Airship- Failures-DisastersPart 2 BANNER

Fire and Explosion Risks from Lift Gases and Fuels

It’s at last time to confront the eight-hundred pound gorilla in the airship design classroom: hydrogen fires and explosions.

Hydrogen, the lift gas that provides the most lift, is also the most dangerous. The following table provides lower (LFL) and upper (UFL) flammability limits at room temperature and standard pressure for mixtures of various gases with air. The mixture is flammable when the gas content is within those limits. The fuels in the table might be burnt, not only to power the engines, but also to heat air in the case of a hot air (thermal) airship or to boil water if the airship uses steam lift. For the lift gases, I also give the buoyant lift provided; this assumes the gas (except air) is pure, and behaves as an ideal gas at NTP (normal temperature and pressure, 15oC and 1 atmosphere pressure).


LIFT GAS LFL% UFL% Mass Lifted lbm/1000 cf
Hydrogen 4 75 71
Helium NA NA 66
Methane 5 15 34
Ammonia` 15 28 32
Steam 100oC NA NA 40
Hot Air 110oC NA NA 19
Propane 2.1 10.1 NA
Isobutane 1.8 9.6 NA
Gasoline 1.4 7.6 NA
Kerosene 0.6 5.0 NA
Diesel 0.6 7.5 NA

(Wikipedia/Flammability Limit, etc.)


At least for hydrogen, the flammability limits widen with increased temperature, although the UFL is affected more than the LFL (College of the Desert, Fig. 1-6). For hydrogen, the concentrations that support an explosion include 18-59% (Edeskufy 102).

Fires and explosions have occurred as a result of ignition of fuels carried by an airship, hence their inclusion in the table. A Goodyear ZPG-3W, the Vigilance, suffered an explosion at 300 feet. (See ). There have also been incidents involving propane on hot air balloons. Crashworthy fuel tanks were not developed until a decade after the US Navy decommissioned its last nonrigid airship (Stahl).

The Royal Anne, the Testbed and, the Czarina Evdokia use steam propulsion, and of course there are safety considerations in handling steam that have already been discussed by Evans, “Steam: Taming the Demon” (Grantville Gazette 11). Those concerns are exacerbated when the steam can come into contact with gas bags holding flammable lift gases such as hydrogen, especially since rubber (a common bag material) is surprisingly permeable to steam.


Hydrogen Fires

The perils of hydrogen became apparent even in the balloon era. In 1785, Rozier constructed a hybrid balloon with separate compartments for hydrogen and hot air. The fabric caught fire near the gas valve. (Payne 12).

The Vaniman Akron (not to be confused with the later USN rigid) was lost in 1912 to a hydrogen fire (Payne 50). In September 1910, the LZ6 caught fire and burned in the hangar due to a mechanic’s error–he used gasoline to clean the gondola (Payne 71). And the LZ zeppelin statistics provided in part 1 show that hydrogen fires were a significant cause of airship loss.

For a hydrogen fire to occur, you need two things in contact: a flammable mixture and a source of ignition.

A freshly filled gas cell should contain hydrogen of sufficient purity to be above the UFL. For the typical purity of hydrogen produced by various processes, see Cooper, “Hydrogen: The Gas of Levity” (Grantville Gazette 38). While a freshly filled gas cell might be 95% hydrogen and 5% air, over time air will diffuse in (reducing the purity of the cell) and hydrogen will diffuse out into the interstitial space between the gas cells and the aerodynamic envelope (in the case of a compound aerostat) or into the ambient air (in the case of a simple aerostat).  The British R38 (2.7 million cf) leaked 1 million cf of hydrogen each month. (Robinson 37, 43).

The Germans called this process “rotting” and when the hydrogen content got too close to the UFL, they would vent and refill the cell with fresh gas. Venting and refilling of course has its own potential for creating flammable mixtures.

Violent maneuvers, possibly the result of storm conditions or lift mismanagement, can result in a gas cell being torn, and thus a much more rapid diffusion. However, in a compound aerostat, the hydrogen would be retained by the outer envelope for a time.

In flight, hydrogen gas might be vented, either deliberately to stop an ascent, or as a result of a stuck valve. The hydrogen would be carried out of the cell by a flexible conduit to the surface of the envelope, where it would escape into the atmosphere. It would rise, and, if the airship were moving forward, be left behind.

Where that gas is vented makes a difference. On some airships, such as the LZ18 (German Navy L2), the vents were on the bottom half of the airship, and the engine car windshields (a new feature) somehow created a suction that drew the hydrogen into the car. (Payne 77). Hydrogen is lighter than air and thus, when vented, will rise upward.


Hydrogen-air mixtures self-ignite at temperatures of 793-1023K (Edeskufy 102). The minimum spark energy needed for igniting hydrogen in air is just 0.017 mJ at earth surface pressure; compare that to 0.29 mJ for methane and 0.24 mJ for gasoline (NASA 2-14). Thus, even a weak ignition source is strong enough. The static charge accumulating on the human body can be up to 25 mJ. (New Zealand OSHA, Guidelines for Control of Static Electricity in Industry, 1999, sec. 3.3.4). Fortunately, the required ignition energy for detonation is higher: 10,000J (Hysafe).

AD-2-crshFor some airship fires, we know how the hydrogen was ignited. Frequently, the engine was the culprit. Woelfert’s Deutschland had an 8 hp Daimler engine with open flame ignition. In 1897, hydrogen expanded and blew off through the escape valve, mixed with atmospheric oxygen, and was ignited by the flame (Payne 28). Engine sparks were the problem on the LZ18/L2, and even just high heat from an engine or its exhaust was a risk (if it exceeded the autoignition temperature).

In 1921, a hull break snapped the gas lines of the R38, releasing hydrogen and spilling fuel over the engine (Payne 121). The R101 met its doom in 1930 when an impact ruptured a gas cell and also displaced the engine.


The LZ40 (L10), in 1915, was struck by lightning while valving gas. Lightning was also the suspected cause of loss of the Dixmude (1923), possibly compounded with a “rotten” gas cell or with venting (Payne 124).The Roma, in 1922, lost control and struck high-tension wires—what one might term, I suppose, “artificial lightning.”

This is perhaps an appropriate point for digression as to how to protect an airship from lightning. The first line of defense is to avoid places and times when lightning is likely to occur, and that of course turns on having good weather prediction capabilities. Still, lightning conditions can develop quickly, so that’s not good enough. Secondly, we want to make critical parts—the envelope, nose cap, empennage (rudder, elevators, fins), gondola suspension cables—electrically non-conductive. Of course, that depends in part on the pace of development of nonmetal materials technology. Finally, we protect the vulnerable areas by electrically bonding them via a common earth system and providing conduction paths to safe static discharge positions far removed from the hydrogen vent holes (Khoury 156-7). Concerning the last point, the skeletal structure of a rigid airship did provide some protection from lightning, serving as a “Faraday cage” if the metal parts were sufficiently interconnected, and there were also methods for electrical bonding of the hull (Robinson 70, Dick 52). And the 1929 NBS Code for Protection Against Lightning, section 251, required that free balloons and airships be provided an effective grounding wire to be dropped just before landing.


Static electricity could be generated by friction or impact. For both LZ4 and the LZ10 (Schwaben), ignition was by static electricity from a torn rubber-coated gasbag fabric (Payne 75).

In the case of the LZ69 (L24), in 1916 a gust lifted the craft, and a ceiling light bulb was broken, which ignited the hydrogen. That rather begs the question of why the hydrogen was free.

During wartime, one must also worry about incendiary bullets.

There are so many theories about the Hindenburg that I am not even going to discuss it!


Increasing Hydrogen Safety

I have considered several possible methods of increasing hydrogen safety. One is to add an inert gas so there is a non-flammable ternary air-hydrogen-inert gas mixture even if, without the inert gas, the binary mixture would be flammable. For non-flammability, we would need 11 volumes Helium (75oF), 10.2 Carbon Dioxide, or 7.6 Water Vapor (187oF) to 1 Hydrogen (Shapiro 1957, Coward 1952).

There are a few problems with this approach. First, the lift is reduced. Helium and water vapor are lighter-than-air but provide less lift than hydrogen, and carbon dioxide is heavier-than-air. Secondly, there may be a problem with maintaining homogeneity of the mixture throughout the gas cell. And finally, in the case of water vapor, heat is necessary, which means fuel costs—but one does get a bit of additional lift from the heated hydrogen component.

A second approach is to interpose additional barriers. For example, a hydrogen bag could be located inside a helium bag. Then any hydrogen diffusing out would mix harmlessly with the helium, and air would have to diffuse through the helium bag to get into the hydrogen bag. The lift for a given total hull volume would be reduced (because helium provides a bit less lift). If helium is not available, the hydrogen bag could be placed inside a nitrogen, carbon dioxide or water vapor bag. Those would carry more of a lift deficit, and generating the water vapor would require burning fuel. The heat would be transferred to the hydrogen.

In WW I, the Germans spread a rumor that their zeppelins were double-bagged, with helium in the outer bag, and thus invulnerable to incendiary bullets. It took a while for the Allies to learn otherwise.

With the exception of the aluminum used in the Schwarz No. 2 and the ZMC-2, all airships have contained lift gas within envelopes that are made of a combustible material. Such materials can be made more resistant to ignition by treating them with flame retardant salts, but they can’t be made non-combustible–they still provide fuel for the fire.


Hydrogen vs. Helium

AD-2-HvsHeThe only safe lift gas is helium; it is non-flammable. (While it can be argued that steam and hot air are nonflammable, generating them requires burning a flammable fuel.) There were essentially two reasons why airships were flown with hydrogen rather than helium on board; for increased lift, and because the helium was too expensive or outright unavailable.

With the ZR1 (Shenandoah), cells 82.5% full, flying at forty knots at 6000 feet, its range was 1790 miles; had it used hydrogen, this would have increased to 3760 miles. And the cost of helium was $120.22 per 1000 cubic feet, versus $2-3 for hydrogen.

On Hindenburg flight 10 (April 6-10, 1936), from Rio to Lowental (104 hours, 7000 miles), the weight of the fueled and crewed Hindenburg, ready to take on passengers and cargo was 454,924 pounds. Since its gross lift was 476,000 pounds (the Germans figured that hydrogen lift was 68 pounds per 1000 cubic feet), that left 21,076 pounds for payload. Had the gas cells been filled with helium (providing only 60 pounds lift per 1000 cubic feet), the gross lift would have been 420,000 pounds, and there would have been a 34,924 pound deficit–which presumably would have been solved by reducing crew, fuel, oil, provisions, ballast, and miscellaneous items.

Of course, the Hindenburg didn’t have a choice; the United States refused to sell helium to Germany. And that brings us to the supply problem. There are essentially two sources of helium; natural gas and the atmosphere. In the case of natural gas, as of 2008, a helium content of 0.2% was considered the economic limit (Smith). The first such wells discovered were in the North Texas-Oklahoma Panhandle-Western Kansas (Hugoton) area.

American (and thus world) production in the Twenties, in millions of cubic feet , was as follows: 1923, 7.7; 1926, 8.8, 1927, 6.0, 1928, 6.4, and 1929, 3.4 (Robinson 150). The Navy got half the helium. Now consider that the ZR1 (2.1 million cf) and ZR3 (2.43 million cf), in hangars, leaked 150,000 and 250,000 cubic feet, respectively, each month, and that the ZR1 valved off 640,000 cubic feet during one transcontinental flight (Robinson 63, 96). The Navy in that time period often only had enough helium to inflate one of its airships.

However, American production as of 2004 also came from Riley Ridge, Wyoming; Moab, Utah; and Southeast Colorado. At that time, there were also producing fields in Algeria (Hass R’mel field); Sichuan, China; Orenburg, Russia; and Odolanow, Poland. (Bowe).

The Odolanow plant was commissioned in 1977-78; until then, Europeans got their helium from the USA. The Odolanow natural gas is 0.4% helium (and also, peculiarly, up to 43% nitrogen) (Zuk and Lim).

Besides those fields, there are some known (as of 2004) sources of helium that had not yet been exploited. In the 1632 universe context, the most important ones of these are the East Hanover Field and the Salzwedel field. The Salzwedel field was discovered in 1968, and natural gas production began there in 1973.

It’s likely that the only subterranean source of helium readily found in Grantville literature is the Hugoton field, and it will be difficult to access for quite a few years.  It is impossible to guess how soon any of the European fields will be discovered and drilled in the new time line, let alone how soon it will be recognized that their natural gas contains helium in recoverable quantities.  And whichever natural gas field is first exploited, there will be further R&D and infrastructure development needed in order to reach the stage of recovering the helium in respectable volumes.

The helium content of the atmosphere is a mere 5.2 ppm. Helium can be partially separated from the air by liquefaction; unfortunately, the liquefaction product is 22.4% helium and 77.6% neon. Since neon’s density is almost the same as that or air, this helium-neon mixture has a density that’s 70% of air, which renders it fairly pathetic as a lift gas. And further purification is quite expensive. I do not believe that atmospheric helium has been commercially exploited for airship lift purposes, which require high purification.


Helium Conservation and Recycling Strategies

A variety of measures were taken in the old time line to conserve helium, and may reappear in the new time line once we are able to fill airships with helium. All of these measures had safety implications. Of them, installing apparatus that recovered water from the engine exhaust was probably the most benign; it did increase drag and engine complexity.

As heavier-than-air fuel was consumed (making the ship “statically light”), the ship could fly “dynamically heavy;” that is, fly with nose down to generate a compensatory amount negative aerodynamic lift. Unfortunately this also increased drag. Moreover, in the event of engine failure, the negative aerodynamic lift would fail, causing the airship to ascend rapidly.

A helium-based airship could attempt to arrange its schedule so that it landed in the morning when the helium was “supercooled” (at a lower temperature than the ambient air, thus providing less lift). That way it wouldn’t need to vent as much helium in order to descend to the ground.

Finally, some helium airships were deliberately flown with some of the automatic vent valves covered. These valves would normally open if the internal pressure rose too high (as could happen if a vertical air current caused the airship to ascend above what was called its “pressure height.”


With helium so rare in the old time line, recycling was important. In 2004 (not too long after RoF), the principal uses of helium were in liquid cooling of semiconductors (29%), as lift gas for balloons and airships (16%), and in metal-arc and plasma-arc welding (12%). The nature of the use dictates the type of impurity encountered. With lift gas use, the impurity will take the form of infill by air.

Room temperature purification of helium uses molecular sieves to absorb the nitrogen, oxygen and carbon dioxide. The feed gas is typically at least 20% helium (probably quite a bit higher if it is old lift gas helium) at a pressure of 40-350 psig and a temperature of 5-45oC. The sieve is then regenerated (the rejected gases evicted) by putting it in a partial vacuum. Unfortunately, there is probably minimal information on the composition and manufacture of these sieves in Grantville literature.

The other approach is by cryogenic purification; the helium is liquefied by compression. It must however be cooled below its critical temperature (5.2oK for He-4) for this to be possible, and the more it is cooled below that temperature, the lower the pressure required. For atmospheric pressure to be sufficient, the cooling must be to 4.2oK for He-4. Helium is mostly He-4. Helium was first liquefied in 1908.

While the basic principles of refrigeration and compression are certainly known to engineers in Grantville, building equipment that is capable of liquefying helium is not a trivial task.


Hydrogen Embrittlement

Hydrogen embrittlement (HE) is the degradation of a structural material caused by the presence of hydrogen in the material. The degradation has an adverse effect on strain hardening rate, tensile strength, fracture toughness, crack propagation rate, and the elongation necessary to cause failure. Failure can be premature and even catastrophic (Sofronis).

Several mechanisms have been proposed (and argued over). But simplistically, molecular hydrogen (H2) itself will not diffuse into a metal. For diffusion, you need atomic hydrogen (H). (Louhan 2) This could be generated by dissociation of molecular hydrogen at the surface, or because cathodic hydrogen (H+) from electrochemical processes (including dissociation of water) is reduced at the surface to generate the atomic hydrogen.

In any event, once absorbed, the atomic hydrogen can recombine to form molecular hydrogen, or react with the material to form hydrides, in either case perturbing the microstructure. Moreover, even atomic hydrogen can act as a barrier to the movements of dislocations, reducing ductility.

AD-2-brttlmntHydrogen embrittlement can occur even when the ambient hydrogen concentration is low, if the absorbed hydrogen is concentrated at points of structural instability. Stress can have a concentrating effect. (Herring).

Hydrogen embrittlement should be distinguished from other forms of hydrogen attack, such as chemical reaction of hydrogen with susceptible chemical functionalities of the material. In general, molecular hydrogen is not especially chemically reactive at room temperature, unless activated by a catalyst. Also, again absent a catalyst, its dissociation requires high temperatures. On the other hand, atomic hydrogen is a strong reducing agent, even at room temperature.


If an airship uses hydrogen for lift, then there are essentially four possibilities for contact between the hydrogen and materials potentially susceptible to attack

1) If the hydrogen is manufactured off-site, and transported in compressed gas tanks to the launch site, then the hydrogen interacts with the tank material (traditionally carbon steel).

2) The hydrogen must be pumped from the tanks or an on-site hydrogen generator to the gas cells, so there is interaction between the hydrogen and the piping material, which could be copper, plastic, rubber or leather.

3) The hydrogen is of course in contact with the gas cell material, probably for weeks or months at a time; the gas cell material may be a varnished cloth, goldbeaters’ skin, rubber, or some sort of composite material.

4) In flight, if net lift becomes excessive, it may be necessary to vent off hydrogen gas. The hydrogen comes in contact with valves (probably copper) and vent tubes (probably varnished cloth, rubber, or leather). After landing, if it is determined that a gas cell is “rotten” (too much air contamination of the hydrogen), the gas cell will be deflated and refilled, probably through the same piping material as for (2) above.

Hydrogen embrittlement is definitely a concern when storing hydrogen in steel cylinders under pressure for long periods. That in turn became commonplace when the switch was made from field production of hydrogen to factory production; the hydrogen would be placed in cylinders that were then transported to the field.

The ISO 11114-4:30005(en) standards for transportable gas cylinders require that if the steel cylinder is holding hydrogen at partial pressure above 50 bar (~50 times surface atmospheric pressure, which is 1.01325 bars), the steel must have an ultimate tensile strength of less than 950 MPa. If the hydrogen pressure is less than 50 bar, then the cylinders may be designed as for ordinary (non-embrittling) gases.

This ISO standard is not based on some mathematical model but on empirical experience, both lab experiments and what equipment failed in the real world. That of course doesn’t mean that it has captured all possible HE problems, but it makes it rather unlikely that we will see HE problems with compliant hydrogen storage tanks or with hydrogen at atmospheric pressure.

Several clarifications are in order. First, the standard is specifying a maximum UTS, not a minimum. A 2003 IGC document explains that in the late 1970s, there was a sudden increase in frequency of accidents with compressed hydrogen transport equipment. It was eventually realized that higher tensile strength steel had a microstructure susceptible to increased absorption of hydrogen when the hydrogen was at high partial pressure.

The flip side is that until the late Seventies—when hydrogen pressures and purities and steel UTCs were increased enough to render the metal more vulnerable to HE—compressed hydrogen tanks were considered to have acceptable levels of safety.

According to Dr. Richard Gangloff, the Ferman W. Perry Professor of Engineering at the University of Virginia, and the editor of Gaseous Hydrogen Embrittlement of Materials in Energy Technologies: The Problem, Its Characterization, and Effects on Particular Alloy Classes (2012), the “standard is valid and well supported for static loading. However, [if] the gas pressure continuously rises or if the loading is cyclic, then resistance to h[ydrogen] cracking is degraded even for these lower-strength steels.” (private communication).

Also, the specific guidance given by ISO is for a particular Cr-Mo steel alloy. The critical partial pressure and UTS for other steel alloys will be different. However, in general, steels that have a body-centered cubic (bcc) lattice are susceptible to hydrogen embrittlement. These include the Cr-Mo steels, which have ferrite pearlite microstructure, and “low alloy” (“tempered martensite”) steels.  On the other hand, steels with a face-centered cubic (fcc) lattice, such as the austensitic stainless steels, are resistant. (Gangloff communication). Thus, Cho (2007) reports that type 316 stainless steel shows high resistance to HE. The higher strength A286 is also resistant (Gangloff communication).


If the airship is a compound aerostat; i.e., the hydrogen is in gas cells that are inside an outer aerodynamic envelope, the main hydrogen-metal contact would be the valves for the gas cells. However, the embrittlement risk is dependent on the partial pressure of the hydrogen gas, and in the gas cells (unlike storage cylinders), that’s atmospheric pressure (~1 bar) or less.

In the 1632 universe airship operations for the foreseeable future, any valves made of steel would be lower tensile strength, and would be in contact with only atmospheric pressure of hydrogen (because the hydrogen would be produced on the field for immediate filling of the gas cells). The filling pressure may briefly exceed ambient, so the gas flows into the cells, but this is temporary; once in the cells it will be left at ambient. So the risk of HE is much less, and embrittlement isn’t likely to be an issue over the lifetime of the valve or even the airship.

The following table provides operating limits for steel in hydrogen service in petrochemical plants. Here the concern is with so-called hydrogen attack, a form of hydrogen embrittlement that occurs when hydrogen at high temperature (above 100oC) enters a metal and forms methane gas at the grain boundaries.

Note that our valves are operating in the extreme lower left, with partial pressure of 14.7 psia or less, and temperature typically around 15-20oC and ranging perhaps as high as 30-40oC for tropical operations. So even C-steel is considered OK.

Both the ISO standard and the just-cited table indicate that there is no problem with hydrogen at atmospheric temperature and pressure. If 1020 steel is exposed to hydrogen at one atmosphere pressure and room temperature, it will dissolve far less than one ppm hydrogen. However, if the same piece were acid cleaned, or exposed to moisture, there will be more hydrogen dissolved. (Louhan 3).

A further point worth mentioning is that HE is inhibited by oxygen in the ambient atmosphere. If oxygen is present, there will be an oxide film on the steel, and this apparently inhibits the dissociation of molecular hydrogen at the surface. (Louhan 3).

The hydrogen produced in the 1632 universe by steam/iron methods is unlikely to be more than 95% pure.  The balance will be air (which is about 21% oxygen, 78% nitrogen). So the gas will contain about 1% oxygen. Nitrogen is without effect, but oxygen inhibits both hydrogen absorption by steel and crack propagation. With hydrogen at atmospheric pressure, 0.6% (6 ppm) oxygen stopped crack growth in H-11 tool steel. (Hancock). Moreover, oxygen was still able to neutralize HE when part of a hydrogen-oxygen-helium gas mixture at elevated pressure (7 MPa), with an oxygen concentration of 29% (Basner).

A typical compressed hydrogen cylinder—I’ll used the ones sold by Praxair as examples—contains at least 99.95% (some 99.9999%) pure hydrogen and the gas is compressed to pressures of 138-460 bar. In contrast, our gas cells are perhaps 95% pure and at a pressure of 1 bar

I have seen reports of hydrogen embrittlement of steel with hydrogen at atmospheric pressure, but that was in an electroplating or acid cleaning operation context, and so there was corrosion to create surface defects that in turn promote hydrogen absorption and thus HE (and the surface defects also promote crack propagation). It is probably not a good idea to remove rust with an acid cleaner from a part regularly exposed to hydrogen, as that would result in surface defects and increase hydrogen absorption.

Hydrogen embrittlement is essentially a reversible process in steels; the hydrogen can be removed by baking the steel at a temperature of 300oF for three hours. (Gunn).

Chances are, given the steel shortage in the early 1632 universe, that valves will be copper, not steel. Copper is more resistant than steel to HE. However, problems have been experienced when copper and its alloys have been heated to temperatures of 400oC and up in a hydrogen atmosphere. If these metals contain oxygen (commercial copper is typically 200-6000 ppm oxygen), the hydrogen diffuses through the metal and reduces copper oxide at grain boundaries to copper and water (steam), the latter then creating intercrystalline cracks. These problems can be observed in samples heated in hydrogen for just a few hours. (Matting). “Oxygen-free” (actually, more like 10 ppm) copper is not subject to this “steam embrittlement” (Louthan 7), but also isn’t yet available in the 1632 universe. On the other hand, the copper on a hydrogen-lift airship isn’t going to be exposed to elevated temperatures. Dr. Gangloff confirmed to me that copper resists hydrogen embrittlement at near ambient temperature.


As to hydrogen embrittlement of the envelope materials, if we have compound aerostats as on the Danish rigid, the goldbeater skin used to hold the hydrogen is under minimal stress, so HE impairment of strength (if any) is not of importance. If we have simple aerostat, perhaps rubber/cloth or varnished cloth, there is some strength requirement for the envelope, but it is only resisting the combined stress imposed statically by the internal overpressure (on order of 500 Pa, i.e., 0.5% atmospheric) and by aerodynamic forces (mostly on the nose where they are taken mostly by the nose cone).

It is unclear at the time of writing whether hydrogen degrades polymers (rubbers and plastics). The very limited studies of polymers have been in the context of polymeric components in use or considered for use in constant or cyclic high-pressure hydrogen service, such as in storage tanks, pipes, and valve seals.

In 1975, Hust (4-14) said “no evidence has been presented which indicates serious embrittling effects of hydrogen on non-metals. However, very little work has been done in this area. . . .” Hust went on to cite a White Sands study to the effect that room temperature vulcanizing silicone rubber, but not PVC, PTFE, or cellulose acetate butyrate, was adversely effected by hydrogen.

According to an article published by the European Commission project Roads2HyCom:

“In the 70s and 80s Brown has performed research towards the influence of helium, nitrogen, argon, oxygen, and water on the mechanical behaviour of some polymer materials (PE, PP, PMMA, PS, PET). The findings of this study are that some effects are noticeable till about 100 C above the boiling point of each gas. The boiling point of hydrogen is -253 C. So, based on the research by Brown it can be expected that hydrogen at normal surrounding conditions will have no physical interaction with polymer materials to the extent that it will cause a measurable deterioration in properties. ”

Fiber-reinforced polymer, in particular polyethylene, has been proposed as a hydrogen pipeline liner material. In 2008, Kane stated that no mechanisms for polyethylene degradation due to hydrogen alone have been reported. However, Kane cautioned that contaminants might provide degradative mechanisms.

The 2012 Sandia report on hydrogen compatibility of materials stated, “we are unaware of hydrogen compatibility studies for common polymer materials” (8100-1). However, it speculated that since hydrogen is quite soluble in polymers, exposure to high-pressure hydrogen might cause damage (blistering or swelling).

In a 2012 paper, Yamabe observed that hydrogen penetrates a rubber material as molecular not atomic hydrogen, and reported that repeated cyclic exposure to high pressure hydrogen (5-90 MPa, atmospheric is 1.01) could cause cracking of synthetic acrylonitrile-butadiene rubber O-rings installed in compressed hydrogen tanks. This cracking is the result of blisters formed when the hydrogen that has diffused into the rubber is rapidly decompressed and therefore expands. While that is a mechanism for degradation by hydrogen, it is not what is conventionally referred to as hydrogen embrittlement and is seen in many elastomers when other high pressure gases are rapidly decompressed. (Fitney 442).

The airship gas cells are only at slightly above ambient atmospheric pressure, and even if they rupture as a result of aerodynamic stress, someone dropping a tool onto the gas cell, etc., the difference in pressure between the lift gas inside and the ambient air outside is only about 500 Pa at most (for a nonrigid), which is 0.5% of atmospheric pressure—probably not enough of a pressure change to cause blister damage, and anyway the blister damage pales beside the rupture. I would presume that the normal deflation of an intact gas cell to refill it with fresh gas is slow enough so that blister damage is not a concern.

Also in 2012, Klopffer reported that after one year under 3MPa (three times atmospheric pressure) at room temperature, the tensile static properties, long-term creep deformation, and ductile fracture of two plastics (a polyethylene and a polyamide) were no more than 10% different from their values in air (Hecht).

Overall, I have not seen anything to suggest that HE is a problem for natural rubber exposed to H2 at atmospheric pressure and temperature, outside a corrosive environment. I’d actually be worried more about embrittlement in winter as a result of cold.

As a check on my literature searching and analysis, I consulted Doctor Gangloff, who wrote me, “your research into this topic is solid.  I am not aware of any published works on low pressure H2 degradation of polymers and my intuition suggests that this is not an important degradation mechanism.”

Finally, even if HE were a problem with the 1632 universe gas cells, that probably would not deter the use of hydrogen any more than the more obvious risk of fire and explosion did—that hydrogen provides more than three times the lift of hot air and doesn’t require fuel consumption to stay aloft are powerful incentives in a world where helium is not going to be an option for decades.


Molecular hydrogen can react chemically, and thereby degrade, polymers with carbon-carbon double bonds. Natural rubber is cis-1,4-polyisoprene, and does possess such bonds. However, for it to be hydrogenated by hydrogen gas, a catalyst, and often also high pressure and temperature, is required (Samran 132).


A final pathway to hydrogen exposure, and possible hydrogen embrittlement, that I considered, was one potentially applicable if the airship is a compound aerostat (gas cells inside larger hull) and also is a rigid airship (outer hull envelope is stretched over a structural skeleton). If a gas cell is ripped open, then there would be release of hydrogen, creating a low partial pressure of hydrogen (say 5% atmospheric) in the interstitial space. This hydrogen-contaminated air would come into contact with the frame (skeleton) of the airship. The skeletons used in rigid airships have been wood, steel and duralumin.  The exposure would be for at most days because of course on landing they would vent out the hydrogen-contaminated air.

According to Dr. Gangloff, wood and aluminum would not be adversely affected by such hydrogen exposure. Also, if the steel were of the lower strength type, it wouldn’t be susceptible, either. “The only possible (but unlikely) degradation would be in fatigue resistance should the frame vibrate for many load cycles. In this case, the presence of oxygen with the hydrogen is beneficial, but water vapor + hydrogen could be detrimental” (private communication).


We will need to worry about hydrogen embrittlement if and when we start shipping hydrogen to airship hangars in compressed hydrogen tanks — but that isn’t going to be in stories published in the foreseeable future.



Airship Successes

To put this two-part articulation of the failures of historical airships in perspective, it is only fair to say something about historical airship successes. The world’s first commercial airship company, the German DELAG, was in operation 1911-14. During its four years of operation, it chalked up 3200 hours in the air, 1600 flights, 100,000 miles, and 37,750 passengers without fatality (although not without crashes and airship losses) (Payne 74). One of the DELAG airships, the LZ10/Schwaben, flew almost 100 flights without mishap in 1911.

AD-2-R34In 1919, the R34 achieved the first aerial crossing of the Atlantic from East to West (taking 108 hours), and also the first aerial round trip across the Atlantic (the West-East passage was 75 hours).

The USS Los Angeles, on behalf of the Navy, conducted 275 hours of scouting, traveling over 14,000 miles in the process (Payne 137). Over its eight-year flying career, it made 331 flights, totaling 4,181 flight hours (

The first aerial circumnavigation of the world was in 1929 by the Graf Zeppelin. It needed three refueling stops, and spent a total of 300 hours in the air. This circumnavigation included the first nonstop flight across the Pacific: the 5400 miles from Tokyo to San Francisco, flown in 68 hours. On its Berlin-Tokyo leg, it set a nonstop distance flight record of 6980 miles (Payne 174).

In World War II, the US Navy used blimps for antisubmarine patrols. Of the 168 USN blimps (134 of which were K-class), about 120 were still in service at the war’s end, despite having flown frequently under adverse weather conditions, and having been exposed from time to time to enemy fire (Payne 242).

Finally, in 1957, the ZPG-2 blimp Snowbird flew nonstop across the Atlantic twice—9,448 miles in eleven days. (Payne 244).

There is no doubt that there will be airship accidents, but airships will find a niche in the 1632 universe.






Helium Recovery and Recycling Makes Good Business Sense. Industrial Gases/Combustion., Donald J. Bowe, Air Products & Chemicals Inc., Allentown, PA, Sept, 2004.


Smith. The Global Helium Outlook. Air Products, 2008.


Belafi. The Zeppelin. 2015.


Payne. Lighter Than Air: An Illustrated History of the Airship. 1977.


Dick. The Golden Age of the Great Passenger Airships, Graf Zeppelin and Hindenburg. 1985.


Robinson. Up Ship! A History of the U.S. Navy’s Rigid Airships 1919-1935. 1982.


Stahl. “Medical Investigation of Airship Accidents,” J. Forensic Sci., 33: 888-98. 1988.


[PASS] Puget Sound Airship Society. “The Zeppelin Airships.”


Mitchell. Resume of German Research Development in Aircraft Construction, Air Service Information Circular, 4 (391): 63 (Nov. 1, 1922).


Hydrogen Safety


College of the Desert. Module 1: Hydrogen Properties, in Hydrogen Fuel Cell Engines and Related Technologies. December, 2001.


Edeskufy. Safety in the Handling of Cryogenic Fluids. 2013.


Hysafe. Hydrogen Fundamentals.


NASA. Safety Standard for Hydrogen and Hydrogen Systems. 1997.



Hydrogen Embrittlement and Related Issues


Basner. The Effectiveness of Oxygen in Preventing Embrittlement in Air Bag Inflators Containing Gaseous Hydrogen. 2006.


Fitney. Seals and Sealing Handbook. 2014.


Gunn. Hydrogen Embrittlement of Steels. 1964.


Hancock. Trans. Met. Soc. AIME, 236: 513-6 (1966).


Hecht. Review on the effects of hydrogen at extreme pressures and temperatures on the mechanical behavior of polymers. Sandia Report, 2013.


Herring, Hydrogen Embrittlement.


Hust. Survey of Materials for Hydrogen Service, in Parrish, Selected Topics on Hydrogen Fuel. 1975.


Kane. Permeability, Solubility, and Interaction of Hydrogen in Polymers—An Assessment of

Materials for Hydrogen Transport. Washington Savannah River Co.: 2008.


Louthan. Hydrogen Embrittlement of Metals: A Primer for the Failure Analyst. 2008.


Matting. “Brittleness in Copper and Copper Alloys With Particular Reference to Hydrogen Embrittlement.”


Roads2HyCom (EC). “Effects of Hydrogen on Materials”,


Samran. A Study of Non-Catalytic Hydrogenation of Natural Rubber. Ph.D. Thesis: 2005.


Sandia Report SAND2012-7321. Marchi. Technical Reference for Hydrogen Compatibility of Materials. 2012.


Sofronis. Hydrogen Embrittlement of Pipeline Steels: Causes and Remediation. 2005.


Yamabe. Hydrogen-induced degradation of rubber seals, in Gaseous Hydrogen Embrittlement of Materials in Energy Technologies: The Problem, its Characterisation and Effects on Particular Alloy Classes. 2012.