“I expect a semi-dystopian future with substantial pain and suffering for the people of the Global South,” one expert said.

  • @[email protected]
    link
    fedilink
    English
    5
    edit-2
    8 months ago

    I don’t think that’s quite true. Where I live it has expanded from nothing to a major power source in just a few years. We’ll need grid storage of some kind to kick fossil fuels completely, but that seems surmountable. Worst case scenario we build pumped air and just eat some round trip losses.

    Nuclear plants take many years to get off the ground, so I’m not sure that’s actually an easier solution. Once they’re up and running at scale they’re actually really cheap per unit production, so I would have agreed with you a decade ago, but as it is solar and wind have just pulled ahead.

    • @dgmib
      link
      English
      28 months ago

      Don’t take my word for it. Look up the numbers for yourself and do the math.

      Search for “National GHG inventory {your country}”.

      You find a report listing (among a bunch of other things) the amount of electricity generated each year by each method, and the emissions from each. Look up the total TWh of electricity produced by fossil fuels.

      Then look at the total TWh from renewables, and rate it has been growing Y-o-Y and extrapolate until it reaches the number needed to eliminate fossil fuels.

      You’ll find it will take decades to build enough renewable capacity to replace fossil fuel based electricity generation.

      And that’s before you realize that only about 25% of fossil fuel combustion goes to electricity generation. As we start switching cars, homes, industries to electric we’re going to need 2x-3x more electricity generation.

      Yes it takes a long time to bring on a new nuclear plant, roughly 7-9 years. If it was remotely realistic that we could build enough renewable power generation in that time to replace all fossil fuel generation then I’d agree we don’t need nuclear. But we’re not anywhere close to that.

      It’s also helpful to note too just how much power a nuclear reactor generates. I live in Canada, our second smallest nuclear power plant in Pickering, generates almost 5 times more electricity annually than all of Canada’s solar farms combined. It will take 1000s or solar and wind farms covering and area larger than all of our major cities combined to replace fossil fuels…

      …or about 7 nuclear power stations the same size as Pickering.

      • @ammonium
        link
        English
        28 months ago

        Then look at the total TWh from renewables, and rate it has been growing Y-o-Y and extrapolate until it reaches the number needed to eliminate fossil fuels.

        You’ll find it will take decades to build enough renewable capacity to replace fossil fuel based electricity generation.

        I get ~2 decades when I extrapolate these numbers (from 2010-2023) to get to 2022 total primary energy usage for solar alone.

        Energy usage will grow as well, and keeping that growth is ambitious, but it the future doesn’t look that bleak too me if you look at it that way.

        • @[email protected]
          link
          fedilink
          English
          1
          edit-2
          7 months ago

          Did you use linear extrapolation, or something else? Because it’s an actual paradigm shift happening now, I’d guess some kind of exponential or subexponential curve would be best. That would bring it even faster.

          Extrapolation is tricky, and actually kind of weak, although I think it’s appropriate here. This XKCD explains it really well, and I end up linking it all the damn time.

          • @ammonium
            link
            English
            27 months ago

            Exponential, it fits the curve very nicely. I can give you the python code if you want to. I got 2 decades for all energy usage, not only electricity, which is only one sixth of that.

            I just took the numbers for the whole world, that’s easier to find and in the end the only thing that matters.

            The next few years are going to be interesting in my opinion. If we can make efuels cheaper than fossil fuels (look up Prometheus Fuels and Terraform Industries), we’re going to jump even harder on solar and if production can keep up it will even grow faster.

            • @[email protected]
              link
              fedilink
              English
              1
              edit-2
              7 months ago

              Yes, code please! This sounds amazing.

              E-fuels are a big deal, particularly for aviation. Non-electricity emissions are also something to watch. Hydrogen as a reducing agent seems like it can work very well as long as we do phase out fossil fuels like promised, so that solves steel production and similar. Calcination CO2 from concrete kilns is a very sticky wicket apparently, since they’re extremely hot, heavy, and also need to rotate, which is challenging to combine with a good seal.

              Cheap grid storage is a trillion-dollar question, but I suspect even if new technology doesn’t materialise, pumped air with some losses can do the trick, again subject to proper phase-out of dirty power sources.

              • @ammonium
                link
                English
                27 months ago

                Here you go, you’ll need numpy, scipy and matplotlib:

                from scipy.optimize import curve_fit
                from matplotlib import pyplot as plt
                
                # 2010-2013 data from https://ourworldindata.org/renewable-energy [TWh]
                y = np.array([32, 63, 97, 132, 198, 256, 328, 445, 575, 659, 853, 1055, 1323, 1629])
                x = np.arange(0, len(y))
                
                # function we expect the data to fit
                fit_func = lambda x, a, b, c: a * np.exp2(b * x ) + c
                popt, _ = curve_fit(fit_func, x, y, maxfev=5000)
                
                fig, ax = plt.subplots()
                ax.scatter(x + 2010, y, label="Data", color="b", linestyle=":")
                ax.plot(x + 2010, fit_func(x, *popt), color="r", linewidth=3.0, linestyle="-", label='best fit curve: $y={0:.3f} * 2^{{{1:.3f}x}} + {2:.3f}$'.format(*popt))
                plt.legend()
                plt.show()
                

                Here’s what I get, global solar energy generated doubles every ~3.5 (1/0.284) years.

                • @[email protected]
                  link
                  fedilink
                  English
                  17 months ago

                  Thank you! That does look like a great fit.

                  So that’s just solar, then? Long term, it does seem like the one that’s the biggest deal, but right now there’s also a lot of wind and hydro in the mix, so that’s another point in favour of the assumptions here being conservative.

                  • @ammonium
                    link
                    English
                    27 months ago

                    Yes, just solar. Hydro is bigger now, but it doesn’t have the growing potential. Wind is currently also growing exponential, but I don’t see it doing that for 20 more years. And even if it does, it doesn’t really make a big difference since exponential + exponential is still exponential. If it grows as fast as solar that would mean we’re just a few years ahead of the curve.

      • @[email protected]
        link
        fedilink
        English
        27 months ago

        Sorry for the delay. I’m trying to get this the response it deserves, including gathering figures for Alberta, and some basic mathematical modeling.

      • @[email protected]
        link
        fedilink
        English
        17 months ago

        Alright, I can’t seem to find useful numbers anywhere. We went from 50% coal to nil in just a few years, though, so big changes fast are possible. If you’re in Ontario, you also have to consider your local renewables penetration was really high to start with, because of those waterfalls.

        And yeah, like I said to the other person, exact growth pattern matters. It’s probably exponential-ish right now, not linear, because it’s just unambiguously cheaper to move to renewables, and so just getting ducks in order to do it is the bottleneck.

        • @dgmib
          link
          English
          27 months ago

          I respect you for doing your own research. People need to understand the scope of the problem if there’s going to be meaningful action.

          The reason I’m passionate about nuclear in particular is that only about a quarter of all fossil fuel consumption is from electricity generation.

          Most of the rest is burned in transportation, buildings, commercial and residential applications. We have the tech already to switch most of these things to electricity, and eliminate their direct emissions, but that’s not much of a win if we’re burning fossil fuels generate that electricity. Which is what happens today when electricity demand is increased, we can’t just turn up the output of a solar/wind farm in periods of high demand, but we can burn more natural gas.

          Switching to electric everything (Car, trucks, ships, heat pumps, furnaces, etc) will increase electricity demand by 2-3x.

          Even if renewables growth is held to the exponential-ish curve it’s been so far (doubtful) we still need 15+ years just to get to the point of replacing current global fossil fuel electricity production in the most optimistic case, never mind enough to handle 2-3x demand.

          Massive quantities of new carbon free electricity generation is needed to “unlock” the electrification technologies we need to deploy if we going to avoid the worst of the disaster. If we wait until renewables alone get us there it’ll be too late.

          The more carbon free energy we can build in the next 20-30 years, the more options we have. Even if we can reach a place of excess capacity, there are a lot of things like DAC and CCS, that we could use it for that today result in more emissions from electricity generation than they sequester.