Other alkali metals, such as lithium, could be used in placed of sodium; 22.5 pounds of lithium hydride, reacted with equivalent water, would produce 1,000 cubic feet hydrogen. (Roth 30). They are, if anything, more expensive than sodium so these are strictly laboratory methods.
The same is true of the alkaline earth metals, of which magnesium will probably be the cheapest.
Aluminum amalgam-water. If a small amount of mercury (or mercuric salt) is added to aluminum powder, to make an amalgam, the latter will react with water to form aluminum oxide, hydrogen, and pure mercury. The latter may be reused to make more amalgam. One pound of aluminum yielded 20.5 cubic feet hydrogen. (Taylor 129).
This procedure was practical in the early-twentieth century, thanks to the Hall-Heroult electrolytic process for making aluminum.
The Mauricheau Beaupre "activated aluminum" variant involves adding water to a mixture of fine aluminum filings, mercuric chloride, and mercuric cyanide. One kilogram solid mixture, so reacted, yields 1.3 cubic meters in about two hours. The apparatus required is minimal. (Ellis 525). The aluminum must not contain copper (Teel 70).
Hydrolith Process. This was another field expedient, exploiting the reaction
CaH2 (42 grams) +2H2O (36 g) -› Ca(OH)2 (56 g) +2H2 (4 g).
So only 55 pounds of calcium hydride is needed to obtain 1,000 cubic feet hydrogen. The calcium hydride would be made at base, from a calcium salt (oxide, chloride) and hydrogen in presence of a reducing agent (sodium, magnesium).
The French used this system in the early-twentieth century; the calcium hydride was carried on latticed trays, immersed in water; the hydrogen rose up. This gas was contaminated with water vapor, which was removed by passing it over dry calcium hydride. (Taylor 128). You also need to remove ammonia, and heat evolution can be a problem. A typical six generator wagon produces 15,000 cubic feet/hour. (Greenwood 229).
A related, speculative process uses lithium hydride:
LiH2 (9 grams) +4H2O (72g) -› 4 LiOH (75g)+ 3H2 (6g).
Note the enormous yield of hydrogen relative to the amount of lithium hydride. This would be great for the field. The ratio is good enough so it's feasible (from a weight, not necessarily a safety standpoint) to bring the lithium hydride on board for use at a destination to make more hydrogen. It has even been suggested that the reaction could be used to produce hydrogen while in flight, reacting the hydride with water ballast (warning: this can be a violent reaction!), and then dropping the lithium hydroxide. (Teel 67). But the cost of lithium hydride, which is made by reacting lithium metal with hydrogen; is prohibitive (even 1992 price was $72/kg-Kirk-Othmer).
Silicol Process. The basic reaction was
2NaOH+Si+H2O-›NaSiO2+2H2.
It was first proposed in 1909, and became a popular military field expedient, especially on ships. Not only were the ingredients quite safe to transport, the produced hydrogen was of "very high purity." (Taylor 143).
Initially, commercially pure silicon was used, but this was replaced by the cheaper ferrosilicon, which was used for deoxidizing steel and introducing silicon into alloys. Ferrosilicon may be made by reducing sand (silica) with coke in the presence of iron. The ferrosilicon typically contains small amounts of phosphine, arsine, and hydrogen sulfide. (Greenwood 227), as well as air. (Teel 45). The silicon content has to be over 80% for reasonable effectiveness, and particle size affects the production rate. (Teel 50). The caustic soda must be neither too dilute nor too strong. (Teel 52).
In addition, there is an explosion hazard. The ferrosilicon dissolves only slowly in cold solution, and thus can accumulate. But the reaction produces heat, and as the solution gets hotter, the accumulated ferrosilicon is attacked, leading to rapid evolution of hydrogen. (Teel 57).
A transportable plant can produce 60-120 cubic meters/hour, whereas stationary plants of up to 300 capacity have been constructed. (Ellis 523) [The typical portable plant was mounted on a three ton truck and produced 2,500 cubic feet/hour; the largest portable apparatus produced 14,000 cubic feet/hour. The reaction has also been used for stationary production at up to 50,000 cubic feet/hour. (Greenwood 226).
For the 1929 British R100 airship, "249 tons of caustic soda and 183 tons of ferro-silicon produced 8,610,705 cu.ft of hydrogen (20.3 tons) and 929 tons of sludge [sodium silicate]." (Wilcox). The R100 had a gas capacity of about 5,000,000 cubic feet, so the gas produced was substantially in excess of the capacity. Wilcox's information about production rate is somewhat contradictory. He says that the plant could produce 60,000 cubic feet/hour, but that the highest daily production was 500,000 cubic feet. Also, that it took ten days to fill fourteen of the R100's fifteen gas bags.
An alternative reaction that can use the same apparatus exploits the reaction of aluminum with sodium hydroxide, and was used by the Russians in the Russo-Japanese War. (Taylor 145-6). It can produce 10 cubic meters/hour. (Ellis 523).
Hydrogenite process. This starts with a compressed block of a mixture ("hydrogenite") of silicon, caustic soda, and soda lime, kept in an air-tight container. To use, the container is placed in a water jacket, a match or a red hot wire is applied to a small hole in the lid. The silicon is oxidized to silica, a heat-releasing action. This heat makes possible the reaction
Si+Ca(OH)2+2NaOH-›Na2SiO3+CaO+2H2.
The heat turns the water to steam and eventually this is permitted to enter the generator, increasing yield by a reaction of the silicol type.
While it requires that 50% more material be provided than for the silicol process, much less water is needed, which would be advantage for desert use. (Taylor 168). A production rate of 150 cubic meters/hour is possible(Ellis 521ff). The portable wagon-based apparatus of the French army, featuring six generators grouped around a central washer, produced 5000 cubic feet/hour. (Greenwood 228).
****
Another "dry" method (by Majert and Richter) involves heating a mixture of zinc dust and slaked lime to redness, but the Prussian army deemed it too slow (it took 2-3 hours to fill a balloon). (AGLJ).
Large-Scale Production
Some processes are best suited to production of hydrogen on a large scale and at a low cost. Unless the airship hangar happens to be near the manufacturing plant, the gas will have to be compressed and shipped in containers (which must be returned empty), which increases the cost.
Steam-Carbon. First, water gas (a mixture of carbon monoxide and hydrogen) is produced by reacting red-hot coke or coal with steam at 800 or 1000oC (2002McGHEST):
H2O (18 grams) + C (12 grams) -› H2 (2 grams) + CO (28 grams).
Just making steam, by itself, consumes fuel. According to EB11/Railways, the faster you burn coal, the lower the efficiency. With Indiana block coal (13000 BTU/lb):
Those are for a 1900 locomotive boiler. and a stationary plant might have a higher efficiency. Additional coal would need to be burnt to superheat the steam to the required temperature. The increase in coal consumption to achieve 100oC superheat is 5.5%, for 150, 8.3%, and for 200, 11%. (Stovel 1475). (Superheated steam is more efficient than ordinary steam, however, in terms of the heat content of the steam relative to that of the coal burnt to produce it. (Babcock 137ff).