6 ms·

Some background and context from someone tangentially related to the field: 1. The overall idea here is to take an intermittent energy source (e.g. solar power

Some background and context from someone tangentially related to the field: 1. The overall idea here is to take an intermittent energy source (e.g. solar power) and "store" it as chemical fuel, in this case hydrogen and oxygen. This is what plants do, and we can also view fossil fuels as resulting from the "storage" of millions of years of solar energy. Note also that you get the water back when you burn the hydrogen, so there is no net consumption of water, it's just a carrier. 2. While you can split water without a catalyst, most of the energy gets wasted as heat, so this is not a great way to go if you're trying to do energy storage. 3. Efficient catalysts exist for this reaction, but they are based on rare and expensive metals, typically Pd, Pt, and Ir. As a result, there has been a search for catalysts involving "first-row" metals such as Fe, Co, Ni, etc. 4. There are variety of metrics for an electrocatalyst (efficiency, stability, cost, etc), but it's a fair bet that if this were significantly better than state-of-the-art, it would be in Science or Nature rather than PNAS.

dnautics 9y ago
Beyond the expensive of the metals another problem is duty cycles. Most transition metal catalysis is oxygen sensitive, and it seems like for some reason the first step of splitting water is creating oxygen. Plants go through great lengths to separate oxygen synthesis (photosystem II) from electron consumption. Most hydrogen production in lower organisms (like e coli) occurs entirely in anoxic conditions. Engineered systems for generating hydrogen via algae typically are temporally segregated (harvest light during day, produce hydrogen at night) which defeats the purpose and is also chemically steppy (carbohydrate intermediates).
- novaleaf 9y ago
  > and it seems like for some reason the first step of splitting water is creating oxygen good laugh on that :)
  - nine_k 9y ago
    Consider a split to H + OH as a first step, as opposed to H + H + O.
    - dnautics 9y ago
      What exactly are you proposing? What do you do with the OH?
      - nine_k 9y ago
        This is a possible first step of a water-splitting reaction that does not produce oxygen.
        dnautics 9y ago
        I suggest going back to your chemistry text and reviewing mass balance, conservation of matter, and balancing equations. What is that OH? Is it hydroxide radical? Hydroxide? Hydroxyl radical? Where does it go? Does it recombine to make hydrogen peroxide? Btw... To make oxygen from water the first part of the first step is to split it into H and HO. You can't really break two bonds simultaneously, or anyways it's equivalent to doing them stepwise by the principle of microscopic reversibility.
logicallee 9y ago
can you compare storing an intermittent energy source by means of a chemical fuel, hydrogen and oxygen, as compared with in a battery? Just compare and contrast every aspect that matters. Just to be clear, this is (in effect) a battery, right? So what are its characteristics as compared with, say, lithium ion batteries. (I am particularly interested in weight and in number of duty cycles, which sounds like it's "unlimited" as opposed to lithium ion which is really not that many cycles, right?) I'm also far outside the field, just interested. Thanks!
- epistasis 9y ago
  "Cycles" don't really make any sense here, but you can get some estimates of costs and efficiencies here: https://en.wikipedia.org/wiki/Power_to_gas https://en.wikipedia.org/wiki/Power_to_gas Lithium ion batteries for grid storage typically use an NMC chemistry, and standard warranties are for 10 years of daily cycling.
  - logicallee 9y ago
    (you might want to mention you're not OP in cases like this, as I addressed my comment specifically to them and almost thought I was still talking to them.) of course, I appreciate your reply too. epistasis, why do you say "cycles don't make sense" -- is it because it's "infinite" or because it can't be run as a fuel cell immediately in the same place? Like you can't go back and forth immediately? I read most of the article you linked (which had a couple of occurrences of the term 'fuel cell' - not many - therefore leading me to be confused as to why you said what you just said). I am talking about using it as a fuel cell, same as a battery. If daily cycling means 1 cycle per day, then 10 years means 3650 cycles. How does "power to gas" compare? Infinite cycles?
    - epistasis 9y ago
      Apologies for the confusion. You were asking for an awful lot, in way that didn't make much sense, so I thought I would chip in a bit. OP was talking about burning the hydrogen, as was I. Burning is not typically the term used with what goes on in fuel cells. A battery has a fixed capacity attached to it, it has a fixed power to energy ratio. The component that stores the energy is directly attached to the part that discharges the energy. This is what makes "cycles" make sense. With power to gas, or using fuel cells to consume energy from the process described in the article, the component that stores energy is not directly connected to the part that consumes energy. It's not necessary that they even have a fixed fuel stores attached to them, they could just be hooked up to a pipeline. The rate at which parts wear out would better be described in terms of total energy stored or consumed. That's why I mentioned the cost estimates (€0.10/kWh for hydrogen in that article), because it's perhaps the best way to compare. The rapid drop of lithium ion grid storage has put it at $250/kWh capacity with 3650 cycles, which is ~$0.07/kWh before accounting for taxes, maintenance, siting, etc., which may or may not be included in that above estimate for hydrogen. Both of these numbers blow me away by how small they are. We're in for a wild ride on tech changes over the next few years...

dnautics 9y ago

Beyond the expensive of the metals another problem is duty cycles. Most transition metal catalysis is oxygen sensitive, and it seems like for some reason the first step of splitting water is creating oxygen. Plants go through great lengths to separate oxygen synthesis (photosystem II) from electron consumption. Most hydrogen production in lower organisms (like e coli) occurs entirely in anoxic conditions. Engineered systems for generating hydrogen via algae typically are temporally segregated (harvest light during day, produce hydrogen at night) which defeats the purpose and is also chemically steppy (carbohydrate intermediates).

novaleaf 9y ago
> and it seems like for some reason the first step of splitting water is creating oxygen good laugh on that :)
- nine_k 9y ago
  Consider a split to H + OH as a first step, as opposed to H + H + O.
  - dnautics 9y ago
    What exactly are you proposing? What do you do with the OH?
    - nine_k 9y ago
      This is a possible first step of a water-splitting reaction that does not produce oxygen.
      - dnautics 9y ago
        I suggest going back to your chemistry text and reviewing mass balance, conservation of matter, and balancing equations. What is that OH? Is it hydroxide radical? Hydroxide? Hydroxyl radical? Where does it go? Does it recombine to make hydrogen peroxide? Btw... To make oxygen from water the first part of the first step is to split it into H and HO. You can't really break two bonds simultaneously, or anyways it's equivalent to doing them stepwise by the principle of microscopic reversibility.

logicallee 9y ago

can you compare storing an intermittent energy source by means of a chemical fuel, hydrogen and oxygen, as compared with in a battery? Just compare and contrast every aspect that matters. Just to be clear, this is (in effect) a battery, right? So what are its characteristics as compared with, say, lithium ion batteries. (I am particularly interested in weight and in number of duty cycles, which sounds like it's "unlimited" as opposed to lithium ion which is really not that many cycles, right?) I'm also far outside the field, just interested. Thanks!

epistasis 9y ago
"Cycles" don't really make any sense here, but you can get some estimates of costs and efficiencies here: https://en.wikipedia.org/wiki/Power_to_gas https://en.wikipedia.org/wiki/Power_to_gas Lithium ion batteries for grid storage typically use an NMC chemistry, and standard warranties are for 10 years of daily cycling.
- logicallee 9y ago
  (you might want to mention you're not OP in cases like this, as I addressed my comment specifically to them and almost thought I was still talking to them.) of course, I appreciate your reply too. epistasis, why do you say "cycles don't make sense" -- is it because it's "infinite" or because it can't be run as a fuel cell immediately in the same place? Like you can't go back and forth immediately? I read most of the article you linked (which had a couple of occurrences of the term 'fuel cell' - not many - therefore leading me to be confused as to why you said what you just said). I am talking about using it as a fuel cell, same as a battery. If daily cycling means 1 cycle per day, then 10 years means 3650 cycles. How does "power to gas" compare? Infinite cycles?
  - epistasis 9y ago
    Apologies for the confusion. You were asking for an awful lot, in way that didn't make much sense, so I thought I would chip in a bit. OP was talking about burning the hydrogen, as was I. Burning is not typically the term used with what goes on in fuel cells. A battery has a fixed capacity attached to it, it has a fixed power to energy ratio. The component that stores the energy is directly attached to the part that discharges the energy. This is what makes "cycles" make sense. With power to gas, or using fuel cells to consume energy from the process described in the article, the component that stores energy is not directly connected to the part that consumes energy. It's not necessary that they even have a fixed fuel stores attached to them, they could just be hooked up to a pipeline. The rate at which parts wear out would better be described in terms of total energy stored or consumed. That's why I mentioned the cost estimates (€0.10/kWh for hydrogen in that article), because it's perhaps the best way to compare. The rapid drop of lithium ion grid storage has put it at $250/kWh capacity with 3650 cycles, which is ~$0.07/kWh before accounting for taxes, maintenance, siting, etc., which may or may not be included in that above estimate for hydrogen. Both of these numbers blow me away by how small they are. We're in for a wild ride on tech changes over the next few years...