Solar Energy Is Now the Cheapest Source of Power, Study

I dug down trying to get the actual paper and it appears it's a pre-print with only the abstract available (but maybe I wasn't looking in the right spot).

That being said, the quotes from the author were more to the point of it being a milestone that in the UK that Solar+Battery systems were now less expensive than gas/coal.

To my understanding, this is a milestone vs "raw" production numbers, which you're correct in saying have had solar as the cheapest option for years.

Wouldn't it depend a lot on where you live in terms of weather and/or required pollution controls?

Do you think those are significantly different for solar versus other technologies? Seems they would be right in line with anything else...

Transmission is more of a problem with the centralized forms of power generation? So transmission should be a net positive for solar?

Construction and disposal I'm not sure to be honest, intuitively I don't think those are much more expensive.

Also storage

It doesn't exclude construction. The abstract says they analyzed levelized cost of electricity (LCOE). LCOE includes investments, operations, and fuel, so it includes construction.

Is there a form of power that doesn't have these? A coal plant has all these plus fuel costs, workers to run the plant, parts for the turbines, and ash disposal to deal with.

As usual, nothing, and not covering storage either.

So it's cheapest in sun peak, when it's sunny

Land is also spoiled and occupied for fossil fuel extraction, refinement, and distribution. And at least you can be within the near vicinity of solar panels without experiencing significant health issues, so we need to tabulate the very real public health costs too when comparing. I highly doubt there is a solar equivalent of “Cancer Alley” in the US. Another consideration is that we can slap solar panels on top of existing structures, which you cannot do with fossil fuel at any stage.

I don’t doubt that the costs are considerable but if we’re going to go down that path we need to make sure it’s applied across the board here and I imagine the gap is not as bad as you’re implying. I’m going off my gut though so I could be entirely wrong.

Because solar (for example) can often go on rooftops so not needing new land. And solar and wind are often good in remote areas where population is low and land is cheap.

Because land doesn't need to be vacant to develop solar?

Residential or industrial rooftops are perfect examples. Heck, sound barriers along freeways etc are now even profitable

Because the land cost and amount needed is immaterial. Solar and wind actually contribute to the tax base where it is colocated, tax revenue that would otherwise not be provided (depending on jurisdiction). People overvalue land for some reason, most of the world is not Manhattan or Hong Kong. The US farms 60 million acres for corn and soybean for biofuels alone, for example, and of course that land is much more efficiently used if solar PV is installed for energy.

https://elements.visualcapitalist.com/how-much-land-power-us...

https://www.weforum.org/stories/2021/10/solar-panels-half-th...

https://thebreakthrough.org/issues/food-agriculture-environm...

https://www.canarymedia.com/articles/solar/california-first-...

Edit: your comment has changed a bit, so let me address the new sentence here:

> I don't understand. Why is it that in every single discussion about power- and in particular, solar power- is the cost of land omitted ?

You are incorrect that land costs are omitted; in fact I don't recall any sort of cost comparison that has ever omitted land costs, whether it's from Lazard or NREL or anywhere else. It's part of the CapEx, or as rent, or however it's modeled. NREL in particular looks at overall costs on completed and built systems. Further, the developers building these things know the land costs very well.

Can you point to a discussion or paper where land cost was not included? If it's still an open discussion I'd love to add the high quality solar-is-cheapest studies that all include land costs.

And as for this particular study:

First, doesn't omit the cost of land. Secondly land is only the most important resource in very very few markets. Razing skyscrapers in Manhattan would be a bad idea. Instead of farming crops or the sun in Manhattan, it's done elsewhere and shipped in.

It's also not clear if they factored in battery costs. They mention lithium battery costs have also decreased in cost but not clear if its included. Batteries are a requirement for consistently delivering solar power. You cant just flip the sun on and off.

The cost of land is usually omitted in casual discussions because it's not a significant driver of cost for solar. Quantitative cost breakdowns do include it. See for example https://www.eia.gov/analysis/studies/powerplants/capitalcost..., where the cost estimate for a US$200 million 150-megawatt-AC solar plant in 02019 (six years ago) included US$133,000 per year in land leasing costs, one fourth the cost of "module cleaning" and one sixth the cost of "preventative maintenance".

However, not every single discussion does omit it. For example, we were discussing this yesterday at https://news.ycombinator.com/item?id=45487268 and https://news.ycombinator.com/item?id=45487197, where my calculation was that even in solar-unfavorable places like the northern extreme of Germany, the cost of land barely reaches the same order of magnitude as the cost of the solar modules, even at today's record-low module prices. US$72 million of the US$200 million of the estimate mentioned above was the modules themselves, but today that would be closer to US$15 million.

The atom of truth in your confused assertion is that solar farms do take up enormous amounts of land compared to other kinds of power plants. But, at current human energy consumption levels, it's still only a tiny amount of overall land.

I'm assuming from your past submission history that you're in the US, because you mostly only post US politics stuff. The US generates about 4200 TWh electric per year, which is 480 gigawatts (https://en.wikipedia.org/wiki/Electricity_sector_of_the_Unit...). Under the 2000kWh/year/kWp conditions prevailing in the Californian Mojave Desert, parts of Arizona, and parts of New Mexico, according to https://solargis.com/resources/free-maps-and-gis-data?locali..., this would require about 2.1 terawatts (peak) of solar panels. A square meter of 22%-efficient solar panel produces 220 watts, peak, so this is about 9600 square kilometers, a 110-kilometer-diameter circle. (Although, if you don't leave spaces between the panels, you have to make them horizontal so they don't shade each other; in practice people set them further apart to use more land but less panels.)

How big is that?

It's almost 0.1% of the US, not counting the space between the panels. It's a little over twice the diameter of the VLA radio telescope in western New Mexico, a facility you might remember from the movie Contact. It's slightly larger than White Sands Missile Range (8300km²). It's 15 times the size of Lake Mead, which was created by Hoover Dam to generate electricity. It's about two thirds the size of California's Death Valley National Park (13'793km²). It would take up one eighth of the Navajo Reservation.

Almost everywhere in the US has at least half that much sunlight. Suppose you wanted to site the panels near Springport, Michigan, for some reason. You only have 1200 kWh/year/kWp there, so you need 16000km². The panels would cover a ten-county area centered on Jackson County; they would reach Lansing, and might reach from Ann Arbor to Kalamazoo.

In real life, it wouldn't be a good idea to put it all in one place like that, both because it's fragile and because it creates higher transmission-line losses; it's better to put the panels closer to where they're used, which implies spreading them apart. So probably 0.2% of every state. In an average-sized state like Iowa (145'746km²) it might be 300km², 20% of the size of an average-sized county like Hamilton County. To power the whole state.

But try to keep perspective on the total amount of land we're talking about here: White Sands Missile Range, two thirds of Death Valley, or a small part of the Navajo Reservation to power the whole country.

Why are we assuming that the cost of land isn't included in their LCoE calculation?

Why are you saying land is not included in the cost calculations? I've not read this particular study, but every study looking at the cost of electricity generation I have seen in recent years, essentially looked at investment/returns which obviously includes land use cost. The reason why it's not a big topic, is because it's generally not a big contributor to overall cost, the main drivers being panels and construction. Incidentally, the main power sources where things are not included in the cost are nuclear (typically risk for major accidents is taken on by the government, as no insurance will do it, waste storage cost is often capped as well incalculations) and fossils (health effects of pollution, CO2...).

What's the cite for solar requiring "massive amounts of land and taxes"? That doesn't square with the back of my envelope but I'd be willing to look at numbers.

PV is fairly unique in that land is not required. It can be installed over existing land uses such as rooftops, canals, parking.

Because the land is already owned for the most part?

It's been included in every one I've read. You're not going to find it in broad surveys like this one, because it's a survey. They'll average together the cost of hundreds of projects. But when you look at the projects individually that were the sources, land acquisition is part of the capital costs or land lease is part of the operating costs.

I'm not sure how practical that would be. Transmitting power over distances incurs losses.

I did see a proposal to build out solar in Africa and pipe it undersea to Europe. That seemed wild, and, predictably, it got canned for its impracticality.

Edit - it looks like there are several such proposals, and that they're not all cancelled:

Several travel across the Mediterranean:

https://gregy-interconnector.gr/index_en.html

https://www.ecofinagency.com/news-industry/0210-49221-egypt-...

Here's the one I thought was cancelled, which travels along the western coast of Africa to the UK:

https://en.wikipedia.org/wiki/Xlinks_Morocco%E2%80%93UK_Powe...

https://xlinks.co/morocco-uk-power-project/

https://thenational-the-national-prod.cdn.arcpublishing.com/...

Why transmit it rather than just storing the unused power for later near the source?

The Pan-European grid is already reality. The fluctuating generation costs between e.g. the Nordics and mainland Europe are already driving investment into undersea HVDC cables.

To transmit power you will need battery storage on both sides of long distance power lines. We never build enough of them, but if you can peak shave and trough fill with stored power, you increase the MWH per day available on the far end.

Solar tends to come with battery storage. But you could also just build the battery storage.

Define "real base power"? You think the grid can only operate if every generator is always-on? How do you square that with the fact that the grid is operating just fine with this stuff already?

This is a fallacy, basically. Not least because electricity is by far the most mobile traded commodity in human history. Not enough sun today where you live? Buy your power from across the continent, where they have plenty. Or from your wind generators which are working fine. Or the wind generators across the continent if you have to. Or crank up the hydro dams (most of which rarely run at 100%) a bit to handle the shortfall. Or even fire up an idle gas plant if you absolutely can't get anything else.

The idea that solar and wind aren't (sigh) "real" is a lie that someone sold you. The real world relies on a lot of this stuff already and the promised apocalypse never arrived. Go figure.

The problem with the idea of "base power" is that in the past, the cheapest electricity was from the big thermal generators that take a day to warm up, and operate most cheaply by pumping out at maximum efficiency 24 hours a day. You could then layer on the more expensive electricity sources that could spin up in 15 minutes or a few hours, and match the demand curve, as long as you matched baseload generators to the minimum of the demand curve, and have a cost-optimal electricity mix.

Now that we have cheaper sources of energy for parts of the day, "base" power is a much less desirable concept. It's gone from a simple and straightforward optimization problem that a middle-schooler could solve to a cost optimization problem that markets and linear solvers can solve.

Now that we have cheap storage, and solar-plus-storage is cheaper than coal in the UK, the cost optimization is getting simpler: get rid of all the base load coal!

Base power isn't really all that important for solar, because its already covered by nuclear/other.

Getting "18" hours of power, thats actually important (https://timera-energy.com/blog/iberian-price-divergence-on-i...) if you can cover the evening peak, then most of your energy costs disappear.

24hrs solar is already a thing that is possible for large parts of the world with lithium batteries.

https://ember-energy.org/latest-insights/solar-electricity-e...

That study is already old as the prices for batteries have come down a lot more since then.