KU Leuven team creates solar panel that produces hydrogen from moisture in air
08 March 2019
Bioscience engineers at KU Leuven have created a solar panel that produces hydrogen gas from moisture in the air. After ten years of development, the panel can now produce 250 liters per day—a world record, according to the researchers. Twenty of these solar panels could provide electricity and heat for one family for an entire winter.
A traditional solar panel converts between 18 to 20% of the solar energy into electricity. If that electric power is used to split the water into hydrogen gas and oxygen, you lose a lot of energy. The KU Leuven bioscience engineers solved this problem by designing a solar panel of 1.6 m² that converts 15% of the sunlight straight into hydrogen gas.
It’s a unique combination of physics and chemistry. In the beginning, the efficiency was only 0.1 per cent, and barely any hydrogen molecules were formed. Today, you see them rising to the surface in bubbles. So that’s ten years of work—always making improvements, detecting problems. That’s how you get results.—Professor Johan Martens
Twenty of these panels produce enough heat and electricity to get through the winter in a thoroughly insulated house and still have power left. Add another twenty panels, and you can drive an electric car for an entire year.—KU Leuven researcher Jan Rongé
Today, most hydrogen gas is produced using oil and gas.—Grey hydrogen gas, in other words—not a big win for the climate or the environment. The KU Leuven researchers believe this is about to change.
The solar panel will be under test in Oud-Heverlee, a rural town in Flemish Brabant. The house we visit is well insulated and gets most of its power from solar panels, a solar boiler, and a heat pump. It is not connected to the gas grid. It only uses power from the grid in the winter.
Soon, 20 hydrogen gas panels will be added to this mix. If all goes well, more panels will be installed on a piece of land in the street. This will allow the other 39 families in the street to benefit from the project as well. The hydrogen gas produced in the summer will be stored and converted into electricity and heat in the winter.
The hydrogen gas produced in the summer can be stored in an underground pressure vessel until winter. One family would need about 4 cubic meters of storage—the size of a regular oil tank.
For Johan Martens, a test project like the one in Oud-Heverlee is what he and his team have been working towards for years.
We wanted to design something sustainable that is affordable and can be used practically anywhere. We’re using cheap raw materials and don't need precious metals or other expensive components.—Johan Martens
The actual cost of the hydrogen gas panels is still unknown, as the mass production is yet to start. The researchers, however, say that it should be affordable. The emphasis will not so much be on large production units, but rather on the combination of smaller, local systems. It will also require less energy-guzzling transport of energy, whether it’s gas, oil, or electricity.
Last week, Toyota announced that it wants to produce hydrogen gas with a prototype designed by Johan Martens’s team in 2014. This device is a little screen (10 cm2) that the engineers will scale up to a large panel.
This is interesting because it actually gives the critical numbers needed for an evaluation, or most of them.
We have the size of the panels.
We have the efficiency.
We know that it does not use precious metals, although I would like to know also about rare earths.
We have the size of the storage needed.
What we don't have are figures for the assumed solar isolation etc, although they could be found by a bit of digging.
For the US the energy use would be a heck of a lot more, but as against that solar isolation almost everywhere in the 48 is a lot more.
I am cautiously impressed.
Posted by: Davemart | 08 March 2019 at 03:07 AM
The claim about the size of the required storage tank is simply not credible. Hydrogen has a much lower energy density than hydrocarbons.
Posted by: Engineer-Poet | 08 March 2019 at 03:30 AM
OK, taking the solar insolation figures for Brussels from here:
And approximating an average by simply adding together the average monthly insolation figures in KWh/m2/dy then dividing by 12 I get:
32.96/12 = 2.73
Times 365 = 996.45
So 1000 KWh pa
At 15% comes to 150KWh/m2 pa in hydrogen
Or we have at 2.73KWh/m2/dy *15% =0.182KWh/m2/dy
So 1 panels at 1.6M2 = 0.2912 KWh per panel/dy
Times 20 = 6KWh per day in hydrogen equivalent, near enough for government work
That is something like 200 grams of hydrogen.
At 60mpge that comes to around 12 miles per day.
I think they are assuming some very light vehicle etc, as well as a super efficient house.
Unless of course I have messed up the arithmetic, which is always possible!
Posted by: Davemart | 08 March 2019 at 03:40 AM
Yep, messed the arithmetic
Or we have at 2.73KWh/m2/dy *15% =0.4095KWh/m2/dy
So 1 panels at 1.6M2 = 0.6552 KWh per panel/dy
Times 20 = 13KWh per day in hydrogen equivalent, near enough for government work
That is something like 400 grams of hydrogen.
At 60mpge that comes to around 24 miles per day
So 8760 miles per year, perhaps reasonable in Belgium
For the house otherwise, I assume that the efficiency would be very high with process heat being used for hot water
In any event the figures are the same 13KWh/dy equivalent for the house, assuming that I have finally got the math right.
Posted by: Davemart | 08 March 2019 at 03:51 AM
One of the things I notice on the illustration is that they have built a box structure to do the job, not a panel.
It may or may not be possible to have something thin enough to put on a rooftop.
If not, then extra space is needed, not too practical in a lot of places, and wherever they put the box equivalent it is quite bulky, never mind worrying about the hydrogen storage tank, unless this is it.
Posted by: Davemart | 08 March 2019 at 04:56 AM
A too good process to be true? May be.
Posted by: Raymondbonnate1 | 08 March 2019 at 05:34 AM
This looks like another positive step towards near future clean H2 economy? Roofs and attics could be used to generate and store enough H2 for homes HVAC and FCEVs?
Posted by: HarveyD | 08 March 2019 at 07:40 AM
I would have thought by the time this comes into widespread use, if it ever does, seasonal storage would be as hydrides, MOFCs, LOHCs or whatever, not in tanks.
Since weight is not important there are a lot of possibilities, although round trip efficiency is important.
Here is how Toshiba do it, at a hotel:
Its up and running now.
Posted by: Davemart | 08 March 2019 at 09:42 AM
The Toshiba H2One thing appears to be over-sold; it only supplies a 12-room annex to the Henn Na Hotel and I saw nothing about bigger follow-up projects. I looked up the hotel itself and the reviews were generally poor, though none of them directly referenced the hydrogen system (but the tiny cube fridge and tea pot may be related).
Posted by: Engineer-Poet | 08 March 2019 at 12:27 PM
Its a prototype, and a world first. No reasonable person would start off by supplying a city block or something.
It is there to provide data on reliability and everything else.
Posted by: Davemart | 08 March 2019 at 02:16 PM
The 4 cubic meter (Volume?) pressure tank @ pressure? = kW/h?
15% efficiency is very impressive obviously when compared to ~ 4% from (20%) solar to H2 electrolysis.
How does this design compare for reducing sunshine in winter. I.E. is it better described as a storage device buffer for night time and winter?
The difficulty as with its advantage is that is is produced specifically for storage so is really a small piece of a functional system . High efficiency P.V. nearer to 30% would way better than half the real estate required for direct consumption is especially relevant for household where battery performance is unlikely to be temperature affected .
Posted by: Arnold | 08 March 2019 at 02:40 PM
That is the whole point of the hydrogen economy, and one which critics miss by seeking to compare efficiencies of electric straight from a solar panel or whatever with a system using hydrogen.
This house will stay warm day and night winter and summer in the none too sunny latitude of Belgium.
Outside of the tropics sunshine just does not arrive when it is needed enough without considerable storage.
And there is no reason at all why this could not be combined with conventional solar panels where that would be appropriate.
Posted by: Davemart | 09 March 2019 at 01:30 AM
The uncritical acceptance of claims like this is at the root of the political deadlock between romantic idealists on the one hand and scientists and engineers on the other.
It's past time to inject some reality into this discussion. First, 4 cubic meters is an ENORMOUS tank; it's over 1000 gallons. The typical propane "pig" where I live is 500 gallons or less, and it's made of mild steel. To get decent energy capacity the hydrogen tank is going to have to operate at pressures on the order of 700 bar. Steel can fail catastrophically and compressed gas at such pressures literally forms a bomb; safety requires fussy and expensive high-strength composites like graphite fiber.
On to energy density. Propane has roughly 91,500 BTU of energy per gallon, and a 500 gallon tank is filled to 80% full for an energy capacity of 38.6 GJ. Hydrogen at 700 bar stores a mere 4.7 MJ/liter, so the 4 m³ tank will store just 18.8 GJ; that is not even half the energy held by a propane tank less than half the size. Can you actually keep a house heated and lit on so little energy? Depends on the house I suppose, but you are going to have compromises like a small refrigerator.
Next is cost of the tank itself. This 2016 projection finds that tank cost may drop to as little as $10.54/kWh at high production volume, or $1970 per 5.6 kg tank. One kg of H2 has 142 MJ/kg, and my cross-check of the numbers is roughly congruent with the cost figure given assuming that the energy is given for the raw hydrogen and not the output of whatever it is fed to. The 4 m³ tank holds 18.8 GJ or just over 5200 kWh; its cost is going to run something like $55,000!! Then you have the cost of the hydrogen production panels, the compressor, and your furnace/fuel cell/whatever.
What else could you buy with $55,000? Well, at a pessimistic $6000/kW you could get something in excess of 9 kW of nuclear power plant. Assuming it operates continuously from October 1 to March 30 (182 days), that would give you 39,312 kWh or 142 GJ. That would serve roughly 8 houses as efficient as the one running on the 4 m³ hydrogen tank... oh, and it would heat several times as many if its waste heat was utilized.
Do you understand now why I call it "hypedrogen"?
Posted by: Engineer-Poet | 10 March 2019 at 06:33 AM
I was outlining what is claimed, as it is absurd to preface every statement with ' if they can make it work'.
This technology is clearly about generating the hydrogen, nor essentially concerned with the storage which they are just going to use what is available, so the switch in focus is missing the point.
There are umpteen ways of storing the energy including feeding it into the natural gas grid.
If however the focus has to be on storage rather than what this technology is actually about at this stage, then although the provided link does not work for me since it is about a 5.6Kg tank it is clearly referring to tanks for an FCEV car.
Why anyone would imagine that that has much to do with mass storage in these very large tanks escapes me.
For a car weight is important as well as cost, and you don't need anything like the same volume of casing materials for a far larger tank, and what is more one which is buried so the surrounding ground will provide some of the structural support.
I don't have figures to hand for the cost of such large storage tanks, nor am I about to research it, since it has nothing to do with the main thrust of this article.
The important breakthrough is reaching 15% solar to hydrogen, and that is what this research is all about.
Posted by: Davemart | 10 March 2019 at 08:06 AM
BTW hydrogen for bulk storage in these sorts of quantities is stored for instance on garage forecourts at 900 bar or more, not 700bar.
If it were at 700 bar there would be no pressure differential to enable the fuelling of the tank in the car.
So the resemblance between these underground structures and the tank in a car is tenuous, and attempting to calculate costs from the latter shows not only a non-existent grasp of the technologies, but is about as sensible as trying to calculate the cost of an oil tanker from that of a 1 gallon can in the back of your car.
Those who think they know it all fall flat on their faces.
Posted by: Davemart | 10 March 2019 at 08:36 AM
Since weight is not an issue, they could store bulk H2 using adsorbants.
Posted by: SJC | 10 March 2019 at 01:08 PM
There you are wrong. The wall thickness scales linearly with the radius of the tank for a given pressure, so the ratio of wall volume to useful volume is roughly constant. You may be able to get by with cheaper fibers, e.g. high-strength glass instead of carbon, but you're going to need a great deal of it and the cheap crap won't do when you're literally confining the energy of a bomb.
This, from the guy who just said "I don't have figures to hand for the cost of such large storage tanks".
I gave you my source. You won't do any research, but you're still certain I'm wrong. This is rather amusing in a way. BTW, one of the things I uncovered but didn't use was a commercial source for steel tanks such as are used in welding. However, they have working pressures on the order of 2200 psig so I didn't consider them valid for comparison. Also... steel is subject to hydrogen embrittlement and corrosion, so really isn't suitable for use outside a controlled industrial setting.
If you had bothered to read the linked paper, you'd see that the price difference between single-tank vehicles and dual-tank vehicles favors single tanks but is quite small. The big difference comes with production volume. Larger tanks are going to require bigger winding machinery, curing ovens, etc. and will likely cost more as a consequence.
There's a reason Elon Musk chose 18650 laptop cells for his first few generations of EV batteries: they were already being produced by the billions and the cost/kWh was unbeatable at the time.
Posted by: Engineer-Poet | 10 March 2019 at 01:24 PM
Safety first is always front and centre. I wouldn't sleep well with such a tank under the house - 1kg or 100klms driving worth can fashion a very capable explosive. For every problem there should be an easy? solution.
One possible answer being developed is the ammonia storage from C.S.I.R.O.
'If it can be made to work' goes to both trust which is in short supply these days and also to the simpler question of engineering an integrated system meeting the expectations assumed from the multiple components. We need to be adding the losses rather than seeing only excellent component efficiencies.
All through the lens of the various limited options for decarbonising the world's energy services which most agree is critical for limiting climate 'catastrophe'
Some would say going nuclear would be a solution but regardless the apparently unsolveable concerns surrounding weapons and toxicity (again trust is broke), practically nuclear does not scale as well with distributed power R.E,s across developing country economies, island grids or remote locations. That is to say it has an industrial rather than human scale.
Since N.P. has been produced since 1950's is well enough understood and there are current developments and trials of say various incl S.M.R. there would appear to be not much point in 'banging on' about virtues compared to R.E. Legislation will be in the hands of regulators and small community of 'experts' rather than business and public action.
Posted by: Arnold | 10 March 2019 at 02:56 PM
No buts. Nuclear and hydro are the only proven solutions we have. Even biomass causes emissions through land-use change.
Every country that developed nuclear weapons did it BEFORE or WITHOUT having a nuclear energy program. USA, USSR, China, India, Pakistan, N. Korea... none started with nuclear electric plants. Those came later or not at all.
Hogwash. When the first oil price shock hit, France's grid was mostly powered by petroleum and made up more than 1/5 of US fossil fuel consumption for electric generation. Over the next 17 years, France had almost eliminated fossil-fired generation and the US had almost completely replaced petroleum in that role as well. France de-carbonized at about 2%/year over this period, while eliminating one major source of air pollution. France became both cleaner and greener.
What's Germany doing with its Energiewende? Its emissions reductions have stalled out. We have industrial-scale economies which require industrial-scale solutions. No country has ever decarbonized with unreliable energy of any kind, and given the lack of serious engineering proposals they never will. "Renewables" are a front for the fossil fuel industry, to keep them in business forever.
One would expect the virtue that it's the only proven solution would be worth something, but with idiots who value feelings over facts and pursue romantic visions that cannot be turned into engineering reality we are headed for some very bad times indeed.
Posted by: Engineer-Poet | 10 March 2019 at 07:41 PM
In the early 1970s, plutonium breeder reactors which produced essentially free fuel were thought to be the solution for providing more clean, economical energy to a growing U.S. economy. The Carter administration scotched the idea because it is cheap to enrich fuel grade plutonium to weapons grade plutonium, and studies even then identified terrorists getting their hands on fuel as the most significant risk to broadly-deployed nuclear power. This left the USA with the need to enrich Uranium, which is much harder to enrich from fuel grade to weapons grade, on a much larger scale and at less cost than was possible with the existing gas diffusion facilities. I helped build the centrifuges that were to provide economical enriched Uranium fuel at scale. By the mid 1980s, the clatter of later-morphing-into-"green" protests decimated the voice of reason, the enriching program was dumped, nukes in America were shouted into history, and coal plants happily and with barely a whimper of protest moved in to fill the void. Sometimes you get what you ask for.
Asking for Hydrogen? Better find out what you're really asking for.
Posted by: LUH3417 | 11 March 2019 at 07:06 PM
And that was correct, even if the Clinch River reactor was the wrong vehicle. As of some years ago the USA was sitting on enough depleted uranium to run the whole country for several centuries, and has no doubt accumulated more since then.
Your timeline is wrong. US advanced reactor research continued until 1994 when John Kerry was the key vote in favor of cancelling all advanced nuclear reactor research in the USA. Carter had been gone for 13 years at that point. Clinton was president.
Your physics is wrong. "Enrichment" of plutonium is grossly impractical given that the bomb-grade isotope, Pu-239, is in the middle of heat-producing Pu-238 and extremely high spontaneous fission isotope Pu-240. Isotopic separation by weight will give you one of the extremes, not the middle. On top of this, the plutonium is fairly radioactive and the enrichment system will have to be decontaminated for maintenance.
Used (not "spent") nuclear fuel has so much gamma radiation from Cs-137 and Sr-90 that it is effectively impossible to steal; anyone making off with it without many tons of shielding would die from radiation exposure before they could get far. To do that at all, they would have to remove spent fuel from cooling pools or welded dry casks. Given that both are well-guarded, this is effectively impossible in the west. Anyone trying this would be captured or killed before they could do squat.
Terrorists are not stupid enough to have failed to figure this out. This is why not one terrorist action in the WORLD has gone after nuclear fuel, but "environmentalists" harp on it day in and day out (their mission is to enstupidate the public and continue the use of fossil fuels indefinitely).
Posted by: Engineer-Poet | 11 March 2019 at 07:55 PM
I'm surprised they are even talking breeder reactors - didn't they use much up in their depleted uranium spray in middle east aid?
But what would he know? While E.P. is busy poo pooing and bellicose blustering the graph below shows what was actually happening.
Germany has a target of reaching 65 per cent renewable energy for its electricity network by 2040, but it is already capable of reaching such levels for shorter periods of time, as illustrated last week.
For “week 10” of 2019, the week finishing on March 10 (Sunday), Germany sourced 64.8 per cent of its electricity generation from renewables.
As this graph below illustrates, the bulk came from wind (48.4 per cent), with solar contributing 5.1 per cent, biomass 7.6 per cent and hydro 3.5 per cent.
Throw in nuclear, and the share of zero emissions electricity sources reached 77.7 per cent for the whole week, in the biggest economy in Europe and one of the biggest in the world.
Posted by: Arnold | 11 March 2019 at 10:30 PM
"As of June 2000, the United States depleted uranium stocks (including USEC tails) contain about 480 000 tU." (source) There's another 70,000 tons or so of used LWR fuel, the bulk of which is slightly-enriched uranium (~1% U-235) and 0.8% or so is transuranics including plutonium.
A number of people are. They're called "realists".
65% falls far short of what is required to halt CO2 concentrations from increasing; we need at least 80% across ALL sectors. Aiming for 65% in the electric sector alone is planning to fail.
While the Green fanatics in Germany are planning to fail, France already proved that nuclear can provide 80+% decarbonization of a formerly-fossil dominant grid without actually aiming for that as a goal; it was an unplanned byproduct of replacing imported petroleum. What Germany will never be able to do with "renewables", France accomplished by accident.
As for "why even talk breeder reactors", several reasons:
Posted by: Engineer-Poet | 12 March 2019 at 09:58 AM
Here's a link to a 1977 NYT report of Carter cutting breeder research funding:
And here's a link to a 1997 brief summary of the previous 2 decades of nuclear policy:
Note the references to reactor plutonium use in weapons, including this:
"The facts were classified SECRET, but the U. S. had actually exploded a device made of "reactor-grade plutonium" at the Nevada Test site in 1962."
As you observed, "...not one terrorist action has gone after nuclear fuel...".
Did the anti-breeder policy begun by Carter (and anti-proliferation in general) and finished by Clinton help keep the fuel out of bad hands?
Hard to say.
But its easy to say that nuclear energy has proved to be far safer by any rational measure than energy from any fossil fuel.
Posted by: LUH3417 | 12 March 2019 at 12:52 PM
REs are gaining ground at a very fast rate, as cleaner energy sources and will continue to do so, at least until new types of more efficient lower cost NPPs can be built and operated, sometime between 2060 and 2100?
Hydro plants, with very large water reservoirs (as big as the 5 Great Lakes) can be and will be used to store surplus Wind and Solar energy for extended periods. Wind/Solar will become the main energy sources to charge your electrified vehicles and feed your HVACs. Hydro will be used (as a back up e-energy source) during peak demand periods only, much the same as you would use batteries and/or stored H2.
Hydro and Wind energies potentials are often colocated and can be mated easily to produce clean lowwer cost energy 24/7 and 365 days/year.
Posted by: HarveyD | 12 March 2019 at 01:40 PM