In 1904, this was the method used to produce hydrogen for German army balloons. (Englehardt 123).
Water Thermolysis. The thermal decomposition of water requires temperatures in excess of 2000oK, and of course reactor materials that can tolerate the temperature. (Yurum 24).
Splitting Hydrogen Sulfide. The use of hydrogen sulfide as a source of hydrogen has been proposed, but not commercialized. One possibility is to react it with iodine, producing sulfur and hydrogen iodide, and then decompose the latter. Another is to react it with methane, forming hydrogen and carbon disulfide. (Kirk-Othmer 13: 874). These methods probably do not appear in Grantville literature.
Thermal Decomposition of Hydrocarbons. When exposed to sufficient heat (1200-1300oC for methane, 500oC for acetylene), hydrocarbons dissociate into their component elements. (Ellis 471).
The Rincker-Wolter system is of some interest because they started with oils and tars, and the demand for tar in163x is limited. The required temperature was 1200oC for the hydrogen to be of acceptable purity. In 1912, a plant producing 3500 cubic feet/hour would cost $2575 plus "erecting expenses," and with the oil at 4 cents/gallon, the hydrogen cost would be $1.75/1000 cubic feet. (Ellis 473ff). A semiportable plant with such capacity has been successfully mounted on two railway trucks. Greenwood 193 reports a cost of 550 pounds sterling for the plant and 2s/6p to 4s/0p per 1000 cubic feet.
A variation on this is the Carbonium process; acetylene gas is compressed to two atmospheres and exploded by an electric spark, yielding carbon (deposited as lamp-black) and high purity hydrogen. You need an explosion chamber, and the lamp black is scraped off the walls. If there's a market for the lamp black (one kg per cubic meter of hydrogen), this method can be advantageous. (Ellis 473). The good news about the process is that it was used to supply hydrogen for the zeppelins at Friedrichshafen. The bad news is that the factory was destroyed by an explosion in 1910! Before this slight mishap, the cost of production was 4 shillings per 1000 cubic feet. (Greenwood 192).
Catalytic steam-hydrocarbon reforming. Per McGHEST2002, volatile hydrocarbons (from natural gas) are reacted with steam over a nickel catalyst at 700-1000oC, forming hydrogen and carbon monoxide, the latter being converted to carbon dioxide by reaction with water at 350oC over an iron oxide catalyst. If the hydrocarbon were methane (which has the highest hydrogen:carbon ratio), the first reaction would be
CH4+H2O-›CO+3H2.
The encyclopedia notes that carbon dioxide may be removed by scrubbing with aqueous monoethylamine. However, there's a much easier method; pass the gas mixture through water under high pressure; the carbon dioxide reacts with water to form carbonic acid and dissolves; the hydrogen doesn't dissolve and bubbles to the surface.
Even with these hints, the method may take a while to get working. In the old time line, experiments began in 1912, but the first real success, with methane over a nickel catalyst, came in the 1920s. It wasn't commercialized until 1931. (Smil 113). Note that the reforming catalyst isn't necessarily the simple metal; the 1962 ICI process used nickel-potassium oxide-aluminum oxide (Weissermel 18).
In the 1632 universe, the likeliest source of the volatile hydrocarbons would be coal gas, but natural gas would also be an option. However, we do have to find the right catalyst, and the feedstock may include substances (sulfur, chloride) that poison the catalyst.
Immediately prior to the RoF this was the dominant method of producing hydrogen. the Kirk-Othmer Encyclopedia of Chemical Technology reports that it had a thermal efficiency of 78.5%, versus only 27.2% for water electrolysis, and a net hydrogen production cost of $7.19 (1985 dollars)/100 m3, versus $22.63 for electrolysis. (13:853). The theoretical energy consumption is 300 BTU/scf hydrogen, and a typical one is 320. If the natural gas price is $4/million BTU, feedstock and utility costs are 65% total operating costs. (Udengaard).
Catalytic steam-methanol reforming. This is a related process:
CH3OH-›CO+2H2.
Steam reacts with the carbon monoxide to form carbon dioxide and more hydrogen. The process has been proposed for modern field use, with a one ton trailer-mounted generator producing 150 cubic feet/hour with fuel consumption of just over one gallon/hour. (Philpott).
In the old time line, it had higher operating costs but lower fixed costs than the hydrocarbon-based scheme. (Blomen, 150). Other advantages are that methanol is free of sulfur and the reaction can be run at a lower temperature (300°C). (Liu 65). However, the problem is that experimentation will be needed to find appropriate catalysts (likely to be copper, zinc oxide or palladium-based).
Decomposition of Ammonia. The first problem is producing the ammonia. Nowadays it's made by the Haber process from nitrogen and hydrogen, but obviously if our goal is hydrogen, we are taking a different route. There is ammonium carbonate in New World guano deposits, and ammonium sulfate is a potential byproduct of the manufacture of boric acid in Tuscany. We also must determine the decomposition conditions. Ammonia can be decomposed by heat alone but a catalyst helps. (Lunge 580ff) Ammonia is another possible target for catalytic reforming. (Udengaard).
Fermentation. The Weiszmann process for the manufacture of acetone and butyl alcohol by fermentation of starchy foods (maize, potatoes) also produces hydrogen and carbon dioxide in equal volumes; 5.5 cubic feet of mixed gas per pound of maize fermented. (Taylor 166). Both acetone and butyl alcohol are quite important industrial chemicals, and so the sale of hydrogen needs to only cover the cost of its separation from carbon dioxide. The real problem is isolating the necessary fermentation organism.
****
Disaster Scenarios. As the hydrogen is produced, it mixes with any air that is present, soon creating flammable or even explosive mixtures. Ideally, several volumes of an inert gas (carbon dioxide, nitrogen) or liquid (water) are run through the production chamber first, to drive out the air. Also watch out for leaks from the gas hoses.
If the reagents are stored close together, and their containers are ruptured, an uncontrolled reaction can occur.
Some of the reactions are exothermic, so even if the reaction is in the proper vessel, the temperature has to be monitored.
Comparative Operating Costs
I was able to find some comparative operating cost data on the different production processes. Some sources include labor and overhead (interest and depreciation on fixed costs), and others don't.
Figure that one 1900 US dollar is 4.2 contemporaneous British shillings, or 0.5 1632 shillings if deflated based on Allen's laborers' wage rates, and 1.25 if using Allen's London CPI. (Two shillings is equivalent to one Dutch guilder. ) That same 1900 dollar is $19.57 in 2000 if inflated using the Sahr CPI.
The NTL economy in 1635 is going to be very different than that of pre-RoF Europe, and also different from that of OTL early-twentieth century Europe. Hence, be cautious about putting a lot of faith on cost conversions. It's probably better to use the table to get a sense of relative rather than absolute costs, but even that's dangerous; individual inputs (e.g., electrical energy) could be cheaper or more expensive in the new universe, even different from one region to another.