Technology

The cost of solar and wind is becoming so attractive, it’s hard to see why anyone would do otherwise.

The elephant in the room (at least for the US; I’m not as familiar with UK policies) are the subsidies. It sparks new investments because many of the incentives are specifically related to new projects. Other ones mess with the valuation of the equipment, making long term tax burden much lower. It’s not the only energy industry to receive subsidies. But it’s pretty asinine to continue to support the one that’s destroying our world.

“In one case, it’s going to profit, amplifying the incumbent status of the oil and gas industry. In another, under more aggressive decarbonization policy and low oil and gas prices, it’s actively working against the climate goal by spurring additional production.”

https://www.canarymedia.com/articles/fossil-fuels/subsidies-really-do-matter-to-the-us-oil-gas-industry-one-in-particular

Both however are so much cheaper than Nuclear and especially than oil/drilling fuels that its hard to see much real investment in those older technologies.

I keep telling people that the economics of nuclear - especially new plants - just doesn’t work, but here and on Reddit it seems to be a very bitter pill that many are not ready to swallow.

The time of nuclear energy has come and gone. We missed it.

I’m not some anti-nuclear energy hippie. I took nuclear reactor design courses at uni. But you just can’t make money that way anymore.

And these numbers don’t even take into account the question of what will happen to fossil fuel companies if society / states start to make them accountable for even a tiny fraction of the environmental damage their business model has caused.

Working in the mining sector, I do worry about the ability of metals production to keep pace with demand

If mining can't keep up, recycling will fill in the gaps. We toss a lot of metal into landfill and regularly replace electronics that could easily be repaired. I think we'll be fine.

I think some people underestimate how much work will need to be done on grids before solar/wind can dominate?

I took a team of three electricians about four hours to install solar on my home. They were done by lunch time and did a second home in the neighbourhood the same day, and that included connecting the panels to the grid — most of the power we produce is sold to the grid. It didn't cost much either - the four hours labour was about a third of the cost to give you some idea.

It's a relatively small system, but even on an overcast day they produce more power than we use even under peak load situations. On a sunny day it produces more power than we use in a week.

Obviously we're still drawing power from the grid overnight... but I was happy to learn (after seeing the metrics produced by our new solar system) that we don't use much power overnight and could easily use less - it's mostly just our kitchen fridge which is close to end of life anyway and we will be replacing it with a more efficient one.

Overnight grid power can easily be covered by wind/hydro. NIMBY is definitely a struggle there, but it can be managed. And nuclear is even worse - nobody wants to live near a nuclear power plant. Nobody even wants to live near a road that is occasionally used to transport radioactive materials.

But the real deathblow for nuclear power is it takes an average 7.5 years to build a nuclear power plant. That doesn't compare favourably at all to 4 hours for an adequately sized solar system and anyone building a nuclear power plant runs the risk that by the time they start operations, they might find nobody wants to buy the power they're producing. In seven years time there's a pretty good chance I will have added a battery to my solar system and we'll only be drawing power from the grid if there's something wrong with our own power system. There's no way I'd get behind investing billions of dollars in a nuclear power plant - there's just too much risk.

A huge portion of our overnight power consumption right now is big industry consumers that operate at night because baseload power is really cheap at night. That demand won't last - as more and more solar is installed those operations are transitioning to daytime operations. It's already started, aluminium for example is now about a hundred dollars per ton cheaper to produce with solar than with fossil fuels. Which means the entire industry needs to switch over, and fast, or else they won't be able to find anyone who will buy what they're producing.

With high efficiency DCDC converters now cheap and good, we totally could do it.

I can see in the future a standard for DC power with a completely different power socket that saves up the round trip of DC to AC then back into DC for all our electronics. There are fairly substantial benefits to be had but as I think it through the usable DC is 3 to 12V and appreciable load will mean those cables will push a lot of amps. Not sure the economics will work out without it being high voltage and that fits nothing DC today. If it's high voltage then everything needs a converter again.

Setting a DC standard today will be a world wide nightmare, we need it but it's got to be dumb for longevities sake. It's far off I think.

USB is the new defacto standard DC outlet. It's only suitable for fairly low power (240W) though. That's enough to charge your laptop but not enough to run a gaming PC.

Not for the reason you're thinking. DCDC conversion has gotten more efficient, but the problem is that nothing wants a specific voltage. Your LED strip might want 12 or 24 or 48V. Your phone wants 5V. Your single board computer might want 3.3V. Meanwhile, what's flying out of your solar panel array might be over 300V (depending on how they're wired together). You end up with layers of conversions that each take an efficiency hit. Add it all up, and it's better to convert the solar output to AC and then convert once to DC at the point of use. Which is exactly what we're doing.

Plus, items with larger draw, like a clothes dryer or electric stove, are going to want AC, anyway.

However, there's another thing that might make DC lines viable in houses. Power over Ethernet equipment has gotten relatively cheap. The base power levels are easily enough to run lights or charge a phone, and the more powerful versions can handle a laptop. What makes it especially interesting is that you don't need a licensed electrician to run it. You can't be a complete imbecile installing it, but don't need a license, either. That could dramatically reduce install costs in new homes. You need a proper electrician to install the breaker box and run a few high draw lines, but skip the rest.

As a bonus, any smart controlled devices no longer need wifi or zigbee or bluetooth or anything. They're already connected to the network by being plugged in.

PoE has been mostly an enterprise thing, but mass production for residential use should further bring prices down.

What about all the appliances expecting AC input?

Putting panels on houses is the least cost effective way to do solar. Every house only has so much space on the roof, you have to work around vents and other obstructions, and the inverters will be smaller (meaning less efficient than one big one). It's a custom project for every single house.

Conversely, if you buy a field, you can make racks that slap on every panel the same, there are fewer obstructions to work around, and you can buy one big inverter to cover the whole thing. This is where solar gets cheap. Really cheap.

I know everyone wants to stick it to power companies and have energy independence. If you want that, look for community solar projects where everyone pitches in a little to get a larger deployment. If your state or locality is blocking those projects, then start going to campaign events and asking them about. Hold them to it with votes and campaign contributions.

Money. Why spend x extra dollars when you don't have to?

I think California has a new law that mandates this, but until their is loud and constant consumer demand home builders are going to continue doing the bare minimum.

What do you mean? We totally don't have datacenters running AI models looking at purchasing small modular nuclear reactors just to run them.