upvote
I believe transmission is an issue so having the solar array nearer to where the energy is needed is important. Not that we couldn't spend a lot of effort to figure out better transmission methods but that's one of the reasons why blanketing a desert is not as effective as it seems on the surface.
reply
Is transmission an issue due to physical limitations, capital expenditure, or some bad govt policy? China has high voltage lines from their western deserts to eastern cities, so it makes me think the physics of moving the electricity from unpopulated -> populated is easy enough, this just seems like permitting/government nonsense that prevents us from having nice things.
reply
Capex and cost of maintenance is definitely something that factors in this kind of decisions, especially for smaller, commercial projects.
reply
All of the above, the nimby opposition is often fierce.
reply
In a desert?
reply
Well, the point of the transmission is to take it from the desert to the places people live.
reply
Yes and neighbors tend not to be a problem in a desert.
reply
Right, but if we're talking about taking energy from the desert to cities, the HV transmission lines we build have to go through a lot of non-desert...
reply
You are still wasting a lot of energy with long distance transmission lines. We are arguing with physics.
reply
It’s as usual a mix of all the issues. Recall that the United States has had essentially flat energy demand for much of the last twenty years or so, so there are no incentives to build new transmission lines.

Also the OP was thinking that the land under the solar panels could be put into better use, but such ultra high voltage transmission lines themselves come with restrictions on what can be built beneath them.

reply
the scale of transmission lines requires large amounts of capital, but also societal buy-in, aka NIMBYs are the problem, at least in western democracies.

Be careful with the term "government nonsense". It's a very slippery slope from there to losing environmental regulations and eventually representative democracy.

reply
Hmmm, very good callout there re: government nonsense. I can definitely see the slippery slope argument, but I look at something like the california high speed rail project and I think "government nonsense" is a good general term for why that has been such a disaster project that is way over budget and way behind schedule. I realize that is a lazy over-generalized use of the term, but I don't know if I agree that characterizing it as a NIMBY problem (which that obviously played a part) solves it. I think it is maybe a frustration with bad processes - processes designed/envisioned with good intentions, but in practice and in combination with a slew of other processes - you get paralysis.
reply
The purpose of a system is what it does.

CAHSR is very effective and spreading money around to a patronage network. Its purpose just isn’t to actually provide high speed rail.

reply
Everything is government nonsense when you think about it. At least we will always have clean, beatiful coal I guess.
reply
As coal burn declines, rising power demand will be primarily met by renewables - https://news.ycombinator.com/item?id=50500905 - November 2026

- Coal’s share of generation will likely decline from 17% in 2025 to 15% by 2027 as 13 GW of capacity is scheduled to retire by the end of next year, EIA said.

Not many coal plants left to go.

https://www.sierraclub.org/coal/coal-plant-map

reply
Thank God for small mercies. I was just being cheeky/bitter about how when it comes to energy, everything is about government subsidies/investment but so far, fossil got so much more.
reply
> China has high voltage lines from their western deserts to eastern cities,

Long lines are AC, they must come from dams or coal generators, still with a huge loss. Solar is DC. It can be transformed at a loss.

reply
China's ultra high voltage grid is DC for long distance transfer from remote western and northern hubs to eastern cities. This uses 24 long lines with 1 at 1100 kV and the rest at 800 kV. It has about 40000 km of lines.

They also have 20+ ultra high voltage AC lines (1000 kV) for shorter region interconnects.

Historically long transmission lines were AC because long lines have to be high voltage to reduce losses¹. Boosting AC voltage for high voltage transmission can be done with a transformer. DC requires more complicated equipment which wasn't really commercially feasible until early this century.

¹Losses are proportional to current. Power is current x voltage. To transmit a given amount of power the higher the voltage the lower the current, and hence the lower the losses.

reply
Thanks for adding some data to this otherwise somewhat vibesy thread.
reply
Look up the Xizang-to-Guangdong UHVDC Project. The whole project build out will have a transmission capacity of 10 gigawatts and has lines stretching from Tibet to Hong Kong.
reply
Long lines are often DC. Talk to an electrical engineer for details, but for long distances DC works better. For distances at the scale of a US state works great, but as you get longer distances DC becomes better again.
reply
> Long lines are often DC.

There are only a few dozen HVDC lines in the USA, so that is inaccurate.

It is far more likely for a NEW long-distance line to be DC compared to the existing installation base.

reply
To briefly describe why:

Current flowing through a conductor produces heat. This scales quadrtatically with the current, and linearly with the resistance. The resistance depends on the cross-sectional area of the conductor and its length. Long wires have more resistance for the same size of wire.

Transformers can raise the voltage & lower the current, or the opposite, in direct proportion.

So you want long-distance transmission to use the highest voltage and lowest current possible, so that the least amount of power is wasted heating up the wires, and so that you don't have to have ridiculously thick wires.

Transformer efficiency rises with frequency. Standard AC grid frequency is quite low, <100Hz everywhere in the world. For small scale utility transformers that's not a big problem, but for transmitting a lot of power the transformer efficiency matters much more. When a transformer heats up that heating happens all throughout it, but heat only dissipates on the surface, so bigger transformers get harder & harder to cool.

Recent (last decade or two) advances in semiconductor design has enabled some really big transistors which can switch a heck of a lot of power very quickly. Switching DC power on & off to create a square wave creates enough changing current that transformers can use it. Transformers don't strictly need AC, they just need changing current through them.

So it ends up being more efficient to convert AC from generators to DC, switch it on & off really fast to create a really high frequency square wave, run that through a step-up transformer designed for that high frequency, and smooth the result out with some big capacitors to clean up the DC. Then that high-voltage DC can be transmitted with relatively low losses on relatively small cables, and go through a similar process to convert down to low(er)-voltage AC at the receiving end.

That wasn't possible until recently, the transistors to switch that much power fast enough didn't exist. Thus the existing grid mostly being AC, with HVDC preferred for new lines.

reply
There are very few really long lines in the US. There are a lot of lines that cross a state or two, but no farther. AC works just fine for those. When you are crossing China, the US (or any continent) DC becomes much better, but most lines are not nearly that long.
reply
Yep. And don't forget solar has availability factor that isn't all that great. This further increases required transmission capacity.
reply
Sort of. Battery storage colocated with solar generation reduces curtailment and can increase transmission utilization. With enough solar and some battery storage, you can achieve the equivalent of base load/firm generation 24/7. Every battery storage installation also becomes a grid health support node, able to assist in voltage and frequency response, as well as providing synthetic inertia.

CATL Sodium Battery Hits LFP Cost Parity in 2026: 15,000 Cycles, 30-Year Life - https://news.ycombinator.com/item?id=50399657 - November 2026

Ember Energy: Solar electricity every hour of every day is here and it changes everything - https://ember-energy.org/latest-insights/solar-electricity-e... - June 21st, 2025

https://ember-energy.org/app/uploads/2025/06/Ember-24-Hour-S... [pdf]

- 1 kW of stable solar power across 24 hours of the day can be achieved on an average day in a sunny place like Las Vegas with 5 kW of fixed solar panels and a 17 kWh battery.

- The sunniest regions in the world can get as close as 97% of the way to 24/365 solar – stable supply every hour of every day of the year.

- Achieving 97% of the way to 24/365 solar in very sunny regions is now affordable at as low as $104/MWh, cheaper than coal and nuclear and 22% less than a year earlier.

reply
Does that price include any extra transmission costs? Is that a thing?
reply
This study is seven years old, but explains the nuance of transmission costs in the US. Broadly speaking, the closer you can get low carbon generation to demand and load centers and avoid long haul transmission, the better (check out https://web.archive.org/web/20261114074622/https://www.eia.g... | yellow is solar, purple are battery storage).

https://emp.lbl.gov/news/new-national-lab-study-quantifies-0

https://emp.lbl.gov/publications/improving-estimates-transmi...

Improving estimates of transmission capital costs for utility-scale wind and solar projects to inform renewable energy policy - https://linkinghub.elsevier.com/retrieve/pii/S03014215193058... - December 2019

- Estimating the overall costs of transmission needed to integrate variable renewable energy (VRE) onto the grid is challenging. An improved understanding of these transmission costs would support electricity system planning as VRE penetrations increase. This paper brackets VRE transmission capital costs using multiple approaches based on interconnection studies, actual transmission projects, capacity-expansion simulation models, and aggregated U.S. VRE-related transmission expenditures. Each approach possesses advantages and drawbacks, and combining the approaches lends confidence to the results. The resulting range of average levelized VRE transmission costs is $1–$10/MWh, which is generally lower than earlier estimates in the literature. These transmission capital costs can increase the direct plant-level levelized cost of VRE by 3%–33%, based on levelized costs of energy of $29–$56/MWh for utility-scale wind and $36–$46/MWh for utility-scale solar. As VRE deployment continues to expand, policy makers can use this information to (1) assess the benefits of transmission avoidance and deferral when comparing distributed energy resources versus utility-scale projects, (2) evaluate the potential costs of large-scale public transmission investments, and (3) better analyze system-level costs of utility-scale VRE technologies. Future research can expand on the framework presented here by providing a review of operation and maintenance costs for transmission systems.

I don't have a link handy at the moment, but there is also a project to use machine learning to maximize the power existing transmission lines can carry using real time monitoring of their health metrics to push harder towards maximum safe utilization (versus more conservative limits).

reply
Much of what you post is only partially correct. It isn’t wrong, just incomplete.

If solar and wind are as cheap as all of your citations suggest, why are data centres still reliant on the power grid and even literal gas turbines?

The surest sign that your studies and links are only partially correct are the actions of the capitalists. How many data centres can you name that use wind turbine and solar panels instead of a grid connection? These are the precise folks who would jump at the opportunity to ditch the constrains of the power grid. Even better, building power generation near data centres avoids the transmission losses as well.

The reality is that electricity is complex because it depends on time and space, where electricity is available as well as when.

Also, how many industrialised economies can you name that are supplying increased electrical demand with only renewables and batteries? Yes, not replacing, growing demand.

To be clear, most of what you post is true, just not a fair representation of the constraints of reality.

reply
> If solar and wind are as cheap as all of your citations suggest, why are data centres still reliant on the power grid and even literal gas turbines?

The grid because while solar and wind are cheap, they are not concentrated enough for a data center. You can't get enough wind/solar on your data center's land as to power it. So you will need a connection to some form of grid, even if it isn't the commercial one.

I live in city where the power company produces more power from wind than all customers purchase (they sell it elsewhere, and I assume are buying power when the wind isn't blowing). There are a lot of data centers around me because they can get in on that cheap wind power from the grid. They still have backup generators for when the grid goes down.

reply
what this illustrates is that reliability is difficult to price. Renewables still have that problem to solve.

I'm confident that we will eventually solve that problem at scale, but it is then a mischaracterisation to present renewable energy as a drop-in replacement for existing grid capacity.

As an aside, the datacenter push has reignited research into microgrids, which are essential to a future with more renewable energy. Same with agri-PV and rural microgrids.

reply
The World Crossed a Major Solar Milestone [3TW cumulative installed capacity] - https://news.ycombinator.com/item?id=49151977 - August 2026

- The world has reached the 3-terawatt threshold of solar power, with the first terawatt taking 10 years to deploy and the next two terawatts taking less than five years.

- Forecasters expect the world to deploy more than 9 terawatts of solar by 2036, with poor countries' share of solar deployment expected to start growing again from this year onward.

- The main constraint for developing countries to adopt solar power is the lack of accessible financing, and the growth of solar power is also limited by the need for energy storage, such as lithium-ion batteries, to absorb excess power.

Global energy demand growth was met by diverse range of sources in 2025, led by solar and then gas - https://www.iea.org/news/global-energy-demand-growth-was-met... - April 20th, 2026

- All major fuels and technologies expanded to meet rising demand, but at very different rates. Solar PV was the single largest contributor to growth in global energy supply in 2025, accounting for more than 25% of the increase – the first time on record that a modern renewable source has led global primary energy supply growth. Natural gas took the next largest share, at 17%, reflecting its role in power generation in many countries. Overall, renewable sources and nuclear met nearly 60% of all growth in energy demand – and power generation from these sources exceeded total growth in electricity demand.

Solar power generation drives electricity generation growth over the next two years - https://www.eia.gov/todayinenergy/detail.php?id=67005 - January 16th, 2026

- In our STEO forecast, utility-scale solar is the fastest-growing source of electricity generation in the United States, increasing from 290 BkWh in 2025 to 424 BkWh by 2027. Almost 70 gigawatts (GW) of new solar generating capacity projects are scheduled to come online in 2026 and 2027, which represents a 49% increase in U.S. solar operating capacity compared with the end of 2025.

https://ourworldindata.org/grapher/installed-solar-pv-capaci...

https://ourworldindata.org/grapher/installed-global-renewabl...

Data Centers are Helping Fuel a U.S. Battery Boom - https://e360.yale.edu/digest/data-centers-battery-storage - September 2nd, 2026

- Across the U.S., tech companies have courted backlash for using polluting gas generators to power massive new data centers. But energy industry data shows that in addition to building out natural gas, firms are adding huge amounts of battery storage.

- The energy demands of data centers are so great that large projects may need approval from regulators before connecting to the power grid. But instead of waiting for that approval, tech companies keen to build the infrastructure for A.I. are often building gas generators on site. By one estimate, more than 100 gigawatts of onsite gas power, the rough equivalent of 100 large gas power plants, is planned for data centers across the U.S.

[My note: Gas turbine manufacturers have a backlog of ~7 years, as of this comment, see https://news.ycombinator.com/item?id=47493501 for details, plans are not installed gas generators]

US data centers turn to on-site solar, storage, VPPs as grid delays mount - https://www.pv-magazine.com/2026/08/06/transmission-logjams-... - August 6th, 2026

- A Reuters Events report highlights how multi-year transmission delays are forcing data center developers in the United States to deploy 56 GW of on-site generation while turning to Virtual Power Plants to unlock fast capacity.

Data centers are outgrowing the grid. Solar is filling the gap. - https://pv-magazine-usa.com/2026/05/11/data-centers-are-outg... - May 11th, 2026

- Solar’s appeal in this context is logistical. It’s one of the few generation sources that can still be built at meaningful scale on a timetable the market can actually use. The EIA projects 86 GW of new utility-scale capacity will come online in the US in 2026, with solar accounting for 43.4 GW and battery storage another 24 GW. Those aren’t technologies that data center developers are tentatively exploring – they’re what the grid is predominantly being built from right now.

Scaling clean energy and reliability in Michigan - https://blog.google/innovation-and-ai/infrastructure-and-clo... - March 17th, 2026

- Google’s data center operations will be served by 2.7 gigawatts (GW) of new resources for the grid, including solar power, advanced storage technologies and demand flexibility. This Clean Capacity Acceleration Agreement with DTE (the same structure as the Clean Transition Tariff) will bring new, clean resources online, while supporting the state’s transition away from coal-fired power. As part of our standard approach to building new data centers, Google will fully cover its electricity costs and infrastructure needs, helping to ensure that its growth protects local ratepayers and actively bolsters the long-term resilience of the state’s electricity grid.

Google to build data center in Minnesota with new solar, wind power and battery storage - https://www.cnbc.com/2026/02/24/google-to-build-data-center-... - February 24th, 2026

- Google will deploy 1,400 megawatts of wind power, 200 megawatts of solar and 300 megawatts of battery storage to the grid under the agreement with Xcel. The renewable projects will be owned by the utility. They are expected to come online in 2028 and 2029, Shea said.

reply
great, but you haven't answered the question -

How many data centres can you name that use wind turbine and solar panels instead of a grid connection?

Every single data centre above uses a grid connection.

Also, pasting walls of text does not meaningfully advance a discussion. This is just a poor appeal to authority.

reply
What's the fundamental difference between:

Company builds datacenter that needs 2GW and a solar/wind/battery farm that produces a steady 2+GW then wires them together

And

Company builds datacenter that needs 2GW and a solar/wind/battery famr that produces a steady 2+GW and wires both datacenter and generation into the grid?

The later is what the wall of text is demonstrating.

reply
Okay, good luck to you.
reply
> CATL Sodium Battery Hits LFP Cost Parity in 2026: 15,000 Cycles, 30-Year Life - https://news.ycombinator.com/item?id=50399657 - November 2026

That's probably the most retarded piece I've ever seen pick up on HN. Other than insane, unrealistic cell costs, its whole thesis was like "well most of battery cost is not cell, it's the packaging (they have it 5x'd), so adding another $50 per kWh to overall product costs means price parity". Bitch that's $50 on top of a $50 cell. That's 2x the cost, nowhere near "parity".

reply
Batteries are cheap enough even discounting that citation, was the point of the citation.

Global battery industry grows 83% in last five years, boosting deployment to over 300 GW - https://www.ess-news.com/2025/10/28/global-battery-industry-... - October 28th, 2025

https://news.ycombinator.com/item?id=44513185 (“LFP stationary battery storage appears to be at ~$52/kWh in China”) - July 2025 ($35 for the cells alone, as you mentioned https://news.ycombinator.com/item?id=44514720)

TLDR Batteries are now cheap enough to replace fossil generation when paired with low carbon generation.

reply
The thing you have to understand is that you like, don't really make any money as a farmer.

It's not about "where can we put solar panels", it's about "how can farmers lose less money farming".

If you can put solar panels on a pasture and not significantly impact the agricultural value, it's another revenue stream which will help you survive the next agricultural crisis that bankrupts the next swath of small farmers (the next one being the one we're currently in, where diesel prices are up and ag prices are down).

reply
This should be the first comment. So much of the discussion here doesn't get that.
reply
A guy I know works on a sustainability research farm. They basically spend all their time trying out different strategies and technology to figure out how farms can be sustainable without sacrificing productivity. It’s a non profit. They’re well respected in the surrounding farming community, and they get other farmers to adopt new practices.

Anyway there’s my ethos. This farm has solar panels only on top of grazing pastures, cause it keeps cows cool. They’re not like blotting out the sun or anything. It’s a productivity and health boost for the cattle, and renewable energy that runs other systems on the farm.

reply

  >The world has plenty of barren, desert areas
Often (but not always) these areas are actually degraded by human influence.

In these cases, we should regenerate the ecosystems we destroyed, not treat it as "free" surface area that doesn't matter just because we damaged it.

No sane Antique Shop has a "you break it, it's now yours to do whatever you like" policy. I wonder why? :)

reply
> you break it, it's now yours to do whatever you like

I mean, you break it, then you pay for it/deal with the consequences, then the broken pieces are yours to do whatever you like with. That is generally how it works.

I don't think leaving the land as-is is going to be better than putting solar panels on it. In fact, is seems like solar panels might be better than leaving the land alone: https://www.ecoportal.net/en/sahara-solar-panels-double-sahe...

reply

  >I don't think leaving the land as-is is going to be better than putting solar panels on it
I agree, but who said "leave as-is?" I said we should regenerate the land.

If some folks think it's hard to regenerate desert land, it just goes to show that perhaps we should be learning more about ecosystems and how they work.

https://www.youtube.com/watch?v=WCli0gyNwL0

https://www.youtube.com/watch?v=f-sRcVkZ9yg

reply
It is hard to regenerate desert land. It takes a massive amount of coordinated effort. But it's possible, and quicker than one might expect (a time scale of decades rather than centuries, and thankfully significant benefits even within a few years).
reply
My country has none. It’s all forests and grassland. Plenty of solar farms with sheep grazing on them. I quite enjoy using locally generated electricity as opposed to depending on some other country for our energy needs. We still do that ofc for oil and gas as we don’t have those.
reply
The world has e.g. Sahara, barren areas all over Middle East, parts of Central Asia, etc, within or close to the tropical area, and reasonably close to industrial and post-industrial consumers. These place would be perfect to produce solar energy year round, and have little if any competing uses.

But do you think it's the lacking technology that prevents turning vast areas of Syria, Libya, Iraq, Israel, Yemen, Pakistan, or, say, Congo into colossal solar powerhouses? One can name various reasons, but I'd say that they all boil down to the current lack of wisdom in humanity as a species.

reply
Land in the desert is not expensive. If you think this is a wise investment, no one would stop you from doing it yourself and getting rich.

If it were the wise thing to do financially, the Sahara would already be covered in panels.

reply
Like noddingham said, avoiding long distance transmission is one reason. There is a lot of opposition to new electrical transmission projects, which makes them slow and difficult to construct, even in countries that otherwise support renewable energy (like Germany). In the United States, the SunZia transmission line entered operation this year:

https://patternenergy.com/sunzia-comes-online/

It took 3 years to build from start of construction but 20 years from start of planning:

https://en.wikipedia.org/wiki/SunZia_Wind_and_Transmission#H...

reply
If they have an ecological benefit, what's the problem? Just looks?

Most of these "grasslands" were formerly a rich ecosystem of different grasses, bushes, trees, etc. We've mowed it all down and converted most of them into monocultures, basically deserts with grass instead of sand. So putting in some shade isn't necessarily degrading these systems, they're already degraded. If the panels provide shade and places for birds and insects to live, it's improving the situation.

reply
Grazing animals like shade too, like silvopasture. Might as well be something that also benefits humans?
reply
That’s why Europe keeps talking about putting solar on North Africa.

It’s not a bad idea but comes with a number of substantial challenges so hasn’t seen much traction this far

reply
There's an obvious problem with this politically: now you're potentially very dependent on that energy from those other countries. North Africa isn't exactly the most politically stable (or democratic) region in the world.

We saw what happened when Europe was very dependent on Russia for natural gas. Getting (mostly) off of it was possible, but certainly not a fun time.

reply
The technical challenges could be overcome, but geopolitically this would be a disaster.

Why would Europe want to make itself energy-dependent on another continent? It's unnecessarily risky and prone to disruption by hostile nations.

The move from fossil fuels to renewable energy sources is an amazing opportunity to gain back our energy sovereignty, so let's not fumble it by putting our energy generation physically outside of our control.

reply
That's fine for places in or near deserts, but other biomes exist too, often far away from any deserts.
reply
"Barren land" is a name for the most sensitive arid ecosystems on the planet. Solar farms would upset them drastically. So there's that.
reply
deleted
reply
Dunno where you live, but here in California pretty much every square foot of undeveloped earth is covered with some kind of foliage. Grasses, shrubs, chaparral, forests, etc. Grass is just another ground cover, it grows by itself.

There's nothing really "barren" in a natural landscape. Maybe dry lakebeds?

reply
Not a botanist, but it feels like there's a considerable gap between a square foot of earth sometimes supporting substantial plant life, and being "covered" with foliage.

Especially in southern/eastern California - https://en.wikipedia.org/wiki/Deserts_of_California

reply
Sure, but 1) there's still an amazing amount of life when you look up close and 2) desert ecosystems are more fragile than grasslands. Grass grows "like weeds" because that's exactly what it is. There's no risk of running out of grassland.

The big problem putting solar panels over grass is that it dries out in the summer. Inevitable grass fires will damage your panels. Grazing helps with that. Or you can mow.

reply
Barren land? Where's that? Are you thinking US deserts?

https://www.fws.gov/project/wildlife-friendly-solar-energy

reply
grassland tends to be nearer population centres.
reply
As others have pointed out, transmission. But also accessibility - being able to service these things is important.
reply
Why would it be cheaper? That's not a given.

All sorts of reasons make sense though:

* farmer wants supplemental income - if X fewer sheep means you can have > Xprice_of_sheep dollars from solar. Not to mention diversifying income streams.

Transmission losses are real.

* not everywhere has access to a desert.

* A lot of places are barren "deserts" due to overgrazing, and the overgrazing causes a lack of shade that is a positive feedback on desert conditions. Solar panels can provide just enough shade to bring back plants and grazing area.

reply
Because we don't all pool all land available to humanity all over the world; very often, a town, a family, a company or even an individual have some plot of land, and it may make sense to use it for solar electricity production rather than purely for agriculture or animal husbandry (or whatever).
reply
Another thing I'm wondering - would putting solar panels in the desert help moisturize/greenify the desert over time because of the extra shade?
reply
Yes, but keep in mind the ecological impact. We see deserts as dead and devoid of life, but many plant and animal species thrive in these "barren" climates, so changing them as its own adverse effects and pushback.
reply
Deserts are not completely dead, but desertification in general is a way bigger problem to ecological diversity than loss of some desert area. In general, the more water some land has, the more species can thrive in. And deserts are growing, we really do not need to protect them unless in very special cases, we rather need to protect the not yet desert land from becoming one.
reply
Yes it does. China has done it big scale and studies show this effect. (But it does not happen overnight)
reply
I think it should yeah.

Many efforts to reverse desertification, plant trees and grow stuff/crops in the semi-shade. Shade does a lot for keeping moisture in the ground, but it’s much less effective than tree cover, as the roots draw water up from deep below etc.

reply
deleted
reply
[flagged]
reply