[ryanv777]

So, I love fire, and enjoy watching it, and I like the way that it lights up a room. I was wondering, for a sort of night light, if it would be practical to create a device which, with the addition of electricity, would perpetually burn hydrogen. It would have a water reservoir which would create hydrogen and oxygen via electrolysis and then burn it in a glass box filled with oxygen. The produced steam would then be condensed and funneled back into the water reservoir. Sound practical? Would I be able to generate a sufficient supply of hydrogen to sustain a fire as bright as that of a candle, minimum, using the power from a wall outlet? 

[Hunter2]

A hydrogen flame is colorless. You need at least some sodium to get a yellow color.

[Arkcon]

create hydrogen and oxygen via electrolysis and then burn it in a glass box filled with oxygen.

Oxy-hydrogen flames are at least a little bit dangerous.  I used to play with tiny amounts, for an explosive *pop* sound, but an oxy-hydrogen flame, candle-sized, is more appropriately called an oxy-hydrogen torch, used for cutting hard metals.  So one wonders about the strength of your container.

The produced steam would then be condensed and funneled back into the water reservoir. Sound practical?

Not really, no.  Consider, at least some water vapor is produced by burning candles, but we don't often see condensed liquid water in the vicinity.  Its just too hot nearby.

Would I be able to generate a sufficient supply of hydrogen to sustain a fire as bright as that of a candle, minimum, using the power from a wall outlet? 

Now that is an interesting thought problem.  Seems like people were thinking about this problem during the last gasoline crunch.  They wanted to use "spare" electricity from an automobile to generate enough hydrogen and oxygen to make a difference in their car's mileage.  It seems like they had something sustainable, if not absolutely efficient, or indeed efficient at all.

[Enthalpy]

Usual light bulbs consume 40-100W to produce <5W light, so if the efficiency were the same, 100W electrolysis would make as much light - with added sodium and everything needed.

But this is very indirect and probably inefficient. Some of the combustion heat (or radicals) would excite a small proportion of the sodium atoms, and some of the excited atoms would convert the energy into light.

Acetylene with limited air supply makes a strong white light, which unseeded hydrogen doesn't. An electric cycle can regenerate C₂Ca from Ca(OH)₂ and CO₂. Quite indirect!

As opposed, you can make an electric discharge directly in sodium vapour. If well built, this sodium lamp provides 40% efficiency, the best among electric lamps.

Sure, they don't have the fascinating flame.

[Intanjir]

Acetylene with limited air supply makes a strong white light, which unseeded hydrogen doesn't. An electric cycle can regenerate C₂Ca from Ca(OH)₂ and CO₂. Quite indirect!

Flames burning a carbon fuel without enough oxygen for complete combustion will produce a nice red flame thanks to soot. The soot is nice and black and quite diverse in the kinds of molecules present. The result is something not totally unlike a blackbody spectrum:
http://www.physics.sjsu.edu/tomley/ObjectSpectra/MatchFlame.html
https://en.wikipedia.org/wiki/Oxidizing_and_reducing_flames

In any case, in order to regenerate the flame we would need complete combustion, ie no soot and no white light. So I don't think this cycle would work but it was a great idea!

[Enthalpy]

Acetylene does make a nice strong white flame. Since the users survive, the reaction must be complete to CO₂. For sure, an acetylene lamp working normally makes no soot.

What exactly happens with the limited air supply isn't clear to me. It isn't an incomplete combustion. It works with petrol lamps as well, which give more light with the glass on. So do oxygen+acetylene burners, which produce much more light with less oxygen than the hottest ratio, but still make no long-term CO nor soot.

A hypothesis - and nothing more - is that too much air quenches the flame more quickly, be it by cooling or by de-exciting species that would otherwise radiate.

I'm convinced that hydrocarbon flames radiate by non-thermal processes, as the example of acetylene+oxygen shows, where a cooler flame is much more brilliant.

[Intanjir]

I think it could still just be soot that is hot enough that it will eventually combust when exposed to enough oxygen. So I guess my earlier suggestion was wrong in that I didn't consider whether it was possible to have soot in the flame whilst still having complete combustion eventually.

Given its transparency in the visible, I think that the earlier you make CO2 the dimmer your flame will be in general. So I guess this is just the replacement of more vibrant carbon species with the invisible CO2 species.

[Intanjir]

Given the presence of water, I do not think sodium is a good choice for enhancing the flame. I suppose you could introduce soluble flame enhancers by wicking them up them along with the water and then boiling the water in the flame, but this will greatly diminish your achievable temperature.

I think just using a noble gas like neon would work well if everything is a sealed closed system. The hot neon actually in the flame should glow whilst the rest remains invisible.

I think managing the dynamics of the flame may be tricky. Hydrogen burns very readily and I think this would make it difficult to have a self-supporting continual flame. As little as 4% hydrogen in air will explode, though it is more like a 'pop' at this low concentration. Popping like this would blow out any nearby flame. Enriching the oxygen levels in the system is a bad idea since this will only make hydrogen explode sooner.

Unfortunately the atmosphere in a closed system would become enriched with unreacted hydrogen and oxygen further lowering the explosive bound. This enrichment would of course reach an equilibrium level, but I doubt it would be a workable one for a self-supporting flame unless there was a lot of extra inert gas.

I think the best way to make it work would be to use principally an unreactive gas which emitted in the visible and to continually support the combustion via a heated nozzle or a spark. You might be better off keeping the oxygen and the hydrogen separate unless there was a couple extra atmospheres of inert gas to dilute the temperature.