4 ms·

OK, but nobody has enough batteries to store enough for the winter. Overnight? Sure. Winter? That's a whole different deal.

seanmcdirmid 1y ago
First, the only reason solar isn't so common in Alaska is that hydro is even easier to get power from, so your outback communities will often use small scale hydro instead of solar (especially in southeast Alaska). Communities in the north will use diesel generators in the winter (nothing else is viable). Again, I assume you are talking about off grid communities, which is basically all of them except a few cities (and most cities have their own grids disconnected from the rest, especially Southeast Alaska).
cachius 1y ago
That's why you generate hydrogen during summer for gas power plants during winter.
- somanyphotons 1y ago
  What mass of hydrogen is needed per house to last a winter?
  - zparky 1y ago
    rough math: 1000 kwh / mo / house, ~30kwh/kg hydrogen so 30kg H per mo per house. idk how long winters would be, 8 months is 240kg of hydrogen, which if compressed to 10 bar is roughly 300 cubic meters of storage. kinda a lot of space. compressed to 100 bar is like 10kg/m3 which sounds more manageable
    - Retric 1y ago
      Round trip efficiency on hydrogen is horrible. Local hydrogen production could make sense because importing fuel into remote off grid communities is extremely expensive. Rather than building 10x as much solar in the north + battery systems + winter hydrogen storage etc long distance HVDC to cities and the surrounding grid just makes so much more sense. Even better because the state is huge and the population is tiny they can go nearly 100% hydro. Where batteries could be useful is operating those long distance power lines at nearly 100% 24/7 then load shifting via batteries to match local demand.
      - pfdietz 1y ago
        > Round trip efficiency on hydrogen is horrible. For seasonal storage, round trip efficiency is mostly irrelevant; the relevant metric is capex per unit of stored energy.
        Retric 1y ago
        That could be generally true, but it’s not true in this specific instance. A panel in Alaska only collects so much sunlight over the summer before considering efficiency losses from Hydrogen. It would require buying panels that effectively get ~1 month of use over the entire year due to efficiency losses + limited gathering period, and solar isn’t that cheap. So in Alaska you’re just better off only using panels directly in the summer which at least provide several months of electricity per year. In say Texas on the other hand you get energy from a panel year round so a marginal panel purchased to generate hydrogen at say 20% round trip efficiency gets 30% * 9 months + say 70% of average production for the 3 winter months = 4.8 months of winter electricity per year. Of course you also need to pay for the hydrogen generating machine and the hydrogen burning device, but that’s not necessarily problematic.
        pfdietz 1y ago
        If the alternative were, say, diesel, I think a RTE of 40% (which you might get with hydrogen) would be fine in this case, if the capex of the storage system is low enough. It is certainly the case that hydrogen would be better than batteries for this storage use case in Alaska.
        Retric 1y ago
        If the baseline is ~2c/kWh solar in a good location and they are forced to buy 10x as much solar panels to cover winter use they are now spending ~20c/kWh on solar panels but on top of that they also need to pay for hydrogen generating equipment which only gets used for a few months a year, hydrogen storage equipment, and hydrogen specific generators plus presumably a backup diesel or gas based generator + storage system. In 20 years it might make sense but today green hydrogen is several times more expensive than gas even when you can use cheaper electricity, can make use of the equipment year round, and have the benefit of larger economies of scale. Even if the goal is completely about climate change locating that same equipment in the lower 48 states is just a much better idea.
- downrightmike 1y ago
  Hydrogen containment isn't long term, every atom we use to make hydrogen tanks are much larger than hydrogen, so they leak no matter what.
  - pfdietz 1y ago
    It would be stored underground at scale, not in tanks.
    - downrightmike 1y ago
      And tell me again what the ground is made of
      - pfdietz 1y ago
        Ideally, salt. Deeply underground, solution mined to make large cavities. This is a preferred method for storing natural gas. Failing that, deep aquifers (not potable ones; saline water is fine). If the basement rock is close to the surface and is crystalline, it probably involves deep mining to form cavities, which would raise the capex by maybe an order of magnitude. Other options could become cheaper then, say storing ammonia. This is for large scale storage, of course, not for individual residences.
- pfdietz 1y ago
  I used to think hydrogen was the front runner for this storage use case, but as I've pointed out previously on HN it's looking like ultra low capex thermal storage will be superior. https://www.orcasciences.com/articles/standard-thermal-copy https://www.orcasciences.com/articles/standard-thermal-copy
namibj 1y ago
In the lower 48 you'd just size the solar to handle winter and find a way to use summer electricity, like electro winning metal from ore or running low-capex high-power chemical synthesis steps like making ammonia without natural gas. Having energy cost related scheduled (winter) downtime gives the plants proper maintenance windows. With free power but only during surplus peaks in summer when the grid can't transmit a large utility solar farm's entire production, and the day/night/weekday time shifting batteries are also already fully active, you could (looks like the math checks out) electrolytically refine iron ore into iron metal (for later smelting in an arc furnace) just about cost-competitively with (coal-fired) blast furnace operation. The key is to skip most overhead by operating them only to eat otherwise-curtailed production and connecting them to the DC bus between the MPPT and the grid inverter (same as the day/night shifting battery).
skrause 1y ago
You always combine solar power with wind power, that already solves more than 90% of the storage problem. In places like Europe the darker winter is usually quite windy. For example, last Sunday Germany covered more than 100% of its own power load with renewables even though winter is approaching. Only a small part of that was solar power, most electricity was generated by wind turbines: https://www.energy-charts.info/charts/power/chart.htm?l=en&c=DE&week=43 https://www.energy-charts.info/charts/power/chart.htm?l=en&c...
- pfdietz 1y ago
  Usually that works, but not always. It's the last part that's the killer. When you optimize against historical weather data in Europe to produce steady power at minimum cost, a wind/solar/battery/hydrogen system cuts the cost in half compared to wind/solar/battery. As I said elsewhere I'm thinking ultra low capex thermal storage will edge out hydrogen here, though.

seanmcdirmid 1y ago

First, the only reason solar isn't so common in Alaska is that hydro is even easier to get power from, so your outback communities will often use small scale hydro instead of solar (especially in southeast Alaska). Communities in the north will use diesel generators in the winter (nothing else is viable). Again, I assume you are talking about off grid communities, which is basically all of them except a few cities (and most cities have their own grids disconnected from the rest, especially Southeast Alaska).

cachius 1y ago

That's why you generate hydrogen during summer for gas power plants during winter.

somanyphotons 1y ago
What mass of hydrogen is needed per house to last a winter?
- zparky 1y ago
  rough math: 1000 kwh / mo / house, ~30kwh/kg hydrogen so 30kg H per mo per house. idk how long winters would be, 8 months is 240kg of hydrogen, which if compressed to 10 bar is roughly 300 cubic meters of storage. kinda a lot of space. compressed to 100 bar is like 10kg/m3 which sounds more manageable
  - Retric 1y ago
    Round trip efficiency on hydrogen is horrible. Local hydrogen production could make sense because importing fuel into remote off grid communities is extremely expensive. Rather than building 10x as much solar in the north + battery systems + winter hydrogen storage etc long distance HVDC to cities and the surrounding grid just makes so much more sense. Even better because the state is huge and the population is tiny they can go nearly 100% hydro. Where batteries could be useful is operating those long distance power lines at nearly 100% 24/7 then load shifting via batteries to match local demand.
    - pfdietz 1y ago
      > Round trip efficiency on hydrogen is horrible. For seasonal storage, round trip efficiency is mostly irrelevant; the relevant metric is capex per unit of stored energy.
      - Retric 1y ago
        That could be generally true, but it’s not true in this specific instance. A panel in Alaska only collects so much sunlight over the summer before considering efficiency losses from Hydrogen. It would require buying panels that effectively get ~1 month of use over the entire year due to efficiency losses + limited gathering period, and solar isn’t that cheap. So in Alaska you’re just better off only using panels directly in the summer which at least provide several months of electricity per year. In say Texas on the other hand you get energy from a panel year round so a marginal panel purchased to generate hydrogen at say 20% round trip efficiency gets 30% * 9 months + say 70% of average production for the 3 winter months = 4.8 months of winter electricity per year. Of course you also need to pay for the hydrogen generating machine and the hydrogen burning device, but that’s not necessarily problematic.
        pfdietz 1y ago
        If the alternative were, say, diesel, I think a RTE of 40% (which you might get with hydrogen) would be fine in this case, if the capex of the storage system is low enough. It is certainly the case that hydrogen would be better than batteries for this storage use case in Alaska.
        Retric 1y ago
        If the baseline is ~2c/kWh solar in a good location and they are forced to buy 10x as much solar panels to cover winter use they are now spending ~20c/kWh on solar panels but on top of that they also need to pay for hydrogen generating equipment which only gets used for a few months a year, hydrogen storage equipment, and hydrogen specific generators plus presumably a backup diesel or gas based generator + storage system. In 20 years it might make sense but today green hydrogen is several times more expensive than gas even when you can use cheaper electricity, can make use of the equipment year round, and have the benefit of larger economies of scale. Even if the goal is completely about climate change locating that same equipment in the lower 48 states is just a much better idea.
downrightmike 1y ago
Hydrogen containment isn't long term, every atom we use to make hydrogen tanks are much larger than hydrogen, so they leak no matter what.
- pfdietz 1y ago
  It would be stored underground at scale, not in tanks.
  - downrightmike 1y ago
    And tell me again what the ground is made of
    - pfdietz 1y ago
      Ideally, salt. Deeply underground, solution mined to make large cavities. This is a preferred method for storing natural gas. Failing that, deep aquifers (not potable ones; saline water is fine). If the basement rock is close to the surface and is crystalline, it probably involves deep mining to form cavities, which would raise the capex by maybe an order of magnitude. Other options could become cheaper then, say storing ammonia. This is for large scale storage, of course, not for individual residences.
pfdietz 1y ago
I used to think hydrogen was the front runner for this storage use case, but as I've pointed out previously on HN it's looking like ultra low capex thermal storage will be superior. https://www.orcasciences.com/articles/standard-thermal-copy https://www.orcasciences.com/articles/standard-thermal-copy

namibj 1y ago

In the lower 48 you'd just size the solar to handle winter and find a way to use summer electricity, like electro winning metal from ore or running low-capex high-power chemical synthesis steps like making ammonia without natural gas. Having energy cost related scheduled (winter) downtime gives the plants proper maintenance windows. With free power but only during surplus peaks in summer when the grid can't transmit a large utility solar farm's entire production, and the day/night/weekday time shifting batteries are also already fully active, you could (looks like the math checks out) electrolytically refine iron ore into iron metal (for later smelting in an arc furnace) just about cost-competitively with (coal-fired) blast furnace operation. The key is to skip most overhead by operating them only to eat otherwise-curtailed production and connecting them to the DC bus between the MPPT and the grid inverter (same as the day/night shifting battery).

skrause 1y ago

You always combine solar power with wind power, that already solves more than 90% of the storage problem. In places like Europe the darker winter is usually quite windy. For example, last Sunday Germany covered more than 100% of its own power load with renewables even though winter is approaching. Only a small part of that was solar power, most electricity was generated by wind turbines: https://www.energy-charts.info/charts/power/chart.htm?l=en&c=DE&week=43 https://www.energy-charts.info/charts/power/chart.htm?l=en&c...

pfdietz 1y ago
Usually that works, but not always. It's the last part that's the killer. When you optimize against historical weather data in Europe to produce steady power at minimum cost, a wind/solar/battery/hydrogen system cuts the cost in half compared to wind/solar/battery. As I said elsewhere I'm thinking ultra low capex thermal storage will edge out hydrogen here, though.