [Gan05])

INTRODUCTION

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 were found that fullled acceptably the role of Halons for re 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 have no noticeable eect on Earths ozone layer. Some of these agents actually showed design concentrations to quench test res not too dierent from those corresponding to Halons. This last feature is quite desirable, because it means that, in most cases, the existing installations (pipes, valves, manifolds, etc.) could be preserved and only the actual agents cylinders had to be replaced, thus making more aordable the initiative of banning Halons and having them replaced by less environmentally hazardous compounds. Notably, one of the most suitable agents (with an acceptable price tag) was found to be HFC-227ea, which can readily be used as a replacement agent in existing installations of Halon 1301 (see Senecal 2001, [Sen01]). But once the problem of Earths ozone layer depletion was given a solution, the world became conscious of a new threat to the planet: the global warming

INTRODUCTION

of atmosphere and the climate change it might engender. A great number of studies were conducted in order to nd the greenhouse eect attached to any gaseous substance that might be released to atmosphere. Nowadays every industrial compound exhibits in its Material Safety Data Sheet the degree of global warming potential it carries. The most eective clean agents, from extinction powers point of view, normally exhibit a very high degree of global warming potential (GWP), several thousand times higher than CO2 (see gure 2). With raising worldwide concern about climate change, it is only natural to try to nd agents as eective as those, but much less hazardous for the environment. While this research is being conducted, with varying degree of success, growing concern arises in rationalizing the use of existing agents, with the aim that just the strictly sucient amount to extinguish a potential re is released to the atmosphere, and nothing else. There exists founded suspicions that, probably, present regulations and standards concerning such re protection techniques, do not adequately address the necessity of reducing the amount of agent to be released to environment. to the actual minimum which guarantees re extinction. Thus, in the long term, those Clean Agents were not so clean, after all, and a new endeavour is again necessary to replace them with other agents that could undertake the task. In the meantime, the emphasis is set on techniques that might improve the eciency of existing agents, allowing for smaller amounts of them to guarantee the same extinction performance. This could very well be obtained with improved methods of discharge, that take into account the specic hydrodynamic features of this process. This justies the necessity of developing new models that could be used in the design of better systems, which fully exploit the intrinsic capabilities of agents to quench res.

INTRODUCTION

Figure 2: Thermophysical properties of most used extinction agents.

Although it there exist many empirical research regarding the eectiveness of these agents to extinguish various kinds of re (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 the hydrodynamics aspects of the phenomenon, specially from the transient point of view. It is our opinion that a deeper understanding of the dynamical process of agent release, will help in reducing and optimizing the amounts of agent that presently are regulated and established as suitable for extinction applications. The information summarised in this State of the Art report is gathered with the intention of producing a work devoted to such endeavour: to provide a physicomathematical model, and the associated dynamical equations, which will permit

DESCRIPTION OF THE TECHNIQUE

the description, comprehension and predictions of the free discharge process of a highly pressurized liquid agent into atmosphere. It is also expected that this model will open the possibility of incorporating the results with it obtained, into the actual regulation and standards which have relation with this eld of re protection engineering. The actual model will be presented in a separate paper, while this work is devoted to investigate the ndings that exist in the literature regarding this phenomenon. Although the model could be used for a number of extinction agents, it will be particularized for HFC-227ea, as it is of widespread use in industry and its GWP 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, Pitts et al. 1994, [PYG+ 94], or Cooper 1993, [Coo93]) the treatment found in the literature has always been from a hydrodynamic steady state perspective. To our knowledge, there is not yet a model for the problems transient hydrodynamic behaviour. It is our opinion (and the results might conrm it) that such a fast process as this agents discharge (which normally takes less than 10 s, with a very fast evolution in the rst second) could not be accurately described with the equations of steady state hydrodynamics. Instead, full transient regime equations must be used, which will allow for new and interesting hydrodynamic behaviour to emerge. Although simple, the proposed model still has taken into account most aspects that are relevant for this application: 1. Friction. 2. The compression of an ideal gas as the driving force which sets the liquid into motion. 3. The liquids nite volume within the vessel. 4. The phenomenons very fast nature and the transient treatment of it. A very recommendable handbook dealing with the general principles of transient uid dynamics is Durst 2008, [Dur08].

Description of the Technique

A very well tested technique for re-ghting, used throughout the world, consists of discharging an agent very rapidly inside the room which is meant to be protected. It is specially adequate for the very initial stages of re development,
1 HFC-227ea has a GWP of 2900, this is, 1 kg of HFC-227ea has the equivalent global warming eect of 2.9 tons of CO2 . Other extinction agents have a much higher GWP: for instance, HFC-23 has 11700 (14800 for a 100 years period). See thermophysical properties of HFC-227ea in appendix A.

DESCRIPTION OF THE TECHNIQUE

10

According to the nature and/or amount of agent released to the ambient, the re is quenched by one or more of the following mechanisms2 : Heat removal: The combustion reaction is halted due to sudden decrease of temperature, below ammability limit. The agent concentration does not need to be very high, as long as it exhibits enough cooling properties. Oxygen removal: The chemical reaction is halted due to the scarcity of a basic reactive. The agents concentration needs to be very high, as it must displace oxygen from the room below the suocation limit. Reaction hindrance: The combustion reaction is hindered, and eventually halted, due to new chemical compounds that interfere with the radicals created in the reaction chain. The agent concentration does not need to be very high, as long as it has the property of forming intermediate compounds with these radicals. Carbon-based clean agents relies primarily on the rst mechanism, being the others of secondary importance. Inert-gas clean agents use mostly the second mechanism. The third mechanism is typical of Halons. The agent primarily concerned with in this work is HFC-227ea, also known as R227, Heptauoropropane, HFP, C3 HF7 , CF3 CF H CF3 , etc. It is also commercialized under the trade registered names of FM-200, FE-227, Solkaam 227, MH-227, NAF S 227, etc. It is a colourless and odourless gaseous halocarbon. Agent HFC-227ea adopts the chemical formula depicted on gure 3. It has a boiling point of -16.4 at atmospheric pressure, and it is a gas at room temperature. It is slightly soluble in water (260 mg/L).

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

fourth obvious mechanism, removing combustibles from ame, is not mentioned because automatic re extinction systems seldom use it.
2A

DESCRIPTION OF THE TECHNIQUE

11

The molecule CF3 CF H CF3 is a reagent to generate sources of radical CF3 CF CF3 by deprotonation. This ability is clearly understood by considering its 3D molecular model, gure 4 The basic features of automatic re extinction systems based on HFC-227ea could be found in general reference handbooks such as SFPEs [DiN02] and NFPAs [DiN03]. Some standards that regulate the design, installation, use and maintenance of such systems are: NFPA-2001 ([NFP08]), ISO-14520:2006 ([ISO06]), CEA-4045 ([CEA05]), VdS-2381 ([VdS09]), etc. In Spain the regulations are UNE-EN 15004-5:2009 ([UE09a], equivalent to ISO-14520) and UNEEN 15004-5:2009 ([UE09b]).

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

Commercially, the agent HFC-227ea is supplied in cylinders of dierent standardized volumes, normally at 24 bar of relative pressure at normal room temperature. The pressure is maintained with compressed gas N2 , being both HFC-227ea and N2 within the cylinder. Should the application demand more than one cylinder, several of them could be deployed together forming a battery, as shown on gure 5. In such cases, normally the pressure of one cylinder is used to trigger the next one in the battery. The cylinders are connected, via rigid pipe (normally with high pressure rating), to a number of nozzles which warrant the adequate distribution of agent within the protected room (see gure 6). The trigger which releases the agent is normally an automatic re detection system, usually of the very-early-warning type.

DESCRIPTION OF THE TECHNIQUE

12

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

Figure 6: A nozzle discharging clean agent to atmosphere.

Figure 7 shows a typical small installation that demands just one cylinder to guarantee re extinction. The elements which constitute an automatic re suppression 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 a manometer.

DESCRIPTION OF THE TECHNIQUE

13

Valve. This fast opening valve releases the agent to the ambient. It could be electrically or pneumatically activated. Usually the cylinder has a manual 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 uniformly enough inside the protected room. Fire detectors. It could detect the presence of smoke, rising temperature, ame, or any other property associated with re. It sends an alarm signal to the re detection panel. Detection panel. Called Releasing Panel in gure 7, it is the electronic unit which receives the early-warning alarm signal from re detectors, and triggers the agents release, according to whatever sequence has been programmed. Horn/strobe. An audible/visible warning signal for the people inside the room, announcing that agents release is coming. People must leave the room and lock the door upon hearing this horn. Manual release. A push-button which sends a signal to the Detection Panel, causing the release of agent to the room.

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

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

DESCRIPTION OF THE TECHNIQUE

14

Figure 8: A Typical installation to protect an engines room.

Fire suppression with agent HFC-227ea is mainly based in its ability to vaporize and rapidly remove heat from its surroundings. This causes a sudden temperature decrease in the protected room, which plunges below the ammability limit, thus quenching any attempted re 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, what explains the ashing behaviour of HFC-227ea as it exits the discharge nozzles. In the excellent photograph of gure 10 it can be seen how the agent rapidly vaporizes as it displaces some 1 m away from the nozzle.

DESCRIPTION OF THE TECHNIQUE

15

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

The calculation of heat removal from ambient as HFC-227ea vaporizes during the discharge process, is suitable described in general handbooks such as SFPEs [DiN02] and NFPAs [DiN03]. This cooling eect is not the only suppression mechanism taking place in agents discharge: the high speed uid also removes partially some oxygen from the re, and disperses and mixes with pyrolysis products, thus interfering with combustion reaction.

DESCRIPTION OF THE TECHNIQUE

16

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

It is worth mentioning the interesting experimental work related with re suppression, Blanchat et al. 2008, [BBFY08], which shows with detail the successive stages which takes place in a typical re 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 between the uid mixing time scale and the ame chemical time scale. Flames are extinguished when the uid mixing time is short relative to the chemical time required for combustion to occur. This justies the importance of designing discharge systems that confer a high velocity to the agent. The main mechanism to achieve the desired discharge velocity is to N2 -pressurize correctly the agent in the cylinders. Thus, a certain amount of N2 must be added to obtain the right pressurization, but not an excessive one, as the pressure within the cylinders would raise rapidly as room temperature grows, perhaps above the pressure rating of cylinders. The relation between N2 and HFC-227ea inside the cylinder is called lling density (kg/m3 ). In order to understand the importance of this pressurization, gure 11 shows the pressure behaviour of a typical cylinder as a function of temperature, taking into account several lling densities, This diagram has been drawn for a cylinder with 2500 kPa at 21 . Alternatively, the agent could be initially N2 -pressurized to 4150 kPa at 21 , and gure 12 shows the corresponding isometric diagram. The importance of an isometric diagram is that it determines the maximum lling density for an agent in a cylinder with a xed pressure rating. The basic rule

DESCRIPTION OF THE TECHNIQUE

17

is that the cylinder must not become liquid full at 54 (for U.S. DOT) and/or the 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 to 5/4 of the design working pressure. which is 3445 kPa at 54 . For HFC-227ea, as seen in gure 11, this yields a maximum lling density of 1153 kg/m3 . The 2500 kPa pressurization level is not based on characteristics of halon replacement halocarbons, but is a vestige of standard Halon 1301 systems.

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

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

LITERATURE REVIEW

18

A secondary mechanism to achieve a high discharge velocity, complementary to the former, is an optimized piping design. It is thus necessary to have good dynamical models of piping ow which take into account the transient behaviour of the discharge process. Also, it is important that those models describe accurately the dependence of the transient friction coecient on ow parameters, as friction is the most important eect opposing agents motion. Thus, it makes sense to develop such dynamical models of transient ow through pipes, as they will be necessary in any eort to optimize current design of automatic re suppression systems with clean agents. The following section is devoted to assess the current State of the Art of most aspects of these extinction systems. A new mathematical model will be introduced in an additional research work, which derives the transient dynamical equations of agents ow directly from the First Principles.

Literature Review

Although there exist a large number of works devoted to study and/or document the eectiveness of HFC-227ea (and other extinction agents) for various kinds of test res, few articles can be found conducting extensive research on the hydrodynamics behaviour of the discharge process of such agents to protected areas.

3.1

Hydrodynamics of Discharge

Grosshanler et al. 1994, [GGP94], is probably the most comprehensive work ever written on the subject of agents discharge over res in protected areas. It covers most areas of research which have a relationship with this issue: thermodynamical properties of agents, hydrodynamical modelling of discharge, hydrodynamical experimentation of these phenomena, ame suppression eectiveness of such agents, ame inhibition chemistry, agent stability under storage and discharge residue, corrosion of metals, elastomer seal compatibility, and human exposure and environmental impact. It contains interesting developments from the theoretical point of view, viable numerical models, ingenious experimental set-ups, and valuable results and ndings. The chapter devoted to hydrodynamics of agent discharge, written by Pitts et al. 1994, [PYG+ 94], is exhaustive and covers most aspects of this subject. It denes a mathematical model for an experimental set-up in which the process can be studied and monitored, and the models predictions compared with the measurement results. It is complete, full and detailed, and the model is quite analytical, but it is based on the equations found in Shapiro 1953, [Sha53], which are developed under the assumptions of isentropic steady regime ow, and it does not take into account the transient ow equations which are deduced in Shapiro 1954, [Sha54].

LITERATURE REVIEW

19

The work describes an experimental apparatus, shown on gure 13, and the associated mathematical model that explains the agents behaviour as it is discharged through it. This approach to the problem considers the agent as a compressible uid and makes use of the equations for steady isentropic compressible ow. The model takes into account the possibility of agent ashing.

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

It considers also a discharge ow coecient for the orice (nozzle) through which the agent discharges to ambient, according to [Sha53]. The dimensionless equations are then presented and solved numerically, for a set of initial values typical of standard extinction installations. The results are depicted in graphs and then compared to actual measurements. There is a comprehensive set of curves which shows most of the situations that can be subjected to experimental control.

LITERATURE REVIEW

20

The experiments are conducted with a number of dierent 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 only the results concerning HFC-227ea will be highlighted. One of such curves is shown on gure 14, which represent the results for the temporal variation of the internal pressure during downward discharges of HFC227ea. Pressures reported in the gure are gauge pressures. The pressure is nondimensionalized by the actual burst pressure, Pi , which is taken to be the pressure at t = 0 s. The curves show two distinct regions, separated by an inection point. The rst region corresponds to the time interval during which the liquid agent is being propelled from the vessel. The second region corresponds to the period when the remaining vapour (mostly N2 ) is being vented from the vessel. The inection point corresponds to the time at which the liquid agent has just been completely expelled from the vessel.

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

In the work there are abundant pictures taken from high speed cameras. These pictures show with a great level of detail the development of ow in the discharge process. Figure 15 shows the comparison of the actual measured values of discharge pressures for HFC-227ea with the models predicted results. Despite not using transient ow dynamical equations, the agreement is excellent.

LITERATURE REVIEW

21

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

Perhaps the most interesting results, from the hydrodynamical point of view, are shown in gures 16 and 17. They correspond to the dynamical pressure curves as measured by two pressure transducers: one located near the exit orice (some 13 mm downstream, gure 16), and another located further away (some 1.3 m downstream of the vessels exit, gure 17). Although air entrainment alters the jet, far-eld curves are not too dierent from near-eld ones.

LITERATURE REVIEW

22

Figure 16: Two time traces of near-eld dynamic pressure for releases of HFC227ea. Trace B is oset by 100 kPa. Conditions: A, 476 g, release pressure 4.45 MPa; B, 489 g, release pressure 4.41 MPa.

LITERATURE REVIEW

23

Figure 17: Dynamic pressure measured 1.3 m from the vessel for HFC-227ea. For clarity, trace B is oset by 100 kPa. Conditions: A, 489 g, release pressure 4.41 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 dierent agent, R22, with dierent properties than HFC227ea. As remarked above, the discharge treatment is done under the assumption of steady compressible ow. Also Pitts et al. 1993, [PYBG93], develops this same issue from an experimental point of view, but it is less exhaustive than [PYG+ 94]. Wysocki 1996,[Wys96], studies also the discharge process of clean agents, including HFC-227ea. It uses Bernoulli equation in the Hessons adaptation form, and does not consider explicitly the transient nature of dynamics, although most of the work is experimental rather than theoretical. Most dynamical curves are plotted with percentage of agent discharged in abscissa, rather than time. There is just one curve with time in abscissa, namely the time variation of the pressure, both in cylinder and in nozzle, corresponding to the discharge test of a cylinder of HFC-227ea with lling density of 1121 kg/m3 , see gure 18. This

LITERATURE REVIEW

24

pressure versus time tracing taken during a HFC-227ea discharge shows the continual variation in cylinder pressure and a lesser variation in pressure at the nozzle during the course of discharge. The peak in the cylinder pressure at 2.5 s indicates the beginning of ashing, in which agents ow changes from singlephase to two-phase. The peak in the nozzle pressure trace at 9.6 s indicates a change from two-phase ow to vapour ow.

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

DiNenno et al. 1994,[DFF+ 94], develops an interesting study of the clean agent discharges process, based on the model developed for Halon 1301 by Elliot et al. 1984, [EGK+ 84]. It oers a mathematical model for the phenomenon, but this model does not take into account the transient dynamical equations, and it rather considers some quasi-steady ow. The agent is assumed to ow in liquid, two-phase and vapour modes. It shows very interesting experimental pressure data for the rst few seconds of process, as it can be seen in gure 19. In this gure the discharge ow has been divided into ve distinctive sections.

LITERATURE REVIEW

25

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

Zalosh-Wang 1996, [ZW96], develops a mathematical model based on mass and energy conservation, in which no transient dynamical equation is considered. The article shows curves of HFC-227ea concentration versus time, both obtained from analytical calculations and from actual test results. The agreement is good in a long time lapse, but no comparison is made between experimental and theoretical results in the rst few seconds. Actually, the discharge concentration curves are modelled as straight lines. A theoretical pressure curve is oered, but it corresponds to room pressure, not to pipe ow pressure. Thus, scarce information is oered about the phenomenons dynamics. Bird et al. 1994, [BGH+ 94], measures the discharge process of HFC-227ea in 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 describes the computer model used to calculate the discharge process, but the program and/or numerical results are not shown on the paper. TFA is the result of a modication to an already existing computer program for modelling Halon 1301 discharge, described in Coward et al. 1992, [CHM92]. TFA is further explored in Bird et al. 1995, [BGF+ 95], which studies the discharge of HFC-227ea in large piping distribution systems, with several cylinders (4 to 14) installed in battery. The discharge pressure is monitored in up to 9 nozzles, and also inside the pipe. TFAs predicted results are compared to actual measurements. No mathematical model is presented in this work.

LITERATURE REVIEW

26

The ow within pipes during a discharge process is studied in Cleary et al. 1994, [CGY94], which characterizes the ow as transient and two-phase. The work includes an analysis of the discharges rst fractions of a second, but shows only the experimental curves with no theoretical result nor computer program. Pipe pressure traces, corresponding to a transducer located some 2 m away from discharge point, show an interesting transient spike phenomenon at around 700 ms, probably related with N2 owing alone in the pipe. Using the same experimental set-up as in [CGY94], a new paper was published by Cleary et al. 1995, [CYK+ 95], in which the discharge process was studied after the agent exits the pipe and is released into test engines nacelles. The pressure transient spike in the pipe transducers is also present, surely due to the presence of N2 alone in the pipe as the liquid agent runs out of the vessel. In this paper the time range under study is longer than that in the previous work, but again no equations nor model was oered. It is based on the model for Halon 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]. The program is named FSP and it represents an improvement of the computer model presented in [EGK+ 84]. It covers transient phenomena and two-phase ow. The computer model is tested against the same experimental apparatus used in [CGY94] and [CYK+ 95], providing acceptable results when the predicted curves are compared with the actual experimental data. Also transient experimental phenomena are studied in Wysocki-Christensen 1996, [WC96], although it treats inert agent IG5413 , not HFC-227ea. It is mentioned here for its study on pure gas discharge, without consideration for ashing or two-phase ow. Some experimental curves are oered which could be compared with those corresponding to HFC-227ea. A dierent approach is oered in Senecal-Prescott 1995, [SP95]. It shows the results of experiments in which it is measured the pressure variation within a one-way tight room that undergoes the discharge of agent HFC-227ea. The pressure is monitored for relatively long times (some 30 s), compared to the actual discharges time. An interesting outcome of this study is that pressure within the room turns negative (below atmospheric) for some time after discharge (see gure 20). Also the agent dispersion throughout the room is observed and monitored.
3 IG541 is a blend of inert gases, Ar 40%, N 52%, CO 8%, and it acts suocating the 2 2 ames, this is, removing oxygen from re. This agent is stored at very high pressures (normally 200 or 300 bar) and is never in liquid state within the cylinders.

LITERATURE REVIEW

27

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

This same negative pressure phenomenon is reported on Robin et al. 2005, [RFS05]. This work oers a comparison in the pressure curves relative to discharge of IG541 (which is always in gaseous state) and HFC-227ea, which exhibits a ashing behaviour and vaporizes entirely. Figure 21 shows a positive pressure within the room at every instant after IG541 discharge, while gure 22 shows alternatively positive and negative room pressure after HFC-227ea discharge.

LITERATURE REVIEW

28

Figure 21: Development of room pressure during the discharge of 30.9% v/v IG541.

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

LITERATURE REVIEW

29

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

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

Of a dierent nature are other works which deal with the transient hydrodynamics phenomena from another point of view, namely, these extracted from the rupture of a highly pressurized pipe. Although those processes are dierent from the one considered here, some of the equations and the general treatment of the problem might oer clues on how to approach the discharge phenomena studied in this work. Denton 2009, [Den09], is a PhD Thesis which focuses very much on transient ow, although it is not initially addressed to the Fire Protection Engineering Community. It develops a CFD program to simulate highly transient ows, and it oers the results of such simulations for a number of interesting cases, notably the discharge of pressurized uids through orices. Other works which deal with highly transient ow in pipes, in a similar fashion 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 Eciency

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

LITERATURE REVIEW

30

Grosshandler et al. 1994, [GGP94], still constitutes the most comprehensive study on every aspect of re quenching with clean agents. Hamins et al. 1994, [HGG+ 94], to be found within the book [GGP94], is an extensive work which analyses carefully most re extinguishing mechanisms. It includes chemical models of reaction inhibition, comprehensive experimental ndings, and a great number of curves which could be quoted to validate future models in this eld. Bennett 2006, [Ben06], is a very advisable primer for readers who wish to take a rst contact with the subject. It clearly establishes the principles that determine what makes an agent good for re extinguishing. Those principles are supported on experimental evidence which is oered on the work. Linteris 2006, [Lin06], oers a complete account of chemical ame inhibition mechanisms, rather than re suppression. It outlines the conditions for which a chemically active agent can be eective, and when it is most eective. These general principles are demonstrated with numerical and experimental data for dierent kinds of re. Moghtaderi et al. 1998, [MDK98], studies the ignition delay caused in res by agent HFC-227ea. It presents a number of experimental ndings regarding ame inhibition, and characterizes some chemical reactions occurring in the interaction of HFC-227ea with methane inside a ame. Saso 2001, [Sas01], also studies the inuence of inhibitors in res, with special emphasis on catalysis. When the concentrations of the fuel and the oxidizer in a combustible mixture are kept constant, the global reaction rate is controlled by the ame temperature, the global frequency factor, and the global activation energy. These factors are changed and the results are plotted in numerous graphs. Kim et al. 1996, [KSMKK96], oers the results of a real extinction test with clean agents, namely HFC-227ea and HCFC Blend A4 . conducted in a real installation under total ooding technique. The acid gases generated during re suppression are also measured and analysed. Robin et al. 2001, [RRC01], presents a comparison among the extinction power of dierent clean agents, for various classes of test res. Results are given regarding the concentration of agent necessary to warrant a re quenching. The tests under comparison include also inert gases like IG541. HFC-227ea results in one of the clean agents which demands less concentration, thus proving it is one of the most ecient agents. Sheinson et al. 2003, [SAA+ 03], and Sheinson et al. 2004, [SAAM04], introduce a new technique for re suppression, namely the simultaneous discharge of HFC227ea (called HFP in these work) and a very ne water mist (called WSCS),
Blend A, also known as NAF S-III, is composed of 82% of HCFH-22 (CHClF2 ), 9.5% of HCFC-124 (CHClF CF3 ), 4.75% of HCFC-123 (CHCl2 CF3 ), and 3.75% of isopropenyl-1-methyl-cyclohexane (C10 H16 )
4 HCFC

LITERATURE REVIEW

31

which helps in cooling the ambient in which the re develops. Figure 24 shows the improvement on extinction eciency obtained with this method: the two curves correspond to the extinguishing of a 1900 kW re of methyl alcohol with and without WSCS (measurements taken at 4.0 m height). Figure 24 demonstrates that WSCS reduced both the peak overhead temperature and the temperature several minutes after HFC-227ea discharge. This reduction protects the compartment by helping decreasing the compartments temperature below the ashpoint of the fuel.

Figure 24: Temperature comparison for HFC-227ea extinguished res with and without WSCS.

Yang-Keyser 2006, [YK06], makes an interesting study about the times necessary for extinguishing, with dierent agents and test res, for aircraft engine nacelles.

3.3

Considerations About Friction

One key aspect that deserves to be taken into account in the hydrodynamical study of the clean agents discharge process, is the calculation of friction within the pipe. Most authors use correlations that are suitably tested for steady ows, but very little (if anything) is found on the suitable friction factor correlation to be used in case of highly transient ows. The most common empiric correlation regarding friction factor f , namely the Colebrook-White formula, relates f with Reynolds number Re, which in turn

LITERATURE REVIEW

32

relates to velocity. This makes the friction factor f time dependent. The same could be said about the various correlations that have been proposed in the literature for the friction factor (see, for instance, [Bra09], [Cla09], [Den09], [ED98], [GS08], [KM86], [KY78], [RK06], [SG06], [TCK05],...). But all these empirical formulae have been devised for situations of steady state ow. To our knowledge, no such correlation has ever been introduced for transient ows, what constitutes no surprise, given the unmanageable amount of variables which take part on even the simplest instances of transient ows (such as the one considered here: clean agents discharge through pipes). What seems fairly well established is that friction stress in transient phenomena is higher (some would say much higher) than their steady ow counterparts. A detailed study of the friction factor correlations available in the literature can be found in the excellent handbook of Rohsenow et al. 1998, [RHC98]. The tables 5.8 and 5.9 within chapter 5, written by Ebadian-Dong, [ED98], oer a comprehensive collection of such correlations. The case treated here, discharge of clean agents through pipes, correspond to table 5.8, smooth circular ducts. Bratland 2009, [Bra09], presents a very extensive discussion on the friction factor for transient ows, although he proposes no correlation (see chapter 2). He also includes a table with a number of the correlations found on [ED98]. Denton 2009, [Den09], uses the Fanning friction factor in the equations, although the processes considered there are highly transient. Nevertheless the author recommends the usage of Chen correlation (rough pipes, page 84), rather than Colebrook-White. For smooth pipes (the case considered here) he recommends the Techo et al. correlation. Taylor et al. 2005, [TCK05], presents a extensive discussion of the eect of surface roughness on friction within pipes. It proposes new methods for determining such roughness. Kurokawa-Morikawa 1986, [KM86], introduces the concept of transient friction coecient, and it even oers some clues as to how could it be estimated, but no correlation is presented. The work introduces explicitly the uids acceleration in the expression leading to estimate the dierence between transient and quasisteady friction coecients.

3.4

Thermophysical Properties of HFC-227ea

Aside from the data supplied by Dupont (manufacturer of HFC-227ea, not many references could be found oering new ndings regarding the thermophysical properties of agent HFC-227ea. Yang et al. 1995, [YHB95] uses a model based on the extended correspondence states principle, to nd out the solubilities of N2 in several clean agents, which use this gas as a pressuring agent. It also calculates the pressure-temperature relationship for agent/N2 mixtures. The model claims to be accurate within 10% of actual experimental measurements.

LITERATURE REVIEW

33

Baginskii-Stankus 2002, [BS03], studies some thermodynamic and transport properties of HFC-227ea, namely heat capacity and thermal conductivity. It oers correlations to calculate these values within 2% of actual experimental measurements. It centres on liquid HFC-227ea. Benedetto et al. 2000, [BGS+ 00], proposes a equation of state for gaseous HFC227ea, obtained directly from speed of sound measurements. This determination of speed of sound is also valuable in those cases in which the vapour phase of HFC-227ea could choke during discharge process. Chen et al. 2003, [CHC03] oers generalized estimation equations for the thermophysical properties of saturated HFC-227ea. The estimated quantities are enthalpy, entropy, density of saturated liquid and gas, enthalpy of vaporization, and saturated vapour pressure.

CONCLUSIONS

34

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 theoretical and mathematical work seems to be missing, so that a solid framework exists into which insert the experimental ndings and the empirical conclusions. The literature review performed in this work has shown that no explicitly transient dynamical-analytical model could be found for the clean agent discharge phenomenon. Instead, there are CFD modelling and steady ow equations to describe this process. Thus it is pertinent that such a model be developed and proposed to the Fire Protection Engineering Community, as a valuable tool for improving the calculations and design of actual clean agent re suppression installations. Such a work will be presented in a paper to complement this one here. It will contain most of the issues arisen in this Bibliography Research and State of the Art investigation for this discharge process. Perhaps this dynamical model will prevent that some Engineer ever designs a re suppression system like the one shown in gure 25, in which the agents discharge is carried out with no signicant length of pipe, thus making it possible that liquid at very high speed could impinge directly on people standing by.

Figure 25: Extinction system with very small length of discharges pipe

REFERENCES

35

References
[BBFY08] T.K. Blanchat, A.L. Brown, V. Figueroa, and S.S. Yoon. Benchmark enclosure re suppression experiments and modeling. In Suppression and Detection and Applications-A Technical Working Conference (SUPDET 2008), March 2008. 6, 16 J.M. Bennett. Verication of NGP Fire Suppression Principles. In R.G. Gann, S.R. Burgess, K.C. Whisner, and P.A. Reneke, editors, Halon Options Technical Working Conferences (HOTWC)2006, volume NIST Special Publication. National Fire Protection Association, May 2006. 8, 30 E.B. Bird, H.D. Giesecke, T.J. Friderichs, A. Maranghides, and R.S. Sheinson. Results of Benchmark Comparisons of Calculated and Measured Flow Parameters for Discharges of Halon Replacement Chemicals. In Halon Options Technical Working Conferences (HOTWC)-1995. National Institute of Standards and Technology, May 1995. 6, 25 E.B. Bird, H.D. Giesecke, J.A. Hillaert, T.J. Friderichs, and R.S. Sheinson. Development of a Computer Model to Predict the Transient Discharge Characteristics of Halon Alternatives. In Halon Options Technical Working Conferences (HOTWC)-1994. National Institute of Standards and Technology, May 1994. 25 G. Benedetto, R.M. Gavioso, R. Spagnolo, M. Grigiante, and G. Scalabrin. Vapor Phase Helmholtz Equation for HFC-227ea from Speed of Sound Measurements. In Fourteenth Symposium on Thermophysical Properties, June 2000. 33 O. Bratland. Pipe Flow 1: Single-phase Flow Assurance. e-book www.drbratland.com, 2009. 32 A.V. Baginskii and S.V. Stankus. Thermodynamic and Transport Properties of Liquid HFC-227ea. International Journal of Thermophysics, 24(4):953961, July 2003. 33 A.E. Cote et al., editors. Fire Protection Handbook, vols. I & II. National Fire Protection Association, 19th edition, 2003. 29 CEA4045. Fire extinguishing systems using liquied halocarbon gases. Planning and installation. Comite Europeen des Assurances, April 2005. 11 T.G. Cleary, W.L. Grosshandler, and J.C. Yang. Flow of Alternative Agents in Piping. In Halon Options Technical Working Conferences (HOTWC)-1994. National Institute of Standards and Technology, May 1994. 26 Z.S. Chen, P. Hu, and W.L. Cheng. Generalized Estimation Equations for Thermophysics Properties of Saturated Fluids and the Saturated Properties Table of HFC-227ea. In Fifteenth Symposium on Thermophysical Properties, June 2003. 33

[Ben06]

[BGF+ 95]

[BGH+ 94]

[BGS+ 00]

[Bra09] [BS03]

[C+ 03] [CEA05]

[CGY94]

[CHC03]

REFERENCES

36

[CHM92]

R.N. Coward, J.A. Hillaert, and D.M. McCrory. Analytical methods for modeling discharge characteristics of Halon 1301 re protection systems. In Halon Alternatives Technical Working Coference-1992. National Institute of Standards and Technology, May 1992. 25 D. Clamond. Ecient resolution of the Colebrook equation. Ind.Eng.Chem.Res., 48(7):36653671, February 2009. 32 L.Y. Cooper. Dischage of Fire Suppression Agents from a Pressurized Vessel: A Mathematical Model and its Application to Experimental Design. NISTIR, (5181), May 1993. 9, 23 L.Y. Cooper. Dispersion of Fire Suppression Agents Discharged From High Pressure Vessels: Establishing Initial/Boundary Conditions for the Flow Outside the Vessel. In Liquid Atomization and Spray Systems, 6th International Conference Proceedings., volume ICLASS, July 1994. 6, 8 T.G. Cleary, J.C. Yang, M.D. King, C.I. Boyer, and W.L. Grosshandler. Pipe ow characteristics of alternative agents for engine nacelle re protection. In Proccedings on Halon Options Technical Working Conference-1995. National Institute of Standards and Technology, May 1995. 26 G.S. Denton. CFD Simulation of Highly Transient Flows. PhD thesis, University College London, July 2009. 29, 32 P.J. DiNenno, E.W. Forssell, M.J. Ferreira, C.P. Hanauska, and B.A. Johnson. Modeling of the Flow Properties and Discharge of Halon Replacement Agents. In Proccedings on Halon Options Technical Working Conference-1994. National Institute of Standards and Technology, 1994. 24 P.J. DiNenno. Halon Replacement Clean Agent Total Flooding Systems, volume SFPE Handbook of Fire Protection Engineering, chapter 4-7. National Fire Protection Association, 3rd edition, 2002. 11, 15 P.J. DiNenno. Direct Halon Replacement Agents and Systems, volume NFPA Fire Protection Handbook of NFPA Fire Protection Handbook, vols. I & II, chapter 11-2. National Fire Protection Association, 19th edition, 2003. 11, 15 F. Durst. Fluid Mechanics. An Introduction to the Theory of Fluid Flows. Springer-Verlag Berlin Heidelberg, 2008. 9 M.A. Ebadian and Z.F. Dong. Forced Convection, Internal Flow in Ducts, chapter 5. Handbook of Heat Transfer. McGraw-Hill, 3rd edition, 1998. 32 D.G. Elliot, P.W. Garrison, G.A. Klein, K.M. Moran, and M.P. Zydowicz. Flow of nitrogen-pressurized Halon 1301 in Fire Extinguishing Systems. Jet Propulsion Laboratory, JPL Publication(8462), November 1984. 24, 26

[Cla09] [Coo93]

[Coo94]

[CYK+ 95]

[Den09] [DFF+ 94]

[DiN02]

[DiN03]

[Dur08] [ED98]

[EGK+ 84]

REFERENCES

37

[Fri03]

R. Friedman. Theory of Fire Extinguishment, chapter 2-5, pages 283/296. NFPA Fire Protection Handbook, vols. I & II. National Fire Protection Association, 19th edition, 2003. 8, 29 R.G. Gann. Next Generation Fire Suppression System Technology: A National Research Plan. In Halon Options Technical Working Conferences (HOTWC)-1995. National Institute of Standards and Technology, May 1995. 5 R.G. Gann. Next-Generation Fire Suppression Technology Program. In Halon Options Technical Working Conferences (HOTWC)-1998. National Institute of Standards and Technology, May 1998. 5 R.G. Gann. Next Generation Fire Suppression Technology Program: FY2003 Progress. In Halon Options Technical Working Conferences (HOTWC)-2003. National Institute of Standards and Technology, May 2003. 5 R.G. Gann. Next Generation Fire Suppression Technology Program: FY2005 Progress. In Halon Options Technical Working Conferences (HOTWC)-2005. National Institute of Standards and Technology, May 2005. 5 W.L. Grosshandler, R.G. Gann, and W.M. Pitts, editors. Evaluation of Alternative In-Flight Fire Suppressants for Full-Scale Testing in Simulated Aircraft Engine Nacelles and Dry Bays, volume SP. National Institute of Standards and Technology, April 1994. 6, 18, 30 W.L. Grosshandler, A. Hamins, S.R. Charagundla, and C. Presser. Suppression Eectiveness Screening for Impulsively Discharged Agents. Halon Options Technical Working Conference, May 2000. 6 C.T. Goudar and J.R. Sonnad. Comparison of the iterative approximations of the Colebrook-White equation. Hidrocarbon Processing, 87(8):7984, 2008. 32 A. Hamins, G. Gmurczyk, W.L. Grosshandler, R.G. Rehwoldt, I. Vazquez, T.G. Cleary, C. Presser, and K. Seshadri. Flame Suppression Eectiveness, volume NIST-SP 861 of Evaluation of Alternative In-Flight Fire Suppressants for Full-Scale Testing in Simulated Aircraft Engine Nacelles and Dry Bays, chapter 4. National Institute of Standards and Technology, April 1994. 6, 8, 30 A. Hamins, D. Trees, K. Seshadri, and H.K. Chelliah. Extinction of Nonpremixed Flames With Halogenated Fire Suppressants. Combustion and Flame, 99(2):221230, August 1994. 6, 8

[Gan95]

[Gan98]

[Gan03]

[Gan05]

[GGP94]

[GHCP00]

[GS08]

[HGG+ 94]

[HTSC94]

REFERENCES

38

[HTSD03]

J.C. Hewson, S.R. Tieszen, W.D. Sundberg, and P.E. Desjardin. CFD modeling of re suppression and its role in optimizing suppressant distribution. In Proccedings on Halon Options Technical Working Conference-2003. National Institute of Standards and Technology, 2003. 16 ISO-14520-1. Gaseous re-extinguishing systems-Physical properties and system design-Part 1: General requirements. International Standard Organization, 2006. 11 M.E. Kim-E. The possible consequences of rapidly depressurizing a uid. Masters thesis, Massachusetts Institute of Technology, January 1981. 29 S.K. Kim, T.J. Lestina, H.D. Giesecke, and R.S. Sheinson. Development of Computer Models for the Discharge of Halon Alternatives. In Halon Options Technical Working Conferences (HOTWC)-1994. National Institute of Standards and Technology, May 1994. 25 J. Kurokawa and M. Morikawa. Accelerated and Decelerated Flows in a Circular Pipe. Bulletin of JSME, 29(249):758765, March 1986. 32 H. Kashiwagi, S. Oshikawa, J. Yui, and Y. Saso. Eect of Fire Size on Suppression Characteristics of Halon Replacement TotalFlooding Systems. In Halon Options Technical Working Conferences (HOTWC)-2001. National Institute of Standards and Technology, April 2001. 8

[ISO06]

[KE81]

[KLGS94]

[KM86]

[KOYS01]

[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. In Halon 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 Eects in Fluid Piping Systems. Nuclear Engineering and Design, 45:93 100, 1978. 32 G.T. Linteris. NGP Research on Fire Suppression Chemistry. In Halon Options Technical Working Conferences (HOTWC)-2006. National Institute of Standards and Technology, May 2006. 6, 8, 30 B. Moghtaderi, B.Z. Dlugogorski, and E.M. Kennedy. Detailed Chemical Kinetic Modelling Study on High Temperature Ignition of Methane/Air Mixtures Doped With C3F7H. In Halon Options Technical Working Conferences (HOTWC)-1998. National Institute of Standards and Technology, May 1998. 6, 8, 30 H. Mahgerefteh, A.O. Oke, and Y. Rykov. Ecient numerical solution for highly transient ows. Chemical Engineering Science, 61(15):50495056, March 2006. 29

[Lin06]

[MDK98]

[MOR06]

REFERENCES

39

[NBHT98]

T. Noto, V. Babushok, A. Hamins, and W. Tsang. Inhibition Eectiveness of Halogenated Compounds. Combustion and Flame, 112(1/2):147160, January 1998. 6 A. Nouri-Borujerdi and M. Ziaei-Rad. Numerical modeling of transient turbulent gas ow in a pipe following a rupture. Transaction B: Mechanical Engineering, 17(2):108120, April 2010. 29 NFPA-2001. NFPA-Standard on Clean Agent Fire Extinguishing Systems. National Fire Protection Association, August 2008. 11 A.O. Oke, H. Mahgerefteh, I. Economou, and Y. Rykov. A transient outow model for pipeline puncture. Chemical Engineering Science, 58(20):45914604, July 2003. 29 W.M. Pitts, J.C. Yang, R.A. Bryant, L.G. Blevins, and M.L. Huber. Characterization and Identication of Super-Eective Thermal Fire Extinguishing Agents. Final Report. NGP Project 4C/1/890. Technical Report NIST Technical Note 1440, National Institute of Standards and Technology, July 2006. 6, 8 W.M. Pitts, J.C. Yang, B. Breuel, and G. Gmurczyk. Dynamics of the release of alternate Halon replacement agents from pressurized bottles. In Halon Alternatives Technical Working Coference-1993. National Institute of Standards and Technology, May 1993. 23 W.M. Pitts, J.C. Yang, G. Gmurczyk, L.Y. Cooper, W.L. Grosshandler, W.G. Cleveland, and C. Presser. Fluid Dynamics of Agent Discharge, volume SP of Evaluation of Alternative InFlight Fire Suppressants for Full-Scale Testing in Simulated Aircraft Engine Nacelles and Dry Bays, chapter 3. National Institute of Standards and Technology, April 1994. 4, 9, 18, 19, 23 M.L. Robin, E.W. Forssell, and V. Sharma. Pressure Dynamics of Clean Agent Discharges. In Halon Options Technical Working Conferences (HOTWC)-2005. National Institute of Standards and Technology, May 2005. 27 W.M. Rohsenow, J.P. Hartnett, and Y.I. Cho, editors. Handbook of Heat Transfer. McGraw-Hill, 3rd edition, 1998. 32 A.R. Rao and B. Kumar. Friction factor for turbulent pipe ow. 2006. 32 M.L. Robin, T.F. Rowland, and M.D. Cisneros. Fire Suppression Testing: Extinguishment of Class A Fires With Clean Agents. In Halon Options Technical Working Conferences (HOTWC)-2001. National Institute of Standards and Technology, April 2001. 6, 8, 30 R.S. Sheinson, S. Ayers, R. Anleitner, D. Morse, D. Szwarc, L. Levenberry, and A. Maranghides. Combining a Water Spray Cooling System With Heptauoropropane for Total Flooding Fire Suppression. In Halon Options Technical Working Conferences

[NBZR10]

[NFP08] [OMER03]

[PYB+ 06]

[PYBG93]

[PYG+ 94]

[RFS05]

[RHC98] [RK06] [RRC01]

[SAA+ 03]

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. Heptauoropropane With Water Spray Cooling System as a Total Flooding Halon 1301 Replacement: System Implementation Parameters. In Halon Options Technical Working Conferences (HOTWC)-2004. National Institute of Standards and Technology, May 2004. 8, 30 Y. Saso. Roles of Inhibitors in Global Gas-Phase Combustion Kinetics. In Halon Options Technical Working Conferences (HOTWC)-2001. National Institute of Standards and Technology, April 2001. 6, 8, 30 J.A. Senecal. New Technology Delivery System for FM-200 Clean Agent. In Halon Options Technical Working Conferences (HOTWC)-2001. National Institute of Standards and Technology, April 2001. 6 J.R. Sonnad and C.T. Goudar. Turbulent ow friction factor calculation using a mathematically exact alternative to the ColebrookWhite equation. J.Hydraulic Engineering, 132(8):863867, August 2006. 32 A.H. Shapiro. The Dynamics and Thermodynamics of Compressible Fluid Flow, volume I. The Ronald Press Company, 1953. 18, 19 A.H. Shapiro. The Dynamics and Thermodynamics of Compressible Fluid Flow, volume II. The Ronald Press Company, 1954. 18 R.R. Skaggs. Assessment of the Fire Suppression Mechanics for HFC-227ea Combined With NaHCO3. In Halon Options Technical Working Conferences (HOTWC)-2002. National Institute of Standards and Technology, May 2002. 8 J.A. Senecal and R.C. Prescott. FM-200 Suppression Systems: A Conservative Discharge Test Method and In-Room Pressure Variance Upon Discharge. In Halon Options Technical Working Conferences (HOTWC)-1995. National Institute of Standards and Technology, May 1995. 26 J.B. Taylor, A.L. Carrano, and S.G. Kandlikar. Characterization of the eect of surface roughness and texture on uid ow-Past, present and future. Proceedings of ICMM2005, 3rd International Conference on Microchannels and Minichannels, June 2005. 32 K. Tuzla, T. Palmer, J.C. Chen, R.K. Sundaram, and W.S. Yeung. Development of computer program for re suppressant uid ow-Final technical report. Technical report, Lehigh University, September 2000. 26

[Sas01]

[Sen01]

[SG06]

[Sha53]

[Sha54]

[Ska02]

[SP95]

[TCK05]

[TPC+ 00]

REFERENCES

41

[UE09a]

UNE-EN-15004-1. Sistemas jos de lucha contra incendios. sistemas de extinci on mediante agentes gaseosos. parte 1: Dise no, instalaci on y mantenimiento. AENOR, 2009. 11 UNE-EN-15004-5. Sistemas jos de lucha contra incendios. Sistemas de extinci on mediante agentes gaseosos. Parte 5: Propiedades f sicas y dise no de sistemas de extinci on mediante agentes gaseosos con HFC 227 ea. AENOR, 2009. 11 E. Valero Sanchez. An alisis del proceso de descarga de uidos sometidos a altas presiones y temperaturas. PhD thesis, E.T.S. Ingenieros Aeron auticos, Madrid, December 1998. 29 VdS2381. Fire Extinguishing Systems using Halocarbon Gases. VdS Schadenverh utung GmbH, June 2009. 11 T. Wysocki and B. Christensen. Inert gas re suppression systems using IG541 (INERGEN). Solving the hydraulic calculation problem. In Proccedings on Halon Options Technical Working Conference-1996. National Institute of Standards and Technology, 1996. 26 T. Wysocki. Single point ow calculations for liqueed compressed gas re extinguishing agents. In Proccedings on Halon Options Technical Working Conference-1996. National Institute of Standards and Technology, May 1996. 23 J.C. Yang, M.L. Huber, and C.I. Boyer. A Model for Calculating Alternative Agent/Nitrogen Thermodynamic Properties. In Halon Options Technical Working Conferences (HOTWC)-1995. National Institute of Standards and Technology, May 1995. 32 J.C. Yang and D.R. Keyser. Fluid Dispensing and Dispersion. In Halon Options Technical Working Conferences (HOTWC)-2006. National Institute of Standards and Technology, May 2006. 8, 31 R. Zalosh and W. Wang. Mathematical modeling of FM-200 discharge ana leakage from an enclosure. In Proccedings on Halon Options Technical Working Conference-1996. National Institute of Standards and Technology, May 1996. 25

[UE09b]

[Val98]

[VdS09] [WC96]

[Wys96]

[YHB95]

[YK06]

[ZW96]

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 Agent

Properties, 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 Henrys Law Constants........................................................................................................................................................ 11 Recovery, Reclamation, and Disposal ............................................................................................................................... 15 Recovery . ................................................................................................................................................................... 15 Reclamation................................................................................................................................................................. 15 Disposal........................................................................................................................................................................ 15

Introduction
Chlorofluorocarbons (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 physical 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 handheld 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 earths 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 bromine 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.

Uses
FM-200 is listed as an acceptable replacement for Halon 1301 and Halon 1211 in the United States Environmental Protection Agencys Significant New Alternatives Policy (SNAP) program. FM-200 is a suitable fire extinguishing agent for total flooding, portable, and local application systems. 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, telecommunication 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 Properties
Physical properties of FM-200 are given in Tables 1 to 3 and Figures 1 and 2. For complete thermodynamic properties, see DuPont Bulletin T-FM-200.

Chemical Properties of FM-200

Chemical Name Molecular Formula Molecular Weight CAS Registry Number ASHRAE Designation 1,1,1,2,3,3,3-heptafluoropropane CF3CHFCF3 170.03 431-89-0 HFC-227ea

Table 1 Physical Properties of HFC-227ea

Property Chemical name Chemical formula Molecular Wt. Boiling Point, 1 atm, C (F) Freezing Point, C (F) Critical Temperature, C (F) Critical Pressure, kPa (psia) Critical Density, kg/m3 (lb/ft3) Liquid Density @ 25C (77F), kg/m3 (lb/ft3) Density, Saturated Vapor at Boiling Point, kg/m3 (lb/ft3) Vapor Density @ 25C (77F) and 1 atm Specific Heat, Liquid (Cp) @ 25C (77F), kJ/kgC (Btu/lbF) Specific Heat, Vapor (Cp) @ 25C (77F), kJ/kgC (Btu/lbF) and 1 atm Vapor Pressure, Saturated @ 25C (77F), kPa (psia) Heat of Vaporization @ B.P ., kJ/kg (Btu/lb) Thermal Conductivity, Liquid @ 25C (77F), W/mC (Btu/hr-ftF) Thermal Conductivity, Vapor @ 1 atm, W/mC (Btu/hr-ftF) Viscosity, Liquid @ 25C (77F), cP ( lb/ft-hr) Viscosity, Vapor @ 1 atm, cP ( lb/ft-hr) Relative dielectric strength @1 atm, 25C (N2=1) Solubility of Water in HFC-227ea @ 20C (68F), ppm Ozone Depletion Potential Global Warming Potential, GWP (100 yr ITH. For CO2, GWP = 1) Atmospheric Lifetime, years TSCA Inventory Status European List of New Chemical Substances SNAP Status Inhalation Exposure Limit (AELa)
a

1,1,1,2,3,3,3-Heptafluoropropane CF3CHFCF3 170.03 16.34 (2.59) 131 (204) 101.75 (215.1) 2925.0 (424.24) 594.25 (37 .098) 1387 .7 (86.63) 8.4860 (0.52979) 7 .1461 (0.4461) 1.1816 (0.28242) 0.81327 (0.81327) 454.73 (65.9) 131.77 (56.7) 0.060491 (0.034975) 0.013336 (0.0077103) 0.23935 (0.57901) 0.011590 (0.028038) 2.00 600 0.0 (CFC-11 = 1) 3220 34.2 Reported/Included EINECS, Listed (207-079-2) Listed 1000 ppm 8 hr and 12 hr TWA

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.

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

Temperature C 15 10 5 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 Vapor Pressure (kPa) 107 .33 132.23 161.41 195.36 234.58 279.57 330.89 389.08 454.73 528.42 610.79 702.45 804.09 916.39 1040.10 1175.90 1324.70 1487 .40 1664.90 1858.30 2068.80 2298.10 2547 .90 2821.60 Liquid Density (kg/m3) 1539.7 1522.1 1504.2 1486.0 1467 .3 1448.2 1428.6 1408.4 1387 .7 1366.2 1344.0 1320.9 1296.7 1271.4 1244.8 1216.5 1186.2 1153.6 1117 .9 1078.2 1032.8 978.6 907 .8 786.8 Saturated Vapor Density (kg/m3) 8.961 10.921 13.205 15.853 18.905 22.411 26.421 30.996 36.202 42.118 48.833 56.454 65.109 74.956 86.189 99.062 113.900 131.170 151.500 175.870 205.840 244.310 298.000 397 .240 Vapor Density @ 1 atm (kg/m3) 8.4325 8.2412 8.0603 7 .8889 7 .7260 7 .5709 7 .4229 7 .2815 7 .1461 7 .0163 6.8918 6.7720 6.6568 6.5459 6.4389 6.3356 6.2359 6.1395 6.0462 5.9559 5.8684 5.7836 5.7013 5.6215

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

Temperature F 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 Vapor Pressure (psia) 17 .50 21.93 27 .18 33.35 40.55 48.88 58.45 69.38 81.79 95.80 111.54 129.15 148.77 170.55 194.65 221.26 250.55 282.77 318.18 357 .11 Liquid Density (lbm/ft3) 95.51 94.28 93.02 91.73 90.41 89.05 87 .64 86.19 84.68 83.11 81.46 79.73 77 .90 75.94 73.84 71.54 69.00 66.10 62.68 58.32 Saturated Vapor Density (lbm/ft3) 0.63 0.77 0.95 1.16 1.40 1.68 2.00 2.38 2.81 3.31 3.88 4.54 5.30 6.18 7 .22 8.45 9.93 11.76 14.10 17 .33 Vapor Density @ 1 atm (lbm/ft3) 0.5197 0.5069 0.4948 0.4834 0.4726 0.4624 0.4527 0.4434 0.4345 0.4261 0.4179 0.4102 0.4027 0.3955 0.3886 0.3820 0.3756 0.3694 0.3634 0.3576

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

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

Materials Compatibility
Because 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 facilities. 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.

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, potassium, 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 athigher temperatures. Highly reactive materials should not be brought into contact with FM-200 until a careful study is made and appropriate safety precautions are taken.

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 130F for 18 days indicates that the following metals are suitable for use with FM-200 : Aluminum 1100 Copper CDA 110 Stainless Steel 316 Carbon Steel 1020 Nickel 200 Inconel 600 Lead Yellow Brass Aluminum 2024 Cast Iron, grey Stainless Steel 304 Silver 999+ fine

Compatibility with Elastomers

Compatibility tests were performed with several commonly used elastomers. Elastomer coupons were50% immersed in liquid FM-200 for two weeks at room temperature. Additional tests were conducted per ASTM D471 at 100C (212F). All of the elastomers tested exhibited minimal swell with the exception of urethane and Viton A. Results are summarized in Table 4.

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

Table 4 Elastomer Compatibility

Elastomer Linear swell, % Weight Gain, % 0.37 1.44 0.66 1.86 1.41 26.83 0.08 0.06 5.71 Hardness Change, units 0 1.6 0 4.0 2.4 44.0 5.5 6.9 4.6

Exposure at Room temperature, 23C (72F) for 14 days Butyl 0 Nordel EPDM 0.20 Neoprene W 0.05 NBR 0 Hypalon CSM 0.19 Viton A 9.49 Epichlorohydrin homopolymer 0.15 FA polysulfide 0.05 Hytrel TPE 1.33 Exposure per ASTM D471 at 100C (212F) Buna N -3.1 Butyl 3.6 EPDM 1.0 Hypalon -2.0 NR Rubber 1.7 Neoprene G 0.8 Neoprene W -3.6 SBR -1.2 Silicone 2.8 Urethane >10 Viton A 8.4

Table 5 Plastic Compatibility

Plastic High-density polyethylene (HDPE) Polystyrene (PS) Polypropylene (PP) Acrylonitrile-butadiene-styrene (ABS) Polycarbonate (PC) Nylon 6/6 Polytetrafluoroethylene (PTFE) Polyimide (PI) Polyethylene terephthalate (PET) Polybutylene terephthalate (PBT) Acetyl Polyvinyl chloride (PVC) Polyphenylene oxide (PPO) Polyphenylene sulfide (PPS) Weight Gain, % 0.11 0.03 0.06 0.03 0.10 0.17 5.23 -0.11 -0.04 -0.06 -0.04 -0.06 -0.05 -0.38 Surface Condition No Change No Change No Change No Change No Change No Change No Change No Change No Change No Change No Change No Change No Change No Change

Compatibility with Plastics

Compatibility tests were also performed with severalcommonly used plastics. Results are summarized in Table 5.

Cardiac Sensitization
If vapors are inhaled at a concentration of 105,000 ppm and higher, the heart may become sensitized toadrenaline, 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 oremotional stress. Because of possible disturbances of cardiac rhythm, catecholamine drugs, such as epinephrine, should beconsidered 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%) asdetermined in epinephrine-challenged dogs.

Safety
Users 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 document 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) forFM-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, confusion, loss of coordination, and even loss of consciousness. Higher exposures to the vapors may cause temporary alteration of the hearts 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 breathing, give artificial respiration. If breathing is difficult, give oxygen. Seek medical attention.

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 exposed 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.

Spills or Leaks
If a large release of vapors occurs, such as from alarge 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 asupplied 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 necessary 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.

The 68 kg (150-lb) size cylinder of FM-200 is a freestanding, upright returnable cylinder, equipped with a nonrefillable 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 cylinder, is attached to the valves 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 freestanding, upright returnable cylinder equipped with a forklift lifting 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 CGA660 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 meters) 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 container; 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). Liquid 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 trailers. These tank trailers are equipped with 1-1/4 mail EverTite (vapor) and 2 male Ever-Tite valves (liquid). Required power is either 240 volt or 440 volt.

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 Handling

Shipping Information for the United States
FM-200 is a liquefied compressed gas. According to the U.S. Department of Transportation (DOT), a liquefied compressed gas is a gas, which when packaged under pressure is partially liquid at temperatures above -50 C (49CFR 173.115). The appropriate DOT designation for FM-200 is as follows: Proper shipping name: Hazards class: UN No.: DOT/IMO Labels: Heptafluoropropane 2.2 3296 Nonflammable Liquefied Compressed Gas

Bulk Storage Systems

DuPont sells bulk storage systems to itsFM-200 customers. The systems are prefabricated, tested, and ready to install onsite. The units are designed tooptimize 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, asan integrated unit. All systems are equipped with the DuPont Fluorochemical Emissions Elimination Delivery (FEED) System to prevent emissions during deliveries and with dual pumps to provide aninstalled spare. The units are skid-mounted andrequire only placement on a concrete pad andconnection to electrical and process systems.

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

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.

Use personal protective equipment, such as sideshield glasses, gloves, and safety shoes, when handling containers. Avoid skin contact with liquid FM-200 ; it cancause frostbite. Never heat a container to a temperature higher than 52C (125F). Never refill returnable cylinders without DuPont consent. DOT regulations forbid transportation ofreturnable cylinders refilled without DuPonts authorization. Never use a magnet or sling (rope or chain) to lift containers. Lifting may be accomplished by the use of a safe cradle or platform basket that holds the container. Never use containers as rollers, supports, or for anyother purpose than to contain FM-200 . Protect containers from any objects that will resultin a cut or other abrasion in the surface ofthe metal. Never tamper with the safety devices in the valves or container. Never attempt to repair or alter containers or valves. Never force connections that do not fit. Make surethe threads on the regulator or other auxiliary equipment are the same as those on the valve outlets. Keep valves tightly closed, with valve caps and hoods in place when the container is not in use. When storing containers outside, store under a roof and protect from weather extremes. Use a vapor recovery system to collect FM-200 vapors from lines after unloading.

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 . Contact 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 thereceiving container. This method chills the FM-200 as it passes through the transfer line, reducing the pressure in the receiver to induce 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 devicesare readily available through a refrigeration supply house. A detailed discussion of leak detection, along withalist 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:

Table 6 Specifications for FM-200 Shipping Containers

Type Dimensions 68 kg (150 lb) 544 kg (1,200 lb) ISO Container 12 x 46 10 x 56 30 x 53 30 x 56 8 x 8 x 20 (frame DOT Specification 4BW240 4BW400 4BW260 4BW240 51 Net Weight, lb FM-200 150 150 1,200 1,200 37 ,000 37 ,000 Net Weight, kg FM-200 68 68 544 544 16,784 16,784

Tank Trailer 5,000 gallon MC-330 or -331

Figure 3. 68 kg (150 lb) Size Cylinder

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

Figure 4. Liquid/Vapor Valve

Figure 6. Typical ISO Tank

Figure 7. Typical Bulk Storage System

10

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 FM200 . 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 superpressurization of FM-200 at various fill densities, it is necessary to understand the solubility relationship of nitrogen and FM-200 . Extensive experimental work was conducted

by DuPonts Central Research and Development group to develop this information. The Peng-Robinson Equation of State (PREOS) was then used to calculate the following: Weight of nitrogen required for superpressurization Isometric diagrams Henrys Law Constants 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.

Henrys Law Constants

PREOS was also used to calculate the Henrys Law Constants as shown in Figure 12 (English units) andFigure 13 (SI units).

Table 7 Weight of Nitrogen Required for Superpressurization of DuPont FM-200 (English Units)
Weight of nitrogen per lb of FM-200 at 70F Fill Density lb/ft3 360 psig oz 0.534 0.463 0.407 0.361 0.323 0.291 0.263 0.239 600 psig oz 0.944 0.819 0.718 0.636 0.568 0.511 0.461 0.418

40 45 50 55 60 65 70 75

Table 8 Weight of Nitrogen Required for Superpressurization of DuPont FM-200 (SI Units)
Weight of nitrogen per kg of FM-200 at 21C Fill Density kg/m3 2500 kPa (gauge) g 36.4 30.3 25.7 22.2 19.3 17 .0 15.1 4150 kPa (gauge) g 63.8 53.1 45.1 38.8 33.8 29.7 26.3

600 700 800 900 1000 1100 1200

11

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

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

12

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

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

13

Figure 12. Henrys Law Constant for Nitrogen Solubility in DuPont FM-200 (English Units)

Figure 13. Henrys Law Constant for Nitrogen Solubility in DuPont FM-200 (SI Units)

14

Recovery, Reclamation, and Disposal

Responsible use of FM-200 requires that the product be recovered for reuse or disposal whenever possible.

Reclamation
Reclamation refers to the reprocessing of FM-200 recovered 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 DuPonts reclamation program. Contact DuPont for further information.

Recovery
Recovery refers to the removal of FM-200 from equipment and collection in an appropriate container. Recovery does not involve processing oranalytical testing. Recovery is normally performed when a system must undergo maintenance and the FM-200 is then returned to the system after completion. Thereare a number of recovery devices on the market. These devices contain a compressor and anair-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 .

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 .

15

For further information regarding DuPont Fire Extinguishing Agents, contact: www.cleanagents.dupont.com
Americas
DuPont Fluoroproducts Chestnut Run Plaza 702-1274E P.O. Box 80702 Wilmington, DE 19880 Tel: (800) 473-7790

Asia
DuPont Taiwan, Limited 13F, Hung Kuo Building 167 Tun Hwa North Road Taipei, Taiwan 105 ROC Tel: 886-2-25144488

Europe, Middle East and Africa

DuPont de Nemours International S.A. 2, Chemin du Pavillon CH-1218 Le Grand-Saconnex Geneva, Switzerland Tel: 41-22-717-5296

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 ANYMEANS ELECTRONIC, MECHANICAL, PHOTOCOPYING, RECORDING OR OTHERWISE WITHOUT PERMISSION OFDUPONT. 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 ourcontrol,wemakeno warranties, express or implied, and assume no liability in connection with any use of this information. As with any material, evaluationof any compound under end-use conditions prior to specification is essential. Nothing herein is to be taken as a license to operate underorarecommendation to infringe any patents.

K23261 (09/09)