5 ms·

Let's run the numbers! - Every 4 years our production doubles - Meaning every year our production capacity increases by .5 (or 5 times every decade).

Let's run the numbers! - Every 4 years our production doubles - Meaning every year our production capacity increases by .5 (or 5 times every decade). - The first year (2020) starts at 165GWh capacity production - The total energy use of the world (in 2013) was 157,000 TWh. - Annual growth in energy demand was 40% between 1990 and 2008. - 20% for the US - 150% for China Let's calculate solar energy generation capacity using the growth estimates provided. 2020: 165GWh 2030: 825GWh 2040: 4.1TWh 2050: 20.6TWh 2060: 103,1TWh (estimated global demand of 400TWh by this point) ---- 2061: 154,1TWh So assuming: - Production capacity and panel efficiency does not increase (which it likely will) - Solar will be the only source of energy (it's more likely to be one of a few technologies used) - We don't factor infrastructure changes, space or installation time. - We don't factor the consumers being ready for energy in the form of electricity (electrifying industry, cars, homes) - We don't factor in the environmental impact of producing panels. - We assume grid scale batteries that can handle buffering exist already and are installed Seems like there is reason for cautious optimism. We have to see how governments go about rapidly normalising our energy usage to electricity (banning new purchases of gas cars?), upgrading infra and how they will generate the remaining electricity (LNG, nuclear, coal?). We can expect to see our 2013 energy usage met with solar alone by 2060. Might be reasonable to use solar exclusively by the turn of the century (2100) If the growth in energy demand continues to rise at its current levels, by 2060 we might see an energy demand of 400TWh. This indicates that with the currently predicted solar production scaling, solar can cover half of our energy requirements. There are enormous challenges with things like rapidly electrifying the automotive industry, mining the materials required to meet our grid scale demand, fabrication of the components, integrating a technology that isn't plug-and-play compatible with our grid infrastructure and supporting an IO model from every node. Source (I know it's just wikipedia but come on, this is just a comment on HN): https://en.wikipedia.org/wiki/World_energy_consumption#Trends https://en.wikipedia.org/wiki/World_energy_consumption#Trend...

hebrox 6y ago
When I look at this chart [0], seeing that we’re wasting about 2/3 of our energy, I expect the energy demand to drop significantly if we electrify everything and have more local energy sources. [0]: https://flowcharts.llnl.gov/ https://flowcharts.llnl.gov/
- apatheticonion 6y ago
  This is a good point. Seems like there is reason to be optimistic. I certainly think it's possible for the US to be carbon neutral by 2050 - but ya'll better start banning gas cars sooner rather than later (seeing as modern cars will last 20-30 years).
- ahartmetz 6y ago
  Note that much of the waste probably comes from thermal / fossil power generation, which is unavoidable even in theory (cf. Carnot efficiency). That is, unavoidable unless chemical energy is turned into electrical energy directly like in fuel cells (these have smaller but still surprisingly large losses), or the rejected thermal energy is used for heating.
  - hebrox 6y ago
    With local energy sources, I meant, local renewables energy sources. So no chemical energy.
beefield 6y ago
Is there some confusion in the units? IEA claims that global electricity production from solar is in the order of 720TWh https://www.iea.org/reports/solar-pv https://www.iea.org/reports/solar-pv And the total electricity consumption is over 20 000 TWh/a, which, given current trends, means that 100% of current electricity consumption would be covered by solar already next decade. (You were discussing energy, not electricity, but electricity should be the smaller one)
kilotaras 6y ago
> Meaning every year our production capacity increases by .5 (or 5 times every decade). Change is geometric (i.e. fixed percentage), not linear (i.e. fixed number). This gives 18.92% yearly increase (instead of .5), but 5.6568x increase in decade (instead of 5x). By 2060 this gives 169 TWh (instead of 103), and 200 in 2061. We would also expect energy growth to level of at US/EU level once it reaches them, there's no reason to expect Asia to sustain it's energy growth indefinitely. Don't get me wrong: solar is still not enough, storage is big elephant in the room, and we should build nuclear yesterday. It's just not as bleak as your comment made it out to be.
jillesvangurp 6y ago
Mining is indeed needed to produce some things. But then those things are not expendables and can be recycled after decades of use. And we're no longer mining stuff that we literally use once to burn it like oil, gas, coal, etc. I'd call that a net win. LNG, nuclear, and coal are each way too expensive long term. With solar prices continuing to drop, this is only going to get worse. Coal is already being decommissioned in many countries. China is an exception but only because they are growing so rapidly that they can't keep up with clean energy deployments (despite having the largest deployment of that world wide). That's an effect that is short term. They actually announced their grid will be 100% renewables by 2060. New gas plants (the least expensive of those 3) at this point don't make a lot of sense either. The investment pitch is horrible: we'll be producing energy at about 3-4x of the current price of renewables for the next 50 years. By the time those 50 years are over it might be as bad as 30x. It goes from bad to worse. Nuclear needs to get a lot cheaper to be interesting for investors. Grid infrastructure is indeed a short term challenge. Having more solar and wind on the grid is causing grid operators to evolve what they are doing to fix that. A few decades is a long time to address issues though. Batteries and cables seem to be what is needed here. European and North American grids seem to have issues getting permits for new cables. E.g. Germany needs more north south cables so people in the south can actually use the wind power generated in the north. But putting a few hundred km of cables in place seems to be a hard problem from a bureaucratic point of view. It's not a technical problem though. We've known how to make electrical cables for a long time. Though there is of course some innovation in that space as well to enable more efficient long distance connections. Cars are electrifying at the pace we can ramp up battery production. Even so, there are a lot of cars in the world and most of the new ones are still not electric. That will flip around towards the end of this decade. From then, decline of ICE could be pretty rapid. The cost advantages for owners will drive this. EVs are a growth business in a covid year and are very profitable. So, car manufacturers are already going as fast as they can here. Interestingly, you see the same kind of learning effect with battery as you see with solar panels. And with wind power too. All of this combined means things could go a lot quicker than some companies hope/expect.
oblio 6y ago
Maybe I'm misreading your numbers, but: The total energy use in 2013 was 157 000 TWh (and it's probably <<much, much>> higher in 2021, due to China and India continuing to develop). The additional solar power being added in 2061, by your own math, is 154 TWh. So, adding up the whole line, that would mean that solar would still only offer about 154 x 40 = ~6 000 TWh (best case scenario) of electricity in 2061. Faaar from the 157 000 TWh used in 2013 and probably super far from what's going to be required in 2061. Did you mess up your math?
ginko 6y ago
Grandparent wrote 165GW not GWh. 165GW is 1445TWh per year.

hebrox 6y ago

When I look at this chart [0], seeing that we’re wasting about 2/3 of our energy, I expect the energy demand to drop significantly if we electrify everything and have more local energy sources. [0]: https://flowcharts.llnl.gov/ https://flowcharts.llnl.gov/

apatheticonion 6y ago
This is a good point. Seems like there is reason to be optimistic. I certainly think it's possible for the US to be carbon neutral by 2050 - but ya'll better start banning gas cars sooner rather than later (seeing as modern cars will last 20-30 years).
ahartmetz 6y ago
Note that much of the waste probably comes from thermal / fossil power generation, which is unavoidable even in theory (cf. Carnot efficiency). That is, unavoidable unless chemical energy is turned into electrical energy directly like in fuel cells (these have smaller but still surprisingly large losses), or the rejected thermal energy is used for heating.
- hebrox 6y ago
  With local energy sources, I meant, local renewables energy sources. So no chemical energy.

beefield 6y ago

Is there some confusion in the units? IEA claims that global electricity production from solar is in the order of 720TWh https://www.iea.org/reports/solar-pv https://www.iea.org/reports/solar-pv And the total electricity consumption is over 20 000 TWh/a, which, given current trends, means that 100% of current electricity consumption would be covered by solar already next decade. (You were discussing energy, not electricity, but electricity should be the smaller one)

kilotaras 6y ago

> Meaning every year our production capacity increases by .5 (or 5 times every decade). Change is geometric (i.e. fixed percentage), not linear (i.e. fixed number). This gives 18.92% yearly increase (instead of .5), but 5.6568x increase in decade (instead of 5x). By 2060 this gives 169 TWh (instead of 103), and 200 in 2061. We would also expect energy growth to level of at US/EU level once it reaches them, there's no reason to expect Asia to sustain it's energy growth indefinitely. Don't get me wrong: solar is still not enough, storage is big elephant in the room, and we should build nuclear yesterday. It's just not as bleak as your comment made it out to be.

jillesvangurp 6y ago

Mining is indeed needed to produce some things. But then those things are not expendables and can be recycled after decades of use. And we're no longer mining stuff that we literally use once to burn it like oil, gas, coal, etc. I'd call that a net win. LNG, nuclear, and coal are each way too expensive long term. With solar prices continuing to drop, this is only going to get worse. Coal is already being decommissioned in many countries. China is an exception but only because they are growing so rapidly that they can't keep up with clean energy deployments (despite having the largest deployment of that world wide). That's an effect that is short term. They actually announced their grid will be 100% renewables by 2060. New gas plants (the least expensive of those 3) at this point don't make a lot of sense either. The investment pitch is horrible: we'll be producing energy at about 3-4x of the current price of renewables for the next 50 years. By the time those 50 years are over it might be as bad as 30x. It goes from bad to worse. Nuclear needs to get a lot cheaper to be interesting for investors. Grid infrastructure is indeed a short term challenge. Having more solar and wind on the grid is causing grid operators to evolve what they are doing to fix that. A few decades is a long time to address issues though. Batteries and cables seem to be what is needed here. European and North American grids seem to have issues getting permits for new cables. E.g. Germany needs more north south cables so people in the south can actually use the wind power generated in the north. But putting a few hundred km of cables in place seems to be a hard problem from a bureaucratic point of view. It's not a technical problem though. We've known how to make electrical cables for a long time. Though there is of course some innovation in that space as well to enable more efficient long distance connections. Cars are electrifying at the pace we can ramp up battery production. Even so, there are a lot of cars in the world and most of the new ones are still not electric. That will flip around towards the end of this decade. From then, decline of ICE could be pretty rapid. The cost advantages for owners will drive this. EVs are a growth business in a covid year and are very profitable. So, car manufacturers are already going as fast as they can here. Interestingly, you see the same kind of learning effect with battery as you see with solar panels. And with wind power too. All of this combined means things could go a lot quicker than some companies hope/expect.

oblio 6y ago

Maybe I'm misreading your numbers, but: The total energy use in 2013 was 157 000 TWh (and it's probably <<much, much>> higher in 2021, due to China and India continuing to develop). The additional solar power being added in 2061, by your own math, is 154 TWh. So, adding up the whole line, that would mean that solar would still only offer about 154 x 40 = ~6 000 TWh (best case scenario) of electricity in 2061. Faaar from the 157 000 TWh used in 2013 and probably super far from what's going to be required in 2061. Did you mess up your math?

ginko 6y ago

Grandparent wrote 165GW not GWh. 165GW is 1445TWh per year.