The Oswalds get it partly right

The reason for the popularity of hydrogen is that every other energy source translates into it pretty well. It forms a kind of middleware.

A ton oil equivalent is defined (as far as I can google) as the amount of energy required to burn ten to the seventh kilocalories and is a measure taken at the beginning of the energy use chain, not the end. This means that, yes, the Oswald brothers have made a rather large error. Hydrogen actually is rather more efficient than internal combustion engines. But there is another error, that the next stage in our energy system development is going to be a monoculture. Instead it is very likely to be the most diverse period of energy generation in many centuries.

You can make hydrogen from animal husbandry and agricultural wastes. Household wastes can also be turned into feedstocks. Entirely wasted energy such as all the muscle effort engaged in lifting weights and running on treadmills becomes useful as mini hydrogen producers. Solar energy is also likely to be more practical in future as orbital solar power stations that beam their power on wavelengths that are unaffected by cloud cover mean that the UK can use energy collectors far beyond its own shores.

In short, the Oswalds destroyed a straw man but furthered a myth, the myth of monolithic power generating methods.
 

I would dispute that most other energy sources "translate well" to hydrogen.  The only one which appears to do so is syngas (mixture of H2 and CO) which can be shift-converted to pure H2 at some loss of energy content.  Electricity is closest to the universal medium, AFAICT.  When you already have electricity, it makes no sense to take the losses of the conversion to and from chemical energy if you can avoid it and the losses you must take ought to be the smallest you can find (until some future technology makes electricity cheaper than dirt, of course).

Hydrogen may work well in internal combustion engines, but the engines themselves are still going to be very inefficient.  If you have a 70% efficient electrolyzer feeding a 30% efficient combustion engine, your overall efficiency falls to 21%.  Contrast this to what you could have with a 70% efficient battery and a 90% efficient motor:  3 times the go for your buck, and no nitrogen oxide emissions or noise.

You can make hydrogen from animal husbandry and agricultural wastes. Household wastes can also be turned into feedstocks. Entirely wasted energy such as all the muscle effort engaged in lifting weights and running on treadmills becomes useful as mini hydrogen producers.Yes and no.  Direct fermentation of manure and such appears to generate methane, which you'd have to reform (at some expense and with losses) to get H2.  Thermal depolymerization doesn't appear to generate much H2 either; you'd have to convert the hydrocarbons, again at a considerable loss of energy.  The hydrocarbons are more compact, transportable and storable than hydrogen.  Why would you bother converting to a medium with all those built-in disadvantages over what we have right now?

I can see a few reasons to generate hydrogen directly, using solar-powered processes (either photochemical or algal):

1.)  Generation of nitrates; can't run the Haber process without hydrogen.
2.)  Conversion of available CO2 to methanol or other storable fuel.

Getting back to the Oswalds, their paper is bound to get some people to write the concept off and pique the interest of others.  Getting the numbers right ought to push people from the former view toward the latter, and that can only be a good thing.

FWIW, I think that thermal depolymerization may have a bright future as a way of eliminating municipal garbage.  If all the organic trash can be turned into liquid fuels and combustible gas, then you've got the supplemental fuel for your almost-all-electric renewable economy.
 

Electricity works as a universal medium as long as you don't have to store it. A great deal of the problem of renewable energy generation is storage as renewable generation tends to be bursty. Electrical grids don't tolerate large fluctuations in generation either up or down very well. Hydrogen and fuel cells seem to have a better improvement curve and are already starting to displace lead acid and other batteries in certain operations. With hydrogen storage and fuel cell efficiency improvements, the number of practical hydrogen applications are likely to grow reasonably quickly.

You're correct that hydrogen internal combustion engine technology is limited which is why most hydrogen research tends towards the fuel cell side of things which gives you increased flexibility in designing solutions and efficiency gains from those looser constraints (you don't need to have a forward engine compartment in your car, for example). A car that requires a road electrical grid is never going to gain wide acceptance because you'll need to do the entire network before people stop worrying that their new, expensive car will not take them where they might want to go next month.

Fixing the "bursty" nature of renewables makes them a more realistic alternative. Eliminating the need for new fuel distribution chains by using a middleware fuel will bring new energy alternatives online as a practical matter. Hydrogen's where we're going to end up.
 

Electricity works as a universal medium as long as you don't have to store it.Since indirectly-produced hydrogen is mostly going to be just another method of storing electricity, that doesn't offer any useful information.  The issue is how electricity can be stored for various purposes, not whether we'll have to.

A great deal of the problem of renewable energy generation is storage as renewable generation tends to be bursty. Electrical grids don't tolerate large fluctuations in generation either up or down very well.Not true; the difference between peak load and minimum load can be 2.5:1 or so.  What the grid can't tolerate is large instantaneous differences between generation and demand, which is another issue entirely.  You can manage the balance by tuning the supply to keep it in line with demand, or you can tune demand to balance it with supply.  The latter approach is often used with small stand-alone hydropower systems; excess power is diverted to a "dump load" so that the generator is not allowed to overspeed.  This same approach can be used on the grid, where the presence of controllable loads such as electric or plug-in hybrid cars, variable-speed and ice-storage A/C systems, and even electric water heaters can allow load to be varied to match instantaneous supply rather than the reverse.

The same technologies useful for peak shaving could also be used to exploit "bursty" energy supplies.  One example is
flow batteries (demo project document here).  If everyone drove a battery-powered vehicle with 1/3 the capacity of the Li-ion tzero (20 KWH), the buffering capacity in a moderately-sized metropolitan area would be measured in gigawatt-hours.  Plug-in hybrids wouldn't suffer any loss of usability if the electric supply wasn't sufficient for their needs, they'd just switch over to burning fuel.  This is a perfect match to semi-regular but unschedulable generation capacity.

A car that requires a road electrical grid is never going to gain wide acceptance...There are a lot more electrical outlets in the USA than there are gas stations, and a car which seamlessly switches to an engine when it runs out of battery power isn't going to bother its owners at all (except for having to fill the tank, a task that electricity would make unnecessary).

Eliminating the need for new fuel distribution chains by using a middleware fuel will bring new energy alternatives online as a practical matter. Hydrogen's where we're going to end up.That's the problem with hydrogen; it requires a new fuel distribution chain, new pipelines (the current natural gas network would need to be about 3x larger to handle the same amount of energy with hydrogen), a lot of new everything.  Further, a lot of the changeover is full replacement, not incremental; you can't just slap a hydrogen system into a car and expect to drive cross-country.  On top of this, hydrogen is very bulky to store, imposes losses of about 50% when used to store electricity, and is explosive in air to boot.  The only virtue it has is that the raw material for its production falls out of the sky.

I'm of the opinion that efficiency and compact storage are going to be important enough that hydrogen will lose out to some other medium.  The one that I think has the most potential as a transferable fuel (something you can more or less put in a tank) is metallic zinc, which can be handled as a slurry of powder in water.  There is already a company powering demonstration buses with zinc-air fuel cells.  Regeneration of metallic zinc from zinc hydroxide is another one of those loads which can be ramped up or down at will, and is thus ideally suited to the variable nature of wind and solar power.  To top it off, there is no explosion risk and the overall efficiency appears to be 60% or better.  There may be uses for which hydrogen is better, but I don't think that powering vehicles is one of them.
 

I honestly don't know what Blogger is doing to screw with the formatting there; I had two blank lines before and after each quote (it won't let me use the <blockquote> tag) but it's selectively deleting the ones after the </i> tags and not showing this in the preview.

Google should be ashamed of what they've done with Blogger.
 

The dump load part of your comment is exactly what I was talking about. What is a dump load if not wasted energy. Working from that, you shift your "dump load" to hydrogen generation and when the sun isn't shining or the wind isn't blowing, you convert the hydrogen into electricity via fuel cell. This makes power contribution from renewable generating stations more predictable. You not only have an unpredictable power generation component but also a predictable one which means the station is capable of generating some power all the time no matter what the weather and there is no time at which you ever forego power by idling turbines. There's always a customer for more hydrogen whether it's on-site fuel cells or off-site uses.
 

"The dump load part of your comment is exactly what I was talking about. What is a dump load if not wasted energy."

It can also be energy used for a lower-priority purpose.  For instance, when you run out of loads which require electricity immediately, you can switch on a heating element in a hot-water tank and displace the gas which would otherwise have been used to heat the water.

"Working from that, you shift your "dump load" to hydrogen generation and when the sun isn't shining or the wind isn't blowing, you convert the hydrogen into electricity via fuel cell."

There's where you run into the problem of conversion, transmission and storage.  For instance, if I'm trying to drive a vehicle does it make more sense to use electricity to convert ZnO to metallic zinc at 65% overall efficiency, or H2O to H2 at 50% overall efficiency?  What's going to fit better into my vehicle?  What's going to be safer to carry (liability is going to be a BIG issue for any new technology).  If I'm trying to cool a building, does it make any sense to store electricity at all instead of using it immediately to make ice and storing the ice?  It'll take a lot of melting to equal the losses of conversion to and from hydrogen.

Unless the hydrogen pathway somehow becomes much cheaper than other methods of storing and using electricity (or the hydrogen is a direct product of a process rather than produced using electricity), I think other storage media will win on the merits.  Hydrogen has a lot of negatives; the only thing hydrogen has going for it is its producibility from water alone and a certain amount of hype.

"You not only have an unpredictable power generation component but also a predictable one which means the station is capable of generating some power all the time no matter what the weather and there is no time at which you ever forego power by idling turbines. There's always a customer for more hydrogen whether it's on-site fuel cells or off-site uses."

Suppose for a moment that I was in the zinc-cycling business instead of the hydrogen-cycling business.  If I bought a KWH of surplus electricity I'd eventually be able to return something like 600-650 WH to the user compared to the hydrogen guy's 500 WH.  I could sell the zinc metal as vehicular fuel and take in ZnO in return.  That gives me an economic advantage in both markets and a size and safety advantage in the vehicular market (zinc is even harder to burn than gasoline and doesn't need high-pressure storage).

Electrolytic hydrogen gets an advantage if the storage problem is solved and electricity becomes cheap enough that the cost of the losses is negligible, or if the raw materials for other storage media become too great.  Zinc is only about a dollar a kilo right now, and the refining accounts for only a fraction of that; unless there are substantial losses in the system, I can't see zinc costs being high enough to raise the expenses of short-term cycling terribly much.  Long-term storage is a different matter, but hydrogen would have similar problems (of storage vessels rather than media).

I seem to have lost both of our points in this mess, so I'll try to recap:  Yes, I agree that some storage will be used.  It's my opinion that those uses which require storage (such as transport) will be the driver there; other things will piggyback on the economies of scale unless they have contradictory requirements.  Other than that, it makes more sense to use electricity as you get it rather than taking all the costs of storage; a truly renewable economy will have demand-side management polished better than the Hope diamond.