US2928529A - Storage of explosive gases - Google Patents

Description

March 15, 1960 A. v. GROSSE ET AL STORAGE OF EXPLOSIVE GASES 3 Sheets-Sheet 1 Filed Feb. 15, 1958 bins-.3

S 6 Y mm m N 0 R E O v RN Mr 0 0 mm& wi l Q W MM K M 0 M a M 6 w 41 N w w 0 4 m 7, L m w M M am w N I 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 W 5 4 3 E 2 U 4 a 2 I,

March 15, 1960 Filed Feb. 13, 1958 QUE/VLl-l/NG 0/40/5227? (d) FOR SPHERES, M/M/ORO/VS 5 Sheets-Sheet 2 & a

600 mp0,? I

PRESSURE 0F 4/0010 0 INVENTORS z a 4 5 6 7 a .9 r0

PRESSURE M ATMOSPHERE? AR/srm M GROSSE w 445x 6. STREAM;

A ORNEY6 March 15, 1960 Filed Feb. 13, 1958 A. V. GROSSE ET AL 3 Sheets-Sheet 3 CUA/DEMSAT/ON POINT 7MPERA TURE, "K.

fire/n0, MMgM ATTORNEYS United States patefifo 2,928,529 STORAGE F EXPLOSIVE GASES Aristid V. Grosse, Haverford, and Alex G. Streng, Philadelphia, P25, assignors to Research Institute of Temple University, Philadelphia, Pa., a non-profit corporation of Pennsylvania 7 Application February 13, 1958, Serial No. 715,098 13 Claims. (Cl. 206-.6

This invention relates to a processfor storing gases either in pure form or in admixture with other gases in concentrations normally subject to induced rapid or explosive decomposition, under pressures of one to several atmospheres. Certain gases, such as ozone, hydrogen azide, chlorine dioxide and others are endothermic compounds and are dangerous to store except in quite dilute mixtures with inert gases. In high concentrations or pure form, they are subject to possible initiation of explosion or detonation in a storage vessel either by spark, mechanical shock, sudden heat, etc., destroying not only the gas but also the vessel containing them. The presence of contaminants such as organic matter in the vessel may also initiate flame or explosion. To illustrate how easily the gases may be ignited, the oil from a thumb print is sufficient to initiate explosion of pure ozone. It is clear that suchgases cannot be stored or shipped in cylinders as can be done with most other gases, e.g. oxygen, nitrogen, methane, etc.

It is an object of the present invention to provide a method by which gases of the character described can be stored in cylinders under pressures of from one to several atmospheres for long periods of time and in concentrations which are normally dangerous;

In a broad embodiment the invention comprises a process for storing ozone, hydrogen azide, chlorine dioxide and the like in concentrations normallysubject to induced rapid decomposition or explosion which comprises confining said gas or gas mixture in a storage zone filled with a material substantially inert to said gas, said material being of a size and physical form to provide alarge number and aggregate volume of free voids, the diameters of said voids not being greater than the quenching distance of said gas.

g If an explosive gas or explosive mixture of gases is subjected to conditions which will initiate a flame, an explosion can develop in the body of the gas if the flame can be maintained. If the flame can be quenched in a short distance no combustion occurs and therefore no explosion can take place. The distance in which the flame must be quenched to prevent explosion is known as the quenching distance or quenching diameter. This varies with the particular gas or gas mixture and the temperature and pressure conditions. The quenching distances ,of a number of combustible or explosive gas mixtures have been reported in the book Combustion, Flames and Explosions by Lewis & Elbe, Academic Press Inc., New York city, 1951, pages 408-412. These are reported as being from 1.6 cm. to 0.02 cm depending on the gas or gas mixture, at one atmosphere pressure and at 20 C. It has been reported also that the quenching distance for ozone is below about 25 microns (0.0025 cm.). I

The quenching distance is dependent upon the pressure, the gas concentration and the storage temperature. The higher the pressure and/or temperature of a gas of a give'n concentration, the smaller the quenching distance. For a gas" Iik'e ozone mixed with'varying percentages of atmosphere pressure. The upper curve sh ows lthe cone:

t 2,928,529 Pat nted Mar- '1 2 oxygen, the quenching distance increases as the volume percent of ozone in the mixture decreases. We have found that the quenching distance for ozone is considerably greater than heretofore believed. H Figure 1 illustrates how the quenching diameter in spherical voids varies with concentration and temperature for various mixtures of ozone and oxygen. I V Figure 2 shows the quenching diameters of spherical voids at various pressures when pure ozone gasis stored at 195 K. (78 C.) at various pressures below its condensation point. H r V p Figure 3 shows theeifect of temperature on the quench ing diameter of spherical voidsfor pure ozone at PIGS sures from one atmosphere to about 8 atmospheres. W In one set of experiments we completely filled glass cylinders with hollow and somewhat porous spheresof substantially pure fused alpha alumina. The diameter of the spheres was in the range of 250 to 300 microns. To one end of each of the cylinders were sealedl-2 capillaries equipped with stopcocks. The cylinders were evacuated and substantially pure ozone gas was introduced. Internal pressures of from one to three atmospheres at 78 C. were investigated. Afterfilling, the stopcocks were closed. Small pieces of rubber were inserted in the capillary on the outlet side and they were then evacuated and sealed. When the stopcocks were opened the ozone filling the evacuated capillary ignited when it contacted the rubber. The flame travelled dowg the capillary and was immediately quenched when it reached the alumina spheres. The ozone in the packed inder remained unchanged. It was condensedin liquid oxygen and the amount of oxygen produced by decomposition of the ozones was determined as being less than 0.1% by weight. When this experiment was carried out in a similar cylinder except that it contained no packing material, the ozone exploded, shattering the apparatus The size of the greatest free space in the storage zone should be less than the quenching distance of ozone. will permit its being stored for long periods of time cylinders, at pressures ranging up to at least three mosplieres or higher at 78 C. Of course the press e for 100% ozone must be below about 6.8 atmosphe 's since ozone liquefies at about that pressure and 5 tend.- perature of -78 C. V l p 7 When using perfectly spherical particles of the range used in the experiments the diameter of the largest free space in a closely packed vessel may be calculated as follows:

Where d is the diameter of the void and Dis the dial" V of the particles. Although the particles used were not perfect spheres they were nearly enough so that the cal: culation is valid for these purposes. The size of the particles which may be used under varying conditions of temperature and pressure a it using various concentrations of ozone from pure to mixtures with an inert gas, such as oxygen, may determined by reference to the a-ttached figures. T figures represent the quenching distance or quenc I diameter for ozone in a spherical void. I Referring to Figure l which is plotted on a logarithmic scale, the two curves illustrate how the quenching diam v eter changes as the molar concentration of ozone Chan s from 100% ozone to mixtures thereof with oxygen. The lower curve shows the quenching diameters of pure ozone and mixtures in various concentrations with ox at atmospherictemperature (298 K. or 25 C.

spending values 195 K. (78 0.), also'at one atmosphere pressure. Thus, it will be observed that the quenching diameter for pure ozone at 298 K. is about 146 microns and at 195 K. is about 330 microns. By substituting these values for (d) in the above equation,

the maximum diameters of uniform spherical particles which will prevent flame propagation under each of these condtiions; can be determined. These calculate out to be 650 microns and about 1.5 mm.;respectively. Therefore, spherical particles having an average diameter of less than 650 mu will quench an induced flamein a vessel with which they are packed and containing pure oiz'oneat 298" K. It follows that such particles will quench a flame in gaseous ozone stored at any lower temperature and/or concentration. The maximum permissible average particle sizes for gaseous mixtures of ozone andoxygen of any other concentration can be calculated in like manner.

I Figure 2 is aplot of quenching diameter versus pressure for 100% ozone at 195 K. for ozone in spherical voids, It will be observed that if the gas is to be stored at this temperature it must be at less than about 6.8 at

mospheressince that is the pressure at whichpure ozone liquefies at thistemperature. that liquid ozone can be caused to explode regardless of whether packing material is or is not present. Consequently the temperature and pressure and/or concentration should be regulated so that condensation does not occur, otherwise the possibility of an accidental explosion "of liquidozone in the vessel is present. The quenching diameter for pure ozone gas at any selected pressure at which the gas is to be stored is readily determined. Then the permissible particle sizes for safe 'storage at this low temperature can be calculated as described above. l

This curve illustrates the fact that the quenching diameter at a given temperature for pure ozone decreases as the pressure is increased to any point at which the ozone "remains in the gaseous state. A similar relationship exists *at other temperatures. The family of. curves in Figure 3 graphically depict the efiect on quenchingdiameter of pure ozone at various r temperaturesand pressures. These curves show that as "either temperature or pressure or both is increased, the quenching diameter becomes smaller. It is apparent that the maximum permissible average size of the packing material will follow a similar trend. By comparison with Figure 1, it will be seen that a similar eflect is to be It should be'pointed out expectedwhen gas mixtures of lesser ozone concentrationare stored. However, since the quenching diameterincreases with diminishing ozone gas concentration in such mixtures, larger diameter spheres can be used at any given pressure and concentration. Another way of putting it is that a cylinder packed with particles of a diameter suitable for safe storage of pure ozone gas under desired temperature and pressure conditions, can be used to safelystore a mixture of ozone and oxygen or other inert gas, at a considerably higher pressure.

,Under certain circumstances where the possibility of accidentally igniting the explosivegas is remote, spherical particles which will provide void diameters of say about 2 to 3 times the sizes indicated for entirely safe storage, maybe used. Under these circumstances flame propagation is not prevented if the gas should be ignited. The acceleration of the decomposition to the point .of explosive violence is, however prevented, Thus, if the gas should become ignited, a flame front will pass through the bed but no explosion will result. If the gas is ozone, all of the ozone in the cylinder will be converted to oxygen. There would, however, be no damage :erated to raise the temperature too high. We refer to .of pure ozone at two atmospheres pressure and 300 K.

The quenching distanceunder these conditions is about 62 microns. Thus the spheres might suitably be 250- 275 micron average diameter. The container, thus packed, can be filled to about 5 atmospheres pressure and still be in the semi-safe area. Ifignited the ozone will be decomposed as described. To be entirely safe at 5 atmospheres pressure, as will be seen from Figure 3, the voids should have a diameter of less than about 25 microns corresponding to spheres of less than about 110 mu. Thus, the amount of available ozoneinthe cylinder is substantially increased because of the higher storage pressure, without thedanger of catastropic' explosion. r

Another beneficial factor is that the aggregate surface area of the particles with which-the gas comes in contact is less for the/larger particles than it is for the smaller ones. Since almost any solid may, in some degree, tend to catalyze the 'slow decomposition of a gas such' as ozone, the decreased surface area as a result of the use of the larger particles reduces any such catalytic eflect that may exist.

The use of particles providing void spaces somewhat greater than the safe quenching diameter might be, for example, under circumstances suchas a laboratory or plant wherein the handling of the gas and cylinders is well controlled and where any hazard resulting from the heating of the cylinder due to the decomposition of the gas offers no problem, even if it accidentally became ignited. I r

It was found that pure ozone contained in the cylinder packed with alumina spheres as previously described shouldbe stored at low temperatures. The rate of spontaneous decomposition of pure ozoneor mixtures thereof with oxygenincreases at higher temperatures. Thus, under approximately'one atmosphere pressure and 78 C. pure gaseous'ozone (in the absence of the packing material) showed no noticeable decomposition after eight days storage. At 23 0. pure ozone showed decomposition of approximately 0.9% per day. Mixtures with oxygen containing about 50 mol percent of ozone showed about'0.9% of decomposition per day when stored at temperatures of about 0 C. and about 24 C. A mixture containing 40 mol percent of ozone stored at 26 C. showed 4.2% decomposition in seven days, or a rate of decomposition of about 0.6% per day.. Even thoughsubstantially pure, the alumina particles have some tendency to increase the rate of decomposition. Indications are that at 78 C. the rate is about 0.01% per day in the presence of the alumina used as contrasted with no appreciable decomposition of ozone to oxygen in a comparable period when no packing was present. The rate of decomposition is so low, however,

that storage over long periods of time 'is practical. The size of the particles also influences the rateof decomposition. Since this appears to be a surface effect, it is desir- "able to have the minimum surface possible, i.e.. to use particles of the largest permissible'diameter consistent if an explosion were to take place in a container not filled with the spheres.

The only damage sustained would be the loss of whatever ozone was in the cylinder or possibly to cylinder fittings if enough heat is gen with the desired quenching efiect. This is also desirable since with comparable wallthickness of hollow spheres, the space'occupied by the solids is less, thus providing a larger useful storage volume.

It is, therefore, desirable to store the gas in accordease with invention at the lowest feasible temper- S stares in vessels with packing material of largest permissible diameter, where it is to be kept in storage for any substantial length of time. In any event, the storage temperature must be below that at which the gas will spontaneously decompose. As a rule the storage temperature is not above ambient room temperature (25 C.30 C.). In the case of ozone, storage for any extended period will be below C. and preferably 50 C. to -80 C. The storage containers must be completely and uniformly filled with packing material. They must not have any free space larger than the voids formed by the particles of material used. his desirable that the packing material be of uniform size. i.e. having mesh sizes within a narrow range so as to provide the maximum free void space.

Experimental evidence has indicated that mixtures of oxygen and ozone containing up to about 16 mol percent of the latter do not produce a flame at 23 C. and one atmosphere pressure when sparked in a 0.95 cm. inside diameter tube, one meter in length. When the mixtures contain from about 17 to about 25 mol percent of ozone, there is slow uniform burning. Mixtures containing more than 25 mol percent ozone burn more rapidly and less uniformly as the concentration increases, first producing noisy burning and burning with an explosive knock which destroys the tube at concentrations above about 60% ozone.

Thus, it might be reasonably safe to store mixtures of oxygen and ozone containing not more than about 16 mol percent of the latter in ordinary cylinders at one atmosphere pressure and at low temperatures. However, at higher pressures the danger of burning upon ignition likewise increases. Mixtures containing more than about 15% to 16% of ozone at atmospheric pressure or at increased pressures are likely to explode or at the very least to burn completely if they are ignited. Hence such mixtures cannot safely be stored in ordinary cylinders.

The packing material, as previously mentioned, should be of a particle size such that the greatest distance between the particles is smaller than the quenching distance of the pure ozone or ozone mixture at the respective storage temperature and pressure. This is true of other explosive gases. The quenching distance of such gaseseither in pure form or in admixture with inert gases varies with the gas, or mixture of gases, under consideration. The quenching distance for any gas or mixture can be determined by attaching capillary tubes of known diameter between sections of tubes sufliciently large to permit fiame propagation. The tubes should not be large enough to hold dangerously large volumes of gas. The tubes are filled with the gas and are ignited at one end. If the flame passes through the capillary, igniting the gas in the remote tube, the diameter of the capillary exceeds the quenching distance of the gas. By using a series of capillaries of known graded diameter the quenching distance can be determined quite accurately. The tubes should be shieidedto prevent damage in the event of an explosion.

The alumina s heres used in the present experiments were hollow and contained relatively large pores in the walls. The average diameter of these porous spheres are about 250 to about 300 microns. The apparent density was about 1.2 grams per cubic cm. Since the density of pure alumina is 4.0 grams per cu. cm., the free space was about'70%, the remainder, or 30%, being that occupied by the alumina.

In cubic or hexa onal close packing of spheres, the free volume between spheres (assumed to be perfect spheres) from geometric consideration is Thus, if the total free space is 70%, 26.2% is due to the interspherical free space. The ditference, i.e.

70% 26.2% =43.8% is due to the pores and voids in the spheres.

The quenching material em loyed in this invention may be in the form of hollow spheres, microballoons', porous spheres or bubbles either discrete or cemented together. They may comprise foams of catalytically irractive material such as glass or quartz foam. The. ideal foam consists of pentagonal dodecahedrons. Cracks or pinholes should be provided in the walls or else they should be so thin that the stored gas can enter and leave by way of diffusion. By the same token, where hollow spheres are used, there should be cracks or pinholes to ermit entrance and egress of the gas, or the walls should be sufiiciently thin to permit diffusion through them. Solid particles, either spherical or otherwise, of substantially catalytically inert material could be employed provided they pack in such a way that the interstitial spaces are of less than the quenching distance of the gas being stored. As a rule, these are less desirable than hollow spheres, etc. since the total free space available for storage of the gas is reduced and the weight of the packed cylinder is increased to an undesirable extent. By using hollow spheres and the like, the capacity of the storage cylinder or other vessel is increased to the extent of the volume of the available space within the spheres, and the weight of the packing material is reduced. As is noted, the available space for gas occupancy in the hollows and pores of the alumina spheres employed in the experiments comprised almost 63% of the total available space within the packed storage vessel. Hence, from a practical standpoint, materials of this kind will be used for most purposes.

For the storage of certain gases, bubbles, spheres or foams or" organic polymers or other solid resins can be used where there is no appreciable reaction between the organic material and the gas being stored, or where they do not trigger flame or explosion, or catalyze a more gradual decomposition. Thus, hydrogen azide can bestored in vessels filled with bubbles, spheres or foams made of polystyrene or with materials such as the normally soiid polymers of tetrafiuoroethylene, chlorotrifiuoroethylene and the like. The latter two can be used also with ozone, although these are less desirable than the alumina, quartz, or glass spheres, because they may increase the decomposition rate of ozone by virtue of a catalytic effect. Most other organic materials react too rapidly with ozone, either causing relatively rapid decomposition or even resulting in ignition or detonation.

Chlorinedioxide can be stored with particles of the polyhalocarbon resins of the type discussed, but tend to oxidize or react with most other synthetic polymers or resins or to decompose more rapidly in their presence. Even in this instance, the inorganic monocatalytic materials discussed are preferred, since chlorine dioxide can only be stored as a gas at temperatures above its boilmg point 11 C. at atmospheric pressure). Spontaneous decomposition becomes more rapid as the temperature increases, and if this is enhanced by reaction with or the catalytic eifect of the packing material, storage in vessels packed with such materials is not practical.

As previously mentioned, the packing material should be as nearly noncatalytic as possible with respect to reaction with or decomposition of the gas being stored. The alumina spheres employed were substantially pure alumina of the alpha crystalline structure, containing only a few parts per million of other impurities. Being a fused material, the pores were comparatively large so that the total surface area with which the gas was in contact was relatively small. Therefore, even though traces of catalytic materials were present and even though the alumina itself might have some catalytic effect, the surface area was so small that the eifect'on decomposition of the ozone was negligible at 78 C. On the other hand fused alumina containing oxides of metals such as iron, copper, etc. cause ignition of ozone and hence could not be used in this invention, even though the particle size were within proper limits. As previous- 1y mentioned, it is desirable to use particles of a size in the'larger end of the range suitable for storage under theselected temperature .and pressure, since as the par.- ticle sizes are decreased the surface area increases as the inverse square of the diameter. Practically all solids may have at least some catalytic effect, hence the resultant decomposition will increase as a function of increased surface area in contact with the gas. Some reasonable margin of safety should be allowed to compensate for possible contingencies such as an increase in temperature due to failure of refrigeration equipment. Since there is a considerable semi-safe latitude of particle size, this is not too critical except where absolute safety is required, as would probably be the case when the packed cylinders are transported by common carrier.

Thus, it is desirable that the packing material present as low a surface area as possible consistent with a high void space; Precipitated alumina or precipitated silica gel such as that used as a carrier for catalysts, regardless of particle shape, have tremendously largeinternal surfaces. Thismay increase catalytic decomposition to such an extent as to make them unsuitable for use as a packing material. This is particularly true if traces of iron, copper, cobalt or their oxides, etc., or other impurities having pronounced catalytic effect on the decomposition of the stored gas are present. Depending on Eheir nature and concentration, they may even induce ame.

Containers may be of any shape but are conveniently of a cylindrical form similar to those employed in the storage of gases such asoxygen, hydrogen, nitrogen, etc. Since pressures are comparatively low, generally not above about 100-125 p.s.i.g., and frequently of the order of one to five atmospheres, they can be made of relatively light weight materials such as aluminum, although stainless s'teel and the like of relatively thin gage can also be used. .The interior of cylinders may be coated with glass or with resins such as Tefion or Kel-F Packing glands, gaskets, etc. for valves, flanges, etc. may be of Teflon or Kel-F. Although some catalytic action might be expected from certain materials that may be used as a component of the container, the surface area with which the gas is in contact is relatively small and the overall catalytic elfect is generally negligible, particularly at low storage temperatures.

We have described my invention more or less in terms of the storage of ozone but it is not intended to be limited to ozone entirely. All gases which may be stored in accordance with my invention are not to be considered equivalents since, while they possess the common property of being dangerous when stored in comparatively high concentrations under conditions other than in this invention, they possess properties which require some modification with respect to particle size of the packing material, dimensions of the interspherical space and also variations in the type of packing material with which they may be employed. They difier in quenching distances and the temperatures and pressures at which they can be stored as gases.

The term inert as used herein, whether in connection with gases or solids, means that the material is sub stantially non-reactivewith the endothermic gas, and has a negligible catalytic effect on the decomposition of said gas under storageconditions of temperature and pressure.

We claim as our invention:

1. A process for storing an endothermic gas in concentrations normally subject to induced explosive de-l composition which comprises confi ingsaid gas; in a storage zone completely filled with hollow bodies of solid material: substantially nonreactive with the gasand substantially. catalytically inactive withrespect to decomposition of said gas under conditions of storage temperature and pressure, said hollow bodies providing, a large number and combined volume of free voids filled with said gas,q.the diameters of said voids being no greater than the quenching distance of said gas at the storage pressure]. and temperature. T

2. .The process of claim 1 wherein said solid body com prises hollow generally spherical particles.

3. The process of claim lwherein the body comprises hollow generally spherical particles of substantially pure fused alumina.

4; The process of claim '1 wherein the body is in formiof a solidified foam.

5. The method of claim 1 wherein the gas is ozone.

6. The method of storing an endothermic gas in concentrations normally subject to artificially induced explosive decomposition which comprises confining said gas in a storage zone completely filled with hollow particles of a solid material substantiallynonreactive with and catalytically inert with respect to the decomposition of said gas at the storage temperature and pressure, said particles being of a physical form to provide a large number and combined volume of free voids filled with said gas, the average diameter, D, of said particles being i greater than n where dis the quenching diameter of the gasat the storage conditions of the temperature and pressure, and n is a numerical factor not greater than about 3.

r 7. The method of claim 6 wherein n is greater than 1, whereby the gas in said zone, if ignited, will burn with less than explosive violence.

8. The method of claim 6 wherein-n is not substantially greater than 1.

9. The method of claim 6 wherein the packing material comprises hollow generally spherical particles of substantially pure fused alumina.

10. The method of claim 6 wherein the gas is ozone.

11. As an article of manufacture, a pressureand gas tight container having means for admitting and removing gas, said container being completely filled with hollow particles of a solid material and an endothermic gas in a concentration normally subject to artificially induced explosive decomposition, said hollow particles .being of a physical form and sizeto provide a large volume of free voids, the average diameters of which are no greater than the quenching diameter of said gas, said particles being substantially nonreactive with and catalytically inactive with respect to inducing decomposition of said gas at storage temperaturesand pressures.

12. The article of claim 11 wherein the gas comprises ozone.

13; The article of claim 11 wherein the packing material comprises spheres of hollow substantially pure fused alumina.

References Cited in the file of this patent UNITED STATES PATENTS 1,463,498 Burgess July 31, 1923 1,608,155 Barnebey Nov. 23, 1926 2,356,334 Maude et al Aug. 22, 1944 FOREIGN PATENTS 53,502 Switzerland Sept. 23, 1910.

the

Claims (1)

11. AS AN ARTICLE OF MANUFACTURE, A PRESSURE AND GAS TIGHT CONTAINER HAVING MEANS FOR ADMITTING AND REMOVING GAS, SAID CONTAINER BEING COMPLETELY FILLED WITH HOLLOW PARTICLES OF A SOLID MATERIAL AND AN ENDOTHERMIC GAS IN A CONCENTRATION NORMALLY SUBJECT TO ARTIFICIALLY INDUCED EXPLOSIVE DECOMPOSITION, SAID HOLLOW PARTICLES BEING OF A PHYSICAL FORM AND SIZE TO PROVIDE A LARGE VOLUME OF FREE VOIDS, THE AVERAGE DIAMETERS OF WHICH ARE NO GREATER THAN THE QUENCHING DIAMETER OF SAID GAS, SAID PARTICLES BEIN SUBSTANTIALLY NONREACTIVE WITH AND CATALYTICALLY INACTIVE WITH RESPECT TO INDUCING DECOMPOSITION OF SAID GAS STORAGE TEMPERATURES AND PRESSURES.

Images