Hacker News Reader

Comment by imgabe | Hacker News Reader

Why are we not building giant solar arrays in the southwest and high voltage DC lines to send the power to population centers?

View on HN

My guess: We run into problems here in the Netherlands now that everybody with solar panels is producing huge amounts of power around noon... And the grid can hardly take it anymore (sometimes people can't push the power to the grid anymore even). We can't store it, and most people don't have smart devices that can start (like laundry) based on power production.

That problem should first be solved.

Edit: In some provinces we can now neither build more power plants nor add any larger industries because the grid can simply not take it. I guess decentralization is key. For myself I'm looking into a small EV that is usually at home and can be charged during the day. But this only works when the car is at home during day time. Another pro-wfh argument ;). Electric scooters (max 45 kph), electric bikes (max 25 kph) and "speed pedeleces" (max 45 kph, they are in between scooters and bikes [0]) are also becoming really poplar here, they use comparatively little, but it's nice when they replace cars.

[0]: https://www.speedpedelecs.com/ (it's a brand and a category at the same time)

well, luckily for the US, we have massive hydro power dams in the west that are low because of 20 years of drought, so we can actually do pumped storage.

Ideally, we would pump salt water from oceans to the salt water reservoirs and use surplus energy to desalinate to refill the fresh water ones.

Better, just pump sea water to the desert, and let it spread out and evaporate. The mountains will catch the water vapor and fill your reservoirs.

Releasing in Death Valley, you don't even need to pump. A siphon once started will just run indefinitely. (The high point of the siphon would need to be less than 30 feet above sea level.)

You might need to bulldoze up the salt in the fall and take it somewhere.

My thoughts exactly. Solar -> Desalination, with all other energy consumption in the middle. Granted, I am not up to date on desalination. I know MIT released something recently but idk the details. Is energy the bottleneck now or is it still material?

Giant solar farms (like Solar Star in LA) will connect to the higher voltage grid rather than the low voltage grid used domestically. Exporting large amounts of power and moving that power geographically is exactly what this is designed for. There may be issues with lack of capacity in district and national grids. But it is a different problem to the domestic solar issue. It can be solved by reinforcement of existing lines, upgrading substations and building new ones.

The main issue with our grid is mostly that we lack the capacity to transport the power. In the past power generation and heavy power consumption were typically relatively co-located. Now, with wide spread energy production with solar panels the energy is coming from everywhere and this power needs to be transported to where it is consumed.

Moreover, in order not to overload the grid other power generators need to scale down, even if those would be closer.

This is not true. Enron was famous (notorious) for leveraging the, then, capacity of long distance interconnects to move power from far off generation capacity to advantageously priced markets. Just take a look at the sheer size of CalISO, MISO, and ERCOT. You could build solar anywhere and wheel it to where it’s needed. That’s not the long pole in the tent (right now).

My reply was to someone talking about the grid in the Netherlands. That was what “our grid” referenced.

Furthermore, we have three levels of grid. The very high voltage, highways, the regional high voltage, aNd the local grid. Each have their own challenges. For Solar it is mostly the local grid that gets overloaded. On my connection I’m hitting 250V on sunny days instead of the standard 230.

All fair points. I assumed utility scale solar in the US.

Exactly this, (lack of) storage is the problem. Daytime demand isn’t high enough to deal with PV output leading to cut offs on grid feed-in from consumer PV installations. Given that car batteries aren’t an option yet to power homes this is a problem that will stick around for a bit.

>Given that car batteries aren’t an option yet to power homes this is a problem that will stick around for a bit

One was posted here not too long ago.. so this may not be far off. But it's a bit of a chicken-egg situation because powering your home with a car battery is not very practical by itself. Investing in both solar and the car at the same time would also be a hefty chunk of change for any individual household

That’s exactly the point. The few car models that do support grid feed-in require an expensive converter to take DC from the car and deliver AC to the home. The hardware to do that is just expensive and currently that doesn’t pay off

Hardly anybody uses car batteries at this point, and not just because of availability. It's a bit of a pain in the ass compared to just buying LFPs. Even at fairly high retail pricing for complete batteries, 30kWh of LFP is $10K. That'll power a typical home for a day. If you just buy cells and BMSs, it's cheaper. That's what most of the off-grid folks do these days.

The thing is that you don’t want just the car battery but you want to use the capability of the car to deliver power back to the grid (or your house at any rate). The rest of the time you want to use the car as a car. The idea behind it is that you don’t need a separate battery pack at home for the times the grid is over-subscribed.

Car batteries (E.g., in Nissan Leaf) are, in fact, an option to power homes. A controller/converter box is under $900.

I would say that home batteries like the Power Wall seem like a nice solution.

However I am a bit worried about the lifespan of those batteries and what happen to them at their end-of-life (recycling? CO2e cost of making a new one?...)

For standalone houses, seems like a bank of traditional lead-acid batteries like datacenters use is a better option. Doesn't require exotic materials, EOL handling is well understood at this point, etc.

If I'm doing the math right, every ten car batteries in the bank should get you about 8 kWh. The average home in the US uses about 11,000 kWh/year[1], or about 30 kWh/day, so a bank of 20-60 batteries seems like a good starting point, depending on how much reserve capacity one is comfortable with, how variable solar generation is, etc. That's an up-front cost of about $4,000-$12,000 every 5-7 years, but at least at the low end that should actually be cheaper than paying for electricity from the grid over the same period. A 20-battery bank should fit in about the same footprint as a refrigerator.

That's also significantly cheaper than a PowerWall of the same capacity[2], which was $7,500/14 kWh ($535/kWh versus about $250-$275/kWh for lead-acid) before Tesla stopped selling them without a solar panel bundle.

For apartments and other colocated housing, it might still make sense as long as there was some sort of central vault for the battery bank.

[1] https://www.eia.gov/energyexplained/use-of-energy/electricit...

[2] https://solarmetric.com/learn/tesla-powerwall-review-costs-s...

Thanks for the thourough estimations!

However I strongly advise against buying lead batteries, which are very dangerous for the health and environment (arguably not while they have not yet reached their end-of-life, but once expended a lot of them end up being dumped into the wild, and I am not even confident in their recycling).

Use the spare power to pump water up mountains.

Maybe the Netherlands have excess solar power in some days/hours, but what about neighboring countries?

In the end, I'm confident little power is wasted; instead, power is sold on the spot market at low and sometimes negative rates. Datacenters will gobble it up to lower their prices too, making "bottom feeders" like crypto miners hit their "I am willing to pay this much" prices.

Then there's of course various energy storages; I believe one thing they want to implement is that any leftover electricity is put into generating hydrogen gas, which can be stored and later burned cleanly to generate power if needs be.

But yeah, that's a bunch of rambling from an amateur who amortizes grid capacity; in NL we have a problem that the grid is full. It's not really affecting day to day things yet, but it means that new companies - power generating or consuming - are not being connected to the grid because it would cause overloads. I'm not sure if they bleed power off anywhere yet.

There are numerous options for super-cheap energy storage. They just aren't built out yet, because the money is better spent, still, on adding generating capacity. What is being built is factories to produce them.

Excess renewable generation is not a problem; if there is too much sloshing on the grid, you turn them off.

I'm sure that the solution is rather storage than cryptomining (to me it looks like the same as just burning the energy into waste heat).

For example, molten salt storage.

Germany, Belgium and Denmark also have quite a bit of renewable energy and the sun is usually similar, so they usually don't really need it when the Netherlands has too much either.

Building power transportation networks makes building out large solar arrays look cheap. Getting the right of way and environmental impact statements for these things that cross rivers and cities is a massive undertaking.

We are, see the multiple square mile installations in the Mojave desert (in particular Rosamond). Output of that region is ~520 MWh/acre. https://www.google.com/maps/@34.8097025,-118.4587554,15367m/...

Side note, land in the area is available for purchase at $3k/acre. And if you look closely, you can see the plots where classic subdivisions were initially scratched into the surface before land owners realized no one wanted their land even at 3k/acre. https://www.google.com/maps/@35.1339826,-118.0179071,2422m/d...

Edit: when I click the link, Google Maps helpfully informs me that traffic in the region is "light". :)

Desert solar arrays will come to be seen as a dumb idea. But not before we build even more of them. They have to be dusted, and because of the heat, efficiency suffers and they don't last very long.

The place for solar, at maturity, will be sharing farm and pasture land.

“We” aren’t building anything. Utilities are an income producing asset for their investors. There are, in fact, a multitude of utility solar installations going on right now all over the US. These things take time to build and learn from. It’s not a slam dunk and the profits aren’t guaranteed. Not withstanding the unsolved engineering challenges, of which there are many startups exploring.

We’re not stupid it just takes time.

We already have built many solar arrays in the southwest. And the power generated does get sent to population centers, both from the renewable and the non-renewable plants. If you are asking why we don't expand those efforts, that is a good question... but it is absolutely already in place.

It's inefficient. HVDC is _more_ efficient over long distances, but not enough to make it worthwhile

Because it's not economically viable, yet. It'll happen when it happens.

Reply on news.ycombinator.com