Almost Plausible Solar Takeover Plan

[anorlunda]

It would certainly be interesting to see a more detailed dive into a potential 100% PV future.

See #26, the eggs in one basket problem.

 

[Lord Crc]

I would be even more interested in a 100% clean energy grid deep dive, but I got the impression you wanted to rail the thread back to the original premise of 100% PV :)

 

[anorlunda]

I would be even more interested in a 100% clean energy grid deep dive, but I got the impression you wanted to rail the thread back to the original premise of 100% PV :)

Maybe your right. Limiting the scope of this thread won't work. Go for it.

 

[russ_watters]

Quick clarification; the title may give a false impression, but the article/model are about a 100% renewables grid, not a 100% solar grid. So, for example, the stated 4.3x solar overbuild may actually much larger after the other sources are subtracted out.

I'm disappointed I don't get to see the details, but my guess is that last MW of solar only produces 10% as many MWH as the first, so what they are saying is storage is 10x too expensive to be viable and can only be used when it is actually required, independent of economics.

edit: Er...or maybe it's simply that the amount of solar needed to satisfy the grid in winter yields a 4.3x overcapacity in summer. That the summer-winter capacity issue is worse than the day/night one.

 

[anorlunda]

so what they are saying is storage is 10x too expensive to be viable

My reading is that with overcapacity as the base case, then the amount of storage is capped at the amount needed until the next day. With anything less than complete overcapacity, then the worst case need for storage gets into the multiple day region, and there is no definitive way to define the worst case storage required.

Maybe it appeals to me because I'm an analyst. Rating PV capacity by the worst case day greatly simplifies the analysis. Said in other words, it makes PV like more like base load generation.

 

[russ_watters]

Yeah, see my late edit; maybe it is more a matter of physical need than cost. A quick google tells me the summer-winter output variation in my area is 5.6x and the summer-winter load variation is 1.4x. Divide and you get 4x. That's how much peak summer overcapacity you need to satisfy the average midwinter load.

 

[anorlunda]

That's how much peak summer overcapacity you need to satisfy the average midwinter load.

I think that's exactly right. But I did not see in the article where they made allowance for catastrophes such as hurricanes or ice storms, where a significant fraction is destroyed and needs replacement. That's why I doubled the 400% to 800% in the OP.

War is also a necessary consideration. Recall that during the 2003 war in Iraq, the allies used weapons specifically designed to attack the power grid. Without any specific knowledge of future weapons, we must assume that such weapons will exist. The primary defense is diversity. Also in the OP, I mentioned cybersecurity as an advantage of solar PV. Cyber attacks sound highly likely.

 

[mfb]

Yeah, see my late edit; maybe it is more a matter of physical need than cost. A quick google tells me the summer-winter output variation in my area is 5.6x and the summer-winter load variation is 1.4x. Divide and you get 4x. That's how much peak summer overcapacity you need to satisfy the average midwinter load.

Bad news for Europe then. We have similar summer to winter production ratios, but demand is ~10-20% higher in winter.

 

[Svein]

That's how much peak summer overcapacity you need to satisfy the average midwinter load.

Yes - and as I have said repeatedly, the simplest way to store intermittent power supply is to use it to pump water up in an elevated dam. Then use it to generate power whenever you want.

 

[artis]

pardon if this comes across as ignorant, I'll admit I haven't put in a ton of research into this.

if space was not an issue like in the deserts or rural areas could we use flywheel storage to compensate within 24h time frames the overproduction at day and under production at night problem?
I am not sure about the energy density and cost comparison between battery technologies and flywheels but as for longevity and robustness I'm sure flywheels take the upper hand.

 

[Svein]

if space was not an issue like in the deserts or rural areas could we use flywheel storage to compensate within 24h time frames the overproduction at day and under production at night problem?

Friction...

 

[anorlunda]

if space was not an issue like in the deserts or rural areas could we use flywheel storage to compensate within 24h time frames the overproduction at day and under production at night problem?

It has been researched and it has been tried. There were even a few commercial ventures based on flywheels. But none of them have been a success. I don't know the details of why.

Is there a hill nearby? If so then there is a variant on pumped hydro that's easier to achieve, able to scale, and with lower capital costs. Use electric railroad trains with regenerative braking. The trains go up/down the hill just as water goes up/down in pumped hydro.

 

[artis]

@Svein well friction yes but now we have magnetic bearing systems and other methods so in theory one could overcome this. Although probably not the most compact energy density wise.
@anorlunda I am also not aware of any existing large scale flywheel storage systems to be honest

 

[anorlunda]

I am also not aware of any existing large scale flywheel storage systems to be honest

I still don't think they are particularly successful, but they do exist.

It only took 60 seconds searching to find these:

Beacon Power opened a 5 MWh (20 MW over 15 mins)[18] flywheel energy storage plant in Stephentown, New York in 2011[48] using 200 flywheels[49] and a similar 20 MW system at Hazle Township, Pennsylvania in 2014.[50]

A 2 MW (for 15 min)[51] flywheel storage facility in Minto, Ontario, Canada opened in 2014.[52] The flywheel system (developed by https://en.wikipedia.org/w/index.php?title=NRStor&action=edit&redlink=1) uses 10 spinning steel flywheels on magnetic bearings.[52]

Amber Kinetics, Inc. has an agreement with Pacific Gas and Electric (PG&E) for a 20 MW / 80 MWh flywheel energy storage facility located in Fresno, CA with a four-hour discharge duration.[53]

Beacon Power opened a 5 MWh (20 MW over 15 mins)[18] flywheel energy storage plant in Stephentown, New York in 2011[48] using 200 flywheels[49] and a similar 20 MW system at Hazle Township, Pennsylvania in 2014.[50]

A 2 MW (for 15 min)[51] flywheel storage facility in Minto, Ontario, Canada opened in 2014.[52] The flywheel system (developed by https://en.wikipedia.org/w/index.php?title=NRStor&action=edit&redlink=1) uses 10 spinning steel flywheels on magnetic bearings.[52]

Amber Kinetics, Inc. has an agreement with Pacific Gas and Electric (PG&E) for a 20 MW / 80 MWh flywheel energy storage facility located in Fresno, CA with a four-hour discharge duration.[53]

Convergent buys up 40MW of flywheels in New York and Pennsylvania

 

[mfb]

The German Wikipedia has a nice table of typical values. flywheel = Schwungrad, they even have two types, steel and carbon fiber. Both have way more capital cost per stored energy (€/kWh column) than batteries. €1200/kWh means you need over 10000 cycles before this becomes somewhat interesting. Even if you can make a full cycle every day and can sell the electricity at 10 cent/kWh more than you buy it you need 30 years to recover the initial investment. If you can make a cycle every hour then things get much more interesting, but the grid doesn't need that.

 

[anorlunda]

If you can make a cycle every hour then things get much more interesting, but the grid doesn't need that.

Some of them offered frequency regulation service, where plus/minus responses at the time scale of seconds is important. They work OK for that purpose, but the prices we pay for frequency regulation are low.

All the other forms of generation, including wind and solar, can devote a fraction of their capacity to modulate for plus/minus frequency regulation. That makes lots of competitors and low prices. Low prices make it a poor business strategy for flywheels.

 

[paradisePhysicist]

From what I understand, opponents of solar energy usually make claims that using green energy uses "too many resources" or "pollutes the environment also". But this is a fallacy because, making regular cars and motors pollutes the environment the same amount as building solar cars and components. So overall, making solar cars would pollute less, since they no longer pollute after it is made.
source: https://understandsolar.com/solar-uses-more-energy-to-manufacture-than-it-produces/

It is said (online) that solar energy will pay for itself after 4 years. This is for the individual home. Worldwide, can we build enough solar panels for every home? Well using rough estimates, there is about a 1:1 ratio for every car and home. So there are about an equal amount of cars as there are homes. And building a car is much harder than building a solar panel. So why not do as many solar panel installations, as cars? What is the world's excuse? To upgrade a home with a solar panel installation would cost about $11,000-22,000. The price of average car costs much more than that, even 3x more at $37,000. So why cannot we build solar panels? From what I read, solar panels are made out of sand, an abundant resource. What about wind power? Well we need copper for the motor for that. There is more sand than copper. Why not give all the people living in sunny areas solar power, and give those in Alaska wind power? It makes perfect sense.
source: https://www.homedepot.com/c/cost_install_solar_panels
https://mediaroom.kbb.com/2019-06-0...ar-for-May-2019-According-to-Kelley-Blue-Book

The current solar panel efficiency is only 18% or 23%. Imagine if we can get it to 50% or even 75%. Building the solar panels can be even greener if the factories that create solar panels are powered by solar panels.
https://www.solar.com/learn/solar-panel-efficiency/

I am not sure if the $11,000 solar installation is overbuilt by a factor of 4. But in places like Alaska, wind power can be used instead of solar. If the $11,000 solar installations are already overbuilt, even better.
https://en.wikipedia.org/wiki/Wind_power_in_Alaska

There are crystals that can slow down light to 17 meters per second. Maybe even some day we can use the crystals to store and release sunlight energy incase there are clouds above the panels. Once computers become photon computers, it may help continue to make solar research even better.
https://phys.org/news/2013-07-seconds.html

 

[hmmm27]

Well, if we're going all speculative...

- for obvious reasons, find an equatorial'ish desert which has decent SiO2 sand.

- because of cheap land and power, and local target sites, build factories in situ, to produce garden-variety solar panels and glass. Locally-powered of course, but an MNR or two would make for a faster and greener ramp-up, for at least the initial factory.

- to use all that insolation in a non-stupid manner, design and build greenhouses with optical filters that pass the frequency range(s) which the crops need at any given time, reflecting the rest onto the (double-sided, of course) solar panels.

- just keep churning them out and emplacing them.

- concentrate local CO2 into the greenhouses. (Maybe ship in CO2 to saturate the hermetically sealed'ish enclosures)

- PV + crops will use 20% of insolation on a good day (but they're all good days ; that's why we're here), so some way of using/ditching the other 80%ish. And, in that regard...

- Install windfarms on the periphery to catch the heightened windspeeds from the massive solar-tower effect from the totally tanked albedo of thousands of square miles of black.

- when there's so much PV installed that it messes up more than just the local weather pattern and - site-specifically - starts pulling in clouds too far, backtrack a bit and start shipping to other equatorial cloudless areas : ones with crap sand.

Infrastructure requirements:
Rail's a good start for the trunks. Water, fresh from the coast, desalinated enroute or at destination (which means shipping salts back). Power lines : lots of. Battery recycling in situ.

etc.

Just sayin'.

 

[anorlunda]

Well, if we're going all speculative...

You missed the opportunity to use the spoiler feature of our post editor.

Spoiler: Warning sarcasm enclosed

Just sayin'.

 

[Rive]

Others have proposed use rather than curtailment before, but they weren't thinking on the scale of 700% of peak demand.

Well, that's my favorite, actually. I mean, doing the curtailment (and grid balancing, at least daytime) through the curtailment of a slightly overbuilt solar capacity.

But overcapacity at this scale... I bet that even just for the cost (end environmental footprint) of periodically replacing the aged panels we could do something better, nicer.

Still, I get the point. Had to have an estimate for the price and method of 100% PV generation. Just... well, what's the next step?

 

[mfb]

It is said (online) that solar energy will pay for itself after 4 years. This is for the individual home. [...] To upgrade a home with a solar panel installation would cost about $11,000-22,000.

Even if this solar panel installation could make the home fully autonomous, that would need an electricity bill of $250-500 per month. No, I don't pay that much, not even close. And solar panels alone don't make the home autonomous.

The price of average car costs much more than that, even 3x more at $37,000.

And the average home costs even more. What's the point of comparing unrelated costs?

From what I read, solar panels are made out of sand, an abundant resource.

You'll find silicon both in solar panels and in sand, but that doesn't mean they are made out of sand. You can find oxygen both in you and sand, does that mean you are made out of sand, too?

Imagine if we can get it to 50% or even 75%.

If you make up numbers anyway, you might as well speculate about 99%.

I am not sure if the $11,000 solar installation is overbuilt by a factor of 4.

It is not.

There are crystals that can slow down light to 17 meters per second. Maybe even some day we can use the crystals to store and release sunlight energy incase there are clouds above the panels.

That's completely baseless speculation.

Once computers become photon computers, it may help continue to make solar research even better.

That makes no sense.

 

[paradisePhysicist]

The average electricity bill is $96. It would pay for itself after 10 or so years.

"The same is likely true of the United States. According to the Energy Information Agency (EIA), every month the average American household has an average electric bill of $95.55, uses 920 kWh, and pays 10.4 cents per kWh. "
www.off-the-grid-homes.net/average-electric-bill.html

Cars do not pay for themselves (unless pizza driver). Solar panels generate energy thus will provide power to future generations. Online it says that low quality solar panels last for a minimum of 25 years. If solar panels are able to last for 100 years, that means your kids, grandkids and great grandkids are going to be saving a lot of money. So after about 25 years your solar panel will be providing 80% its original capacity of output. But if you have a decent quality panel you will still be getting a lot of energy out of it. High quality panels can offer about 0.3% yearly degradation. (And also remember that future tech will be more efficient and batteries will be better. So even though the solar might be at 80%, the electronics and batteries in your home will probably use less power by then also, provided companies of the future aren't intentionally making inefficient products due to planned obsolescence. )
https://www.solarreviews.com/blog/how-long-do-solar-panels-last

I got the 0.3% value from this website however I am not sure their math is correct. I am not sure if degradation rate is a constant or if it is cumulative. This is crucial to determine the viability of solar long term. Also keep in mind that solars will last a lot longer than 25 years, the 25 years number is due to power being at 80% of original.
https://news.energysage.com/how-long-do-solar-panels-last/

 

[mfb]

The average electricity bill is $96. It would pay for itself after 10 or so years.

This is still ignoring the storage and maintenance cost.

If solar panels are able to last for 100 years

You can always make things look more attractive if you make the assumptions unrealistic enough.

 

[russ_watters]

The average electricity bill is $96. It would pay for itself after 10 or so years.

"The same is likely true of the United States. According to the Energy Information Agency (EIA), every month the average American household has an average electric bill of $95.55, uses 920 kWh, and pays 10.4 cents per kWh. "
www.off-the-grid-homes.net/average-electric-bill.html

This is still ignoring the storage and maintenance cost.

Also ignores that in most jurisdictions you aren't legally allowed to go off grid. And if you go with net zero and forgo storage to sell the electricity back to the electric company, you still have to pay for the grid infrastructure parts of the bill. You mainly save on the generation. At least in the US that's how the vast majority do it.

 

[artis]

@russ_watters How can someone not be allowed to have no electricity if he/she wants to say live without it? Isn't that a breach of basic rights ?

 

[russ_watters]

@russ_watters How can someone not be allowed to have no electricity if he/she wants to say live without it? Isn't that a breach of basic rights ?

It's a public health and safety issue. A house is not considered livable unless it has basic utilities that meet certain minimum requirements.

And here's a real punch in the gut for solar: residential systems without batteries (basically all of them) require a grid connection for power regulation because of solar's intermittency. So if you lose grid power due to an outage, you lose electricity, even on a sunny day.

Individual building solar can help, but it is not the panacea a lot of people think it is.

 

[mfb]

@russ_watters How can someone not be allowed to have no electricity if he/she wants to say live without it? Isn't that a breach of basic rights ?

You can't opt out of police, public streets and so on either. Not all places have a such a law about electricity but it is common. With rooftop solar power becoming more popular and decreasing storage prices we might see these laws go away in the future.

 

[russ_watters]

With rooftop solar power becoming more popular and decreasing storage prices we might see these laws go away in the future.

Storage is the key to enabling off-grid operation. A solar system with no storage and no grid tie simply can't self-regulate reliably most of the time - a single cloud shuts you down or worse browns you out (potential to damage electrical systems). And of course no night operation without it.

"Island mode" is fairly common, though, for commercial/industrial sites (albeit not usually on solar).

 

[russ_watters]

...also, this residential discussion is interesting but can't be a total solution. Residential and commercial rooftop solar can take a significant chunk of our electricity, but certainly much less than half. The topic of the thread - grid scale generation - is still required.

 

[artis]

Well where I live I can cancel my electric connection , especially if that is a private house.
If I had a house and some money to invest I'd probably experiment with having solar panels on rooftops and areas that are of no other use and then a wind turbine, given one doesn't need as much electricity during night/sleep this could in theory provide one with a daily 24/7 minimum load, given some battery storage is also used to drive out the peaks and cover the dips/gaps.But all in all not all places or people have the option to make their own power production facilities to feed them so it seems the grid isn't going anywhere and the best we can do is to incorporate solar and wind into the grid as much as possible and solve the storage problem

 

[russ_watters]

Well where I live I can cancel my electric connection , especially if that is a private house.

One needs to be absolutely sure that's true and there is always a risk that even if it is true today it will become false later. The problem is that when residential solar generation was small it was possible to ignore its disruption to the grid, but as it grows enough to become a relevant factor - even at only a percent or two of total generation - those disruptions can no longer be ignored. Laws/utility rates are changing so that fewer and fewer rate structures enable zero-ing out the bill:

Under traditional net metering, customers are only billed for their net consumption over a billing cycle, meaning that any energy they consume from the utility can later be offset by energy production from their solar installation. As a result, even though a customer’s solar installation only makes power for them during the day, they can use excess energy production during the day to cancel out their nighttime usage and drop their electric bill down to nearly zero...

Changes to net metering policies fall into three general categories: 1) changes to how long excess generation credits can be carried forward and applied to future energy charges, 2) the application of fixed energy charges which cannot be offset with solar energy credits, and 3) changes to the value of electricity sold to the grid from a solar installation compared to the value of electricity bought from the grid.

https://blog.aurorasolar.com/how-net-metering-is-evolving-three-changes-you-need-to-know

These changes reflect the reality that even if you export to the grid as many kWh as you get from the grid (net zero energy usage), you're still receiving a service from the grid that must be paid for by somebody.

 

[mfb]

Completely cutting the connection is as stable as it gets (for the grid): No demand, no supply.

 

[shjacks45]

You argue from the viewpoint of someone who assume that the sun is shining "most of the time". Some of us live in countries where that is not true. Try to calculate the solar energy hitting Alaska in the winter months - and then remember that parts of the Nordic countries and parts Russia lies north of Alaska.

I live in Seattle, overcast 260days a year. Overcast day provides 60% of sunny day energy. Bigger impact is length of day at 45’North. Note overcast usually means wind.

 

[shjacks45]

Solar thermal is cheaper low tech that can be used in 3rd world. Like solar frying French fries in Oregon. Large hydroelectric will be around until the dams silt in. Small hydro built into repair of riparian courses has good $$ return. Also geothermal: Alaska, Idaho, California, and how many volcanoes in Italy. VAWT, say at building rooftops, are a compact way to add local capacity. Light offshore wind can be 60 mph gusts in Seattle’s downtown “canyons”. Thermal recycling to capture waste energy e.g. concrete plant waste gasseshe was run an electric generator for the plant. Japanese natural gas fuel cell technology increases natural gas to electricity efficiency. Solar cell farms and wind farms would help maintain control for the energy industry elite, therefore more likely to get congressional approval.

 

[paradisePhysicist]

thought of another idea, what if you put a bunch of mirrors to boost power on cloudy days... these mirrors work like a periscope to reflect off two angles, there is no limit to how many mirrors cause you can spread 'em apart at any distance...

i am wondering will this work for more help than just cloudy days?... for instance, can a solar panel use the power of two suns, or is it maxxed at only just one sun before the max power is achieved?... if you can get solar panels to accept the power of 10 suns or more the power is astronomical, assuming the actual photon waves do not cancel each other out...

if this idea is useful please consider giving me a nobel prize someday...

 

[shjacks45]

It's a public health and safety issue. A house is not considered livable unless it has basic utilities that meet certain minimum requirements.

And here's a real punch in the gut for solar: residential systems without batteries (basically all of them) require a grid connection for power regulation because of solar's intermittency. So if you lose grid power due to an outage, you lose electricity, even on a sunny day.

Individual building solar can help, but it is not the panacea a lot of people think it is.

Several local installations I've worked around (farms, agricultural, medical, etc.) have both wind, solar, and generators for backup power (the PNW often has power outages even in major cities.) Often cloudy weather brings wind, compensating for lower solar efficiency. Small Wind VAWT also work at night. Coastal cities like Seattle can depend offshore and onshore winds regularly. One customer even added "small hydro" to their power mix. Cloudy days affect the average Solar array less than you think. Usual Solar home/business rooftop array faces South (in the Northern Hemisphere) and is set at a fixed angle based on location latitude. Optimized for noon twice a year, but a pragmatic (cost) compromise. On a cloudy day solar radiation is dispersed and arrives from all direction. An array that follows the Sun will suffer more power loss than cheap fixed array. Generally, cloudy days provide 60% of the power of sunny days. Note that cloudy days are generally late fall, winter, early spring, whose days are shorter than the other half year.

 

[Tom.G]

thought of another idea, what if you put a bunch of mirrors to boost power on cloudy days... these mirrors work like a periscope to reflect off two angles, there is no limit to how many mirrors cause you can spread 'em apart at any distance...

Like this, perhaps? For a sense of scale, see the cars in the parking lot at lower right.

--

--
That is Solar One in the Mojave Desert (California). It was retired in 1988. This is a Solar Thermal unit. The top of the tower, operating at 1000°F, has molten salt circulating in it as a thermal transfer fluid. Then water is boiled to drive a conventional steam turbine/generator system.

Photovoltaic cells are destroyed at those temperatures.

Solar Two and a bit more info at:
https://www.atlasobscura.com/places/solar-one-and-solar-two

Also:
https://climatekids.nasa.gov/concentrating-solar/
But they are V-E-R-Y slow to connect, it may take a few tries.

Cheers,
Tom

 

[mfb]

@paradisePhysicist: With PV this is generally known as concentrator PV. You save some PV area but you increase the complexity of the system, you generally need tracking to follow the Sun, you create heating issues, and it only helps with direct sunshine - on cloudy days your mirrors are not brighter than the sky they obstruct.

 

[anorlunda]

The economics are changing so fast. I suspect that we are already at the point where the cost of one square meter of mirrors is about the same as one square meter of PV panels.

Three years ago, the price of panels (not including installation) was $1/watt. Now it is about $0.12.

 

[shjacks45]

Like this, perhaps? For a sense of scale, see the cars in the parking lot at lower right.

--
View attachment 272590
--
That is Solar One in the Mojave Desert (California). It was retired in 1988. This is a Solar Thermal unit. The top of the tower, operating at 1000°F, has molten salt circulating in it as a thermal transfer fluid. Then water is boiled to drive a conventional steam turbine/generator system.

Photovoltaic cells are destroyed at those temperatures.

Solar Two and a bit more info at:
https://www.atlasobscura.com/places/solar-one-and-solar-two

Also:
https://climatekids.nasa.gov/concentrating-solar/
But they are V-E-R-Y slow to connect, it may take a few tries.

Cheers,
Tom

Gotta note the solar thermal plant that Ore-Ida built in Eastern Oregon to fry french fry potatoes.

 

[Tom.G]

Gotta note the solar thermal plant that Ore-Ida built in Eastern Oregon to fry french fry potatoes.

Will do... but a link would speed things up a bit, please?

 

[shjacks45]

At 1000W per meter squared solar irradiation, for 1 TW that is about a square 30 x 30 km each side if the collection was all in one concentrated localized area.
Of course, solar panels do not collect 1000W per square meter, so the land area required increases substantially. There might be some "not in my backyard" issues to overcome, just because. ( elimination of the sun shining on my spot of land ).
But the amount of land used presently for particular endeavors such as transportation road network, golf courses, mining, just to name a few, usually isn't an issue most people contemplate. But I still suspect location of the solar farms, or panel placements ( such as rooftop or roadway right of ways ) might spark some discussion.

DOT is building walls around the freeways for noise abatement. Dessert areas for solar could fish farm (uses less water than farming) beneath panels and wind. Flat desserts are windy. Places like Denmark use VAWTs on building roofs. Coastal cities like Seattle get daily onshore and offshore breezes. Stormy weather makes less small solar, however typical home (fixed, facing south, angled ~40 degrees) gets 60% sunny day power on Seattle's 260 overcast days a year. Main problems with solar are rapid output decline after solar cell installation and the cost and enviromental costs. Most power used in Seattle area is winter home heating, an inefficient way of using solar energy. Should be more like direct solar: solar water heaters. I have installed "small hydro"; small hydroelectric is being shut down by energy conglomerates as they concentrate their power. This as 15% of power plants (nuclear) will reach end-of-life in next 5 years. Many dams built in the 1930s-1950s will be silted in in 20 years. And how about that Yellowstone caldera. It blew up 100s of megatons worth in the past. Or we could cool it down using Geothermal wells. The heat energy in Yellowstones magma exceed the energy of US coal reserves. Instead of water/steam, direct infrared to electricity "solar cells" (Thermoelectric cells) could be used.
I grew up on an american farm with a windmill that made electricity and pumped our water. We could charge the batteries from the tractor if necessary, but never did. We lived in the Mohave desert and had near constant wind coming down the mountains.

 

[shjacks45]

Will do... but a link would speed things up a bit, please?

Heinz bought Ore-Ida but still using Solar
https://digital.library.unt.edu/ark:/67531/metadc1071603/

 

[Fisherman199]

I have yet read the article, but I do know this: The drawings are already submitted and ground is being broken for infrastructure to accommodate very large solar projects all over the southeast. Big things to take place in the next few years. I would also expect governmental forces to play ever more apparent roles in these projects now that Biden is president apparent.

 

[shjacks45]

This article:
Overbuilding solar at up to 4 times peak load yields a least cost all renewables grid.

I'm normally among those who think that renewable advocates lack realism when they advocate 100% wind+solar grid. This article is the first I've seen that comes close to being plausible. I do doubt the article's numbers for storage and for wind, but the basic idea of massively overbuilding solar is in the right direction.

The basic idea is this. Actual solar generation depends strongly on the season and the weather. But even on the worst case day (winter solstice, thick clouds, and heavy snowfall), solar panels produce a fraction of their rated power. If there is enough overcapacity, then even that fraction will be big enough to supply the demand on that day plus recharge the batteries for that night. No heroics, special tricks or cleverness are necessary. Simplicity and reliability are the keys to successful power.

The article says 400% overcapacity is enough. I say 800% for the sake of argument and to allow for contingencies like hurricanes and ice storms. That is roughly 8 Terrawatts for the USA. At $1/watt installed, that's $8 Terradollars investment we need for the generation part (plus ? for storage). That's nearly 40% of the pre-COVID USA GDP. Big numbers, but conceivable when spread over a number of years.

Side issues:
Terrestrial wind in some regions is subject to 2-4 consecutive weeks with winds <15 knots, so wind production would be nearly zero. That makes terrestrial wind overcapacity less attractive than solar overcapacity. Offshore wind is more dependable, and thus more attractive, but it is only close to the coastline by definition. This whole idea is much easier to visualize with solar.

Solar capacity could be distributed across the continent close to the load centers. Therefore, massive new investments in power transmission or distribution would not necessarily be needed with this change.

What to do with the massive solar overcapacity in summer? The article says "curtailment" meaning shut down portions so we don't generate energy that we can't use. But the excess capacity could be used to produce hydrogen or fresh water production by desalinization. Hydrogen and desalinization are not economical in most circumstances, but if the alternative is curtailment of excess capacity, we might reconsider their economy.

Others have proposed use rather than curtailment before, but they weren't thinking on the scale of 700% of peak demand. The economies of scale would be very significant.

The required overcapacity would be less in southern states than northern states. Fresh water is more scarce in southern states.

We already have the technology (called synthetic inertia) to make solar panels behave transiently like old fashioned steam turbine generators. That allows a non-disruptive transition with respect to grid operations and control. No matter what the ratio of solar to conventional generation, the dynamics of the grid would remain nearly constant.

Of course it would take lots of engineering to study the actual feasibility of this plan. Perhaps 50 engineering-man-years to do the study. If I was not retired, I would bid for the study contract myself I know exactly the team to do it. It is a study that I believe should be funded.

p.s. It's fun to have something other than COVID-19 to think about.

A CEO acquaintance invests in clean energy. His company has small and medium hydroelectric (waterfall bypass like Niagra), wind properties, geothermal, The idea of covering useless dessert land with solar cells fails to understand WHY the land isn't used. The US Southwest is geologically "basin and range" terrane, not flat like Midwest farmland. It's not too dry to farm, it's too rocky. You can actually see the windmills at the Tehachapi wind farm from space (or google satellite view). But a closer view of the Tehachapi mountains shows jagged hillsides not usable for much else. Coastal cities like Copenhagen have seen use of building-top VAWTs which complement Solar to generate in bad weather as well as good. Geothermal is overproducing for Iceland, the Yellowstone Caldera is over a mantle hot spot (like the Hawaiian Is.) So sucking all the energy out of it is unlikely. Urgent because current nuclear plants generate 15% of US power and reaching engineered end-of-life. Many US hydroelectric projects date back to 1930s and dams fill up with silt over time, currently provide 5% of US generation capacity. Of course coal still generates 40+ % of US electricity.

 

[Fisherman199]

I read the article and slept on it. I have many questions but the basic assumption is not wrong (which is not the same as being good and is further yet different form being correct). I still think the article borders on issues with accepting realism but the numbers check out. I think their assumption of 400% overbuilding seems very generous to me. I think 8 or 9 (even 10) times installed capacity would be needed. Current installed capacity is about 1.1 TW. An overbuilt PV/Wind system would be around 8.8 - 11 TW. Being very generous on the numbers it'd be about $1.30 US for each Watt on average. Total investment could range from $11.4 T US to $14.3 T US. That's 53% to 66% of last years GDP. These numbers are massive as will be price increases to customers. If the time-scale is anything less than a decade or two this plan would be madness to implement. That being said, I'd almost give money away to be a part of a study on this ideas feasibility.

 

[anorlunda]

When I wrote that article, the price for panels was about $0.80/watt. Panels only, installation not included. 2 weeks ago, I saw that the bulk price of panels has dropped to $0.12/watt, and the price for a whole solar farm installed, tested and connected to the grid is down to $0.75/watt, or $6 Trillion per 8 TW.

And the price declines don't stop here. Soon we will be at a point where the cost of mowing the weeds in a field once per year will be more than the cost of covering it in a solar farm. It is like Moore's Law on steroids.

So, as you probably know, a planning study projecting 10-20 years into the future would have to take those trends into account and use future prices lower than today's.

I also note that some people are concerned about land use. They fear that we need to cover vast land areas with panels. It is nowhere close to that. Here's a quick calculation.

Power density for one panel:	18 watts/square foot [194 w/m^2]
Square feet in one square mile:	27,878,400
MW per square mile	502 [194 MW/km^2]
Square miles in the lower 48 of the USA	3,119,884 [8080464 km^2]
Percent of the land needed for 8 TW.	0.5%

The ability to generate power locally rather than transmit it across continental distances is a major part of the appeal.

 

[jrmichler]

And there are many dual use opportunities to install solar panels to minimize the impact on land use.

The total area of interstate highways, state highways, and parking lots is about 40% of the needed land area in the @anorlunda calculations above. I like the idea of shade, especially when driving toward the sun. And the idea of shaded parking lots on hot sunny days.

Some farm crops apparently have better yields when partially shaded. Here's one link that has numerous other links in the article: https://www.treehugger.com/agrivoltaics-solar-power-crops-bees-4863595. They discuss crops that have better yields when partially shaded, reduced need for irrigation, and improved working conditions for farm workers. Arable land in the US is about 16.6% of the total land area: https://data.worldbank.org/indicator/AG.LND.ARBL.ZS?end=2016&locations=US&start=1961. Since only 0.5% of the total US land area would be needed for 100% solar, agrivoltaic has the potential of holding enough solar panels to fully meet US needs.

Rooftops could hold solar panels for over 1 TW: https://www.nrel.gov/docs/fy16osti/65298.pdf. Their estimate is based on 16% cell efficiency, so may be low. This would be more practical if solar panels were designed to be the roof, instead of being installed over the roof. Since roofing materials are not getting cheaper, and solar is getting cheaper every year, the possibility of a solar roof cheaper than a conventional roof exists.

 

[mfb]

And the price declines don't stop here. Soon we will be at a point where the cost of mowing the weeds in a field once per year will be more than the cost of covering it in a solar farm. It is like Moore's Law on steroids.

How much can installation costs drop? Installing something on something else isn't a particularly new technology, how much more can be saved?

 

[anorlunda]

How much can installation costs drop? Installing something on something else isn't a particularly new technology, how much more can be saved?

That's a good question. Certainly, the cost decrease exponent for the panels is not necessarily the same as the exponent for installation. But there is still a long way to go on the installation side.

I'm speaking mainly of solar farms installed on open land, rather than rooftops.

I watched a nearby project last summer of about 250K m²/ size. I saw that the start-to-finish time is short and the labor crews small. I estimate 5000-10000 m²/ installed per man-day of labor, so the labor costs were low. They used specialized machines.

The obvious future direction it to make installation machines like agricultural machines that plow/plant/harvest. The machine could roll over open land, leaving a finished installation behind. They drill holes, fill them with concrete, then push posts into the wet concrete, attach frames to the posts, and panels to the frames, and do whatever they need for the wiring. The frames hold everything rigid while the concrete hardens. Lasers and computers let them adjust for land contours. One pass and the job is done. That sounds to me like a level of difficulty comparable to some existing agricultural and road building machines.

In Lake County Florida (my winter home) they had a freeze in the 1980s that killed orange trees. Even today, there remains about 10⁹ m² of abandoned agricultural land in Lake County available for a new use.