Irina Slav writes on the Oilprice website about the water needed to produce green hydrogen. What are your views?
The Green Hydrogen Problem That No One Is Talking About
Gigawatt upon gigawatt of green hydrogen capacity is being planned across Europe, Asia, and Australia. According to proponents of the technology, green hydrogen – the kind produced through electrolysis powered by solar, wind, and other renewable energy sources – is the best way to decarbonize heavy polluter industries. There is much talk about the falling costs of solar and wind and how they will make green hydrogen viable very soon. What nobody seems to want to talk about is water. Electrolysis is the process of breaking down water into its constituent elements – hydrogen and oxygen – using an electric current. The process is performed in an installation called an electrolyzer. When hydrogen advocates talk about the bright future of the technology, they focus on the costs associated with the electricity needed for the electrolysis. But electrolysis, besides electricity, needs water.
Tons of water – literally.
One industry source told Oilprice that the production of one ton of hydrogen through electrolysis required an average of nine tons of water. But to get these nine tons of water, it would not be enough to just divert a nearby river. The water that the electrolyzer breaks down into constituent elements needs to be purified
The process of water purification, for its part, is rather wasteful. According to the same source, water treatment systems typically require some two tons of impure water to produce one ton of purified water. In other words, one ton of hydrogen actually needs not nine but 18 tons of water. Accounting for losses, the ratio is closer to 20 tons of water for every 1 ton of hydrogen.
Speaking of water purification, organic chemists explain that the simplest way to do this is by distilling it. This method is cheap because it only needs electricity, but it is not fast. Regarding the electricity cost, distilling a liter of water requires 2.58 megajoules of energy, which translates into 0.717 kWh, on average.
This doesn’t look like much at first glance, but let’s see how things look on a larger scale. Germany is the country with the most ambitious plans for green hydrogen. The cost of electricity for non-household users in Germany was an average of $0.19 (0.16 euro) per kWh as of last year. At a power consumption rate of 0.717 kWh, the distillation of a liter of water, then, would cost $0.14 (0.1147 euro). For a ton of water, that would be $135.14 (114.72 euro).
However, electrolysis needs as much as 18 tons of water – not accounting for losses during the process – to produce one ton of hydrogen. That means that the cost of water purification for the production of a ton of hydrogen would be $2,432 (2,065 euro). This is based on the assumption that the water would be purified using the cheapest method available. There are other – much faster – methods, but they are also costlier, involving ion exchange resins or molecular sieves.
Other alternatives to distillation, according to chemists, are unreliable at this point.
So, providing the right kind of water for hydrolysis costs money, and while $2,400 per ton of hydrogen may not sound like much, the cost of purifying water is not the only water-related expense in the technology that seeks to make hydrogen from renewable sources. Besides being pure, the water to be fed into an electrolyzer has to be transported to it.
Transporting tons upon tons of water to the site of an electrolyzer means more expenses for the logistics.
To cut these, it would make sense to pick a site where water is abundant, such as by a river or the sea, or, alternatively, close to a water treatment facility. This puts a limit on the choice of locations suitable for large-scale electrolyzers. But since an electrolyzer, to be green, needs to be powered by renewable energy, it would also need to be in proximity to a solar or a wind farm. These, as we know, cannot be built just anywhere; solar farms are most cost-effective in places with a lot of sunshine, and wind farms perform best in places where there is sufficient wind.
Needless to say, these places are not, as a rule, close to waterways, except offshore wind, which seems perfect for the production of green hydrogen. Unfortunately, offshore wind is also the costliest form of the three renewable sources – solar, onshore wind, and offshore wind – normally mentioned in the context of green hydrogen production. According to Rystad Energy, the capital costs of an offshore farm are twice as high as those of its onshore counterpart and four times as high as the costs of a comparable solar installation.
Not all costs associated with the production of hydrogen from renewable energy sources are the costs of those renewable energy sources. Water is the commodity that the process needs, and it is a little odd that nobody seems willing to discuss the costs of water, including the European Commission’s Green Deal Team.
Perhaps the cost of water supply, storage, and purification is negligible compared with other costs that need to be addressed first. Yet it is an actual cost that should be added to the total when estimating how far the technology of producing hydrogen from renewable electricity has progressed and how viable it has become.
For now, experts appear to be unanimous that it is not viable – not without significant government support.
12 thoughts on “Green hydrogen – nobody seems to want to talk about water”
Indeed, this is a very good article !…it is very much welcome if attached to many other problems which H2 produced from water electrolysis…besides efficiency of H2 cycle, there is also the efficiency of gas turbines on H2…since their compressors need much more power to supply the right pressure for GT…anyway, for Romanian “experts” dreaming to replace 3500 MW on lignite with H2, this means drying up Danube to use the wind farms in Dobrogea…or, taking the water from cooling nuclear plants in Cernavoda, or, taking the water from irrigations…etc..etc….water is a much more precious “commodity”, becoming more and more scarce, to electrolyse it for something which doesn’t pay of economically…even for “green energy”
It is something similar to Saudi Arabia case, where, also some smart “experts” proposed “green hydrogen” from their drinking water….which is produced from huge CHP desalination plants firing…guess what…crude, heavy fuel oil, cracked fuel ?!
Thanks so much Catalin. Excellent comment
A somewhat speculative article. I used to run, amongst other things a water purification plant (Cation & Anion columns using resin beads etc) input was mains water. This needed regeneration from time to time – typically producing 50 tonnes of de-i water and needing half a tonne of water to regenerate. This is some distance from the figures given in the article.
Energy is the main input cost for an electrolyser, accounting for between 74 to 80% of cost. The rest is other Opex and capex. This reality thus validates a comment in the article: “Perhaps the cost of water supply, storage, and purification is negligible compared with other costs that need to be addressed first” – absolutely correct.
As for siting of electrolysers, I note that Naturgy in Spain will build 60MW of electrolysers on the site of an old thermal station (coal) which used river water for cooling. Power will be provided by 400MW of PV – located at the old open cast coal mine nearby. From a water balance point of view, there will be little change, indeed less water will need to be extracted from the river than when the coal station was running.
As the writer may be aware, development of electrolysers powered by off-shore wind is well advanced, purifying sea-water is seen as a trivial task. There are also PHS developments. EdF in Chile will combine PV and pumped storage to drive reverse osmosis systems to produce low cost freshwater. No reason why this could not be deployed in Spain/Andalucia – and the water then used for a range of applications including H2 production.
Given the above, the comment in the article: “Other alternatives to distillation, according to chemists, are unreliable at this point” – seems rather strange since reverse osmosis has been around for many many years. & sure, perhaps the water would then need to be de-ied but that is not a big cost.
The concluding comment in the article: “For now, experts appear to be unanimous that it is not viable – not without significant government support”… is rather at odds with realities occurring right now. The Hydeal project (and the Naturgy above) being but two among many. Will they have gov’ support? probably a bit but “significant” don’t think so.
Mike, in very very general terms what you are saying is right, but, the simple “common sense” leads me to the following:
1. to consume electricity (even renewable) in an electrolysis process @60-70% efficiency to generate a “green fuel” (hydrogen), which then, to burn it @30% efficiency (Steam Rankine cycle), or max 50% efficiency (combined cycle), finally to end up at a net 30-35% efficiency, this triggers at least “an eyebrow”…let’s forget about minor things of technology itself, corrosion, security…etc. etc. and technology costs for GT to use straight H2…otherwise, the GT compressors, will have to use a huge part of power generated to compress the H2 (which is very different than CH4)
2. the water, indeed may not be a very significant item cost or quantitywise, BUT, this is a commodity which is becoming more and more scarce…climate change is worsening for the moment, isn’t it !?…so, this cleaner (and cheaper water) is diverted from Hydroplants, or, from irrigations, which values could be assesed
3. RO for desalination….indeed, this one is on the market since long time ago, but, when considering that it is required ~4.5 KWh electricity/m3 of desalinated water, then, we have to look for this “green electricity”(again !!) to produce water, to generate further Hydrogen (green)
These are things which could make anybody have serious “second thoughts” on hydrogen for power generation.
The direct ideal use of excess green electricity is to use it directly in the heat market…which is predominanlty fossil fuelled powered and….it is much bigger than electricity market worldwide !…anywhere
Mr Catlin, indeed, H2made by electrolysers and used in power generation does not, using simple efficiency measures, appear to make much sense.
However, in any power system, the more renewables installed, the more excess electricity is produced. A very rough metric is that once RES contributes more than 30% of load, excess electricity from RES increases in a non-linear fashion. By the time RES accounts for 80% of load, RES generation is around 120% of load over some period of time (weeks, months etc). I model this, it’s my day job. These are not assertions, they are realities and apply as much to countries as they do to regions.
All this excess RES electricity needs to go somewhere. Demand response works for periods of <<24hrs, batteries lack scale when 100sGWh (or perhaps single figure TWh) need to be stored which leaves electrolysers and a repurposed gas network. As for pushing H2 through CCGTs or OCGTs – not difficult, tech exists. But you would only do that in extremis – perhaps in the winter period, probably a couple of times over multi-day periods per month. Again, this is something I & my colleagues look at and model every day.
At a more mundane level, and speaking as an ex-DNO power systems engineer, the existing distribution networks covering the EU and the UK will not, with some minor exceptions, support the widespread use of heat pumps (or indeed, large-scale trickle charging of EVs) – key word “widespread”. I have the metrics covering this and the model delivering these metrics is very robust. So in conclusion, I have no argument about H2 round trip through CCGTs, but this is a narrow focus and avoids the bigger “how do we de-carb energy” picture.
Thank you Mike,
That’s why I love debating the arguments !
It takes 2.38 US gallons of water to make 1 kg of green hydrogen. There is plenty of water to perform electrolysis (water splitting) to make hydrogen. Water Vapor is the largest greenhouse gas in the atmosphere. It accumulates from increased evaporation due to the greenhouse effect. They even have a catchy new name for it: Atmospheric Rivers! This water falls back to earth every 9-10 days as rain, snow, or sleet. When you burn hydrogen in a fuel cell its main byproducts are heat and pure water. So in essence this is a closed loop system – then hit the repeat button!
These issues (including water) we have discussed it long time ago and, indeed, this is a legitimate issue !
There are smart guys thinking to take water from Danube to produce H2, from the largest Hydro plant in EU-Portile de Fier, store H2 and produce power back in the system with the fancy H2 Gas Turbines (hopefully manufactured by Siemens)…the stupidity is present: better reduce the hydro generation available to produce H2 for GT @50% effcy
At least for Romania: we have under analysis a site for Hydro plant with reverse pumping/generating units of 1000 MW net…with the existing hydro already and other facilities, if we add these 1000 MW, we can forget about H2 for the next 20-30 years…and leave the matter to other fans of H2…which is a stupid thing anyway, in my view, even for considering balancing the power systems !…there are so many other ways to achieve this !…also, I’m speaking “from the shoes” of a power utility old dog of more than 22 years…
Thanks so much for this thoughtful comment, Catalin
…and no comparison numbers to water use on Blue Hydrogen production…and associated costs to capture, liquify the carbon, and dangerously transport it to sequestration water costs?