9 ms·

Around here I have a backyard washing machine, a backyard grill, a backyard clothesline, and my neighbors have backyard childcare, backyard parking, and a backy

by kragen 1y ago

Around here I have a backyard washing machine, a backyard grill, a backyard clothesline, and my neighbors have backyard childcare, backyard parking, and a backyard motorcycle. And actually inside my house I have a flush toilet and an oven. Your assertion seems to imply that if instead I had to wash and dry my laundry at a coin-operated laundromat, eat at a cafeteria table with all my neighbors, and piss in a portapotty down the street, and my neighbor's kids were taken away to live at a boarding school, while he had to ride the bus to work instead of driving, that would be evidence of successful policy! We clearly have radically different definitions of success. Yours seems not to be well thought out. The problem with backyard furnaces was that the steel they produced was not just expensive but also inferior. If you could plug in a Tabletop Minimill that produced ingots of specialty precision steel alloys to your order, ranging from maraging steel to tool steel to nitinol, and which was also cheap to operate, that would not be an economic policy failure. That would be an enormous success! Your generalization that any kind of decentralization amounts to collective action failure is wrong. Some kinds of decentralization do; my household water pressure tank is decentralized because of collective action failure, for example, as I explained in https://news.ycombinator.com/item?id=45487610 https://news.ycombinator.com/item?id=45487610. Others don't. We have to analyze the particulars of each case. I've written at https://news.ycombinator.com/item?id=45487051 https://news.ycombinator.com/item?id=45487051 a bit about the particulars of this case.

alexey-salmin 1y ago
> Your generalization that any kind of decentralization amounts to collective action failure is wrong. Some kinds of decentralization do; ... Others don't. That's correct, yes. I too prefer a decentralized washing machine to coin-operated laundromats which are very common in France where I live nowadays. However I still believe that electricity generation has more in common with steelmaking than with laundry. Today solar panels make sense in terms of money ROI but not in terms of KWh ROI, they're largely enabled by economical disbalances around the world and not by their long-term value. The long-term solution in my view is nuclear which coincidentally the French do rather well, it's not all laundromats.
- Slartie 1y ago
  > Today solar panels make sense in terms of money ROI but not in terms of KWh ROI That is clearly wrong. Even the worst-case embodied energy assumptions for solar panels estimate the cost of producing a square meter of solar panel area at 2000 kWh (the best cases are around 300, see https://en.wikipedia.org/wiki/Energy_return_on_investment#Photovoltaic https://en.wikipedia.org/wiki/Energy_return_on_investment#Ph... ). A square meter of solar panel area produces an average of 200 kWh of power per year in Germany (which implies a pessimistic assumption, more sunnier countries can get a multiple of that). This means that even in the worst case, the solar panel has amortized itself from the perspective of embodied energy after 10 years. On average it will be more like below 5 years. Solar panels however have an expected lifetime of well over 30 years and require no maintenance if installed correctly.
  - alexey-salmin 1y ago
    30 years lifetime (if there are no heavy storms), 5-10 years to recoup energy, that's 3-6 EROI. Then you also need to include inverters which are energy-heavy. If we're talking of any meaningful scale, then you also need to include batteries because without them panels get shut down during sunny hours lowering the EROI even further. In the end you're lucky if you get EROI of 1.5-3. That's an extremely bad deal. All other sources of energy, renewable or not, are in the 20-100 territory. The purpose of the energy production system is not to barely sustain itself but to actually produce usable electricity for everything else you want to do. It's the means, not the ends. Solar panels have very solid benefits of course, you can buy them for your house to be more autonomous -- I would do that if I had a house. But from the energy/CO2 perspective what you're doing is buying a sort of "battery" that was charged with cheap coal-based electricity in China and shipped off to US/Europe. This makes sense in certain situations but presenting this as a green solution or as future of humanity is just delusional.
    - Slartie 1y ago
      > Then you also need to include inverters which are energy-heavy. Scientific publications say otherwise. It's quite hard to come by any numbers, but page 11 of https://www.wisdomlib.org/uploads/journals/mdpi-sust/2025-volume-17-issue-3--2071-1050-17-3-852-.pdf https://www.wisdomlib.org/uploads/journals/mdpi-sust/2025-vo... says that typical solar inverters require about 15 MJ/kW of power for their production in total, which would amount to approx. 4 kWh per kW of inverter power. A solar panel square meter produces about 200-250 watts of peak power, so it needs inverter power that cost about 1 kWh to build. Let's triple that, because inverters have a typical lifetime of 10 years in contrast to the 30 years of solar panels. 3 kWh for the inverter is negligible when compared with 300-2000 kWh for the panel itself. So we can just ignore that. Batteries are interesting. According to https://www.mdpi.com/2076-3298/12/1/24 https://www.mdpi.com/2076-3298/12/1/24, it takes about 35 kWh in total to produce 1 kWh of battery capacity. Let's say we'll need 250 Wh of capacity for our 250 WPeak solar panel square meter (the rule of 1 kWh capacity for each 1 kWp is a typical estimate applied when sizing solar installations for residential homes). That makes up about 8 or 9 kWh to produce this battery capacity. Admittedly that doesn't include the raw materials, of which the cell requires quite a few expensive ones. Unfortunately I wasn't able to find a good resource on that, so I resorted to asking ChatGPT for a rough calculation, and it came up with about 140 kWh for our 250 Wh LFP cell, which doesn't sound entirely wrong, as it assumed a cell weight of 1,5 kg and splitted that up into different materials. The weight matches what I would expect from personal experience with LFP batteries. Basically, we can just ignore the inverter and must add about 150 kWh for the battery to our 300-2000 kWh for the panel. That does not substantially impact an EROI calculated from a 1000 kWh assumption for the panel alone. And this is a calculation based on Germany. Again: weather conditions are far from optimal for solar in Germany. It's much better in many regions in China, where solar panels are made. They can easily achieve EROIs of 20+ with solar there, which is probably the reason why China installs absolutely HUGE numbers of panels. But according to you, they must be "delusional" over there.

by kragen 1y ago

alexey-salmin 1y ago

> Your generalization that any kind of decentralization amounts to collective action failure is wrong. Some kinds of decentralization do; ... Others don't. That's correct, yes. I too prefer a decentralized washing machine to coin-operated laundromats which are very common in France where I live nowadays. However I still believe that electricity generation has more in common with steelmaking than with laundry. Today solar panels make sense in terms of money ROI but not in terms of KWh ROI, they're largely enabled by economical disbalances around the world and not by their long-term value. The long-term solution in my view is nuclear which coincidentally the French do rather well, it's not all laundromats.

Slartie 1y ago
> Today solar panels make sense in terms of money ROI but not in terms of KWh ROI That is clearly wrong. Even the worst-case embodied energy assumptions for solar panels estimate the cost of producing a square meter of solar panel area at 2000 kWh (the best cases are around 300, see https://en.wikipedia.org/wiki/Energy_return_on_investment#Photovoltaic https://en.wikipedia.org/wiki/Energy_return_on_investment#Ph... ). A square meter of solar panel area produces an average of 200 kWh of power per year in Germany (which implies a pessimistic assumption, more sunnier countries can get a multiple of that). This means that even in the worst case, the solar panel has amortized itself from the perspective of embodied energy after 10 years. On average it will be more like below 5 years. Solar panels however have an expected lifetime of well over 30 years and require no maintenance if installed correctly.
- alexey-salmin 1y ago
  30 years lifetime (if there are no heavy storms), 5-10 years to recoup energy, that's 3-6 EROI. Then you also need to include inverters which are energy-heavy. If we're talking of any meaningful scale, then you also need to include batteries because without them panels get shut down during sunny hours lowering the EROI even further. In the end you're lucky if you get EROI of 1.5-3. That's an extremely bad deal. All other sources of energy, renewable or not, are in the 20-100 territory. The purpose of the energy production system is not to barely sustain itself but to actually produce usable electricity for everything else you want to do. It's the means, not the ends. Solar panels have very solid benefits of course, you can buy them for your house to be more autonomous -- I would do that if I had a house. But from the energy/CO2 perspective what you're doing is buying a sort of "battery" that was charged with cheap coal-based electricity in China and shipped off to US/Europe. This makes sense in certain situations but presenting this as a green solution or as future of humanity is just delusional.
  - Slartie 1y ago
    > Then you also need to include inverters which are energy-heavy. Scientific publications say otherwise. It's quite hard to come by any numbers, but page 11 of https://www.wisdomlib.org/uploads/journals/mdpi-sust/2025-volume-17-issue-3--2071-1050-17-3-852-.pdf https://www.wisdomlib.org/uploads/journals/mdpi-sust/2025-vo... says that typical solar inverters require about 15 MJ/kW of power for their production in total, which would amount to approx. 4 kWh per kW of inverter power. A solar panel square meter produces about 200-250 watts of peak power, so it needs inverter power that cost about 1 kWh to build. Let's triple that, because inverters have a typical lifetime of 10 years in contrast to the 30 years of solar panels. 3 kWh for the inverter is negligible when compared with 300-2000 kWh for the panel itself. So we can just ignore that. Batteries are interesting. According to https://www.mdpi.com/2076-3298/12/1/24 https://www.mdpi.com/2076-3298/12/1/24, it takes about 35 kWh in total to produce 1 kWh of battery capacity. Let's say we'll need 250 Wh of capacity for our 250 WPeak solar panel square meter (the rule of 1 kWh capacity for each 1 kWp is a typical estimate applied when sizing solar installations for residential homes). That makes up about 8 or 9 kWh to produce this battery capacity. Admittedly that doesn't include the raw materials, of which the cell requires quite a few expensive ones. Unfortunately I wasn't able to find a good resource on that, so I resorted to asking ChatGPT for a rough calculation, and it came up with about 140 kWh for our 250 Wh LFP cell, which doesn't sound entirely wrong, as it assumed a cell weight of 1,5 kg and splitted that up into different materials. The weight matches what I would expect from personal experience with LFP batteries. Basically, we can just ignore the inverter and must add about 150 kWh for the battery to our 300-2000 kWh for the panel. That does not substantially impact an EROI calculated from a 1000 kWh assumption for the panel alone. And this is a calculation based on Germany. Again: weather conditions are far from optimal for solar in Germany. It's much better in many regions in China, where solar panels are made. They can easily achieve EROIs of 20+ with solar there, which is probably the reason why China installs absolutely HUGE numbers of panels. But according to you, they must be "delusional" over there.