Almost Plausible Solar Takeover Plan

[Lord Crc]

No no. It's not true that they don't take EVs into account. In fact the switch to EV means that the electric sector demand will increase very substantially. But when discussing how to meet the total demand, it is unnecessary to mention each of the uses that contribute to the demand.

But in this case they take today's demand, which is unrealistic, and no mention of using EV to shift demand.

If they had angled it as "this is how it could have looked today", the fair enough. But then don't talk about "we can build this by 2030", because it is irrelevant.

There may well be studies that do take this into consideration, so not trying to pan all of them.

 

[Baluncore]

But in this case they take today's demand, which is unrealistic, and no mention of using EV to shift demand.

The article has a paragraph on the subject of EV.
The electrification of transportation will increase demand, so it will obviously increase the capacity that must be installed. But EV comes with some storage so the same degree of overbuilding is not required.

 

[Baluncore]

The unusual thing about the article linked in the OP is that it considers changing the game when considering renewables. It does that by inverting the fraction of the installed capacity. Instead of 1/4 of peak demand, the renewables go to 4 times the peak demand. Advances in storage technology and storage capacity now allow that analysis step to be considered.

The article made it clear that the analysis was for the USA, not Norway, nor Australia. That does not mean that the same analysis cannot be applied in Norway with hydro and wind, or differently in Australia with wind and solar. It is a mistake to think that if there is one difference or exception, the overbuilding analysis fails locally or globally. Everywhere is an exception and needs to be analysed in it's own way. The article does not consider the USA as a single regime. It considers different states as needing different solutions.

The article also made it clear that the electrification of transportation was not part of the analysis. That future issue may be worrying for some people, but it does not change the inverted analysis and the need for overbuilding. The economy of scale will come into play sooner with the transition to EV.

It is clear that as the renewables share of energy production increases, it threatens the profits of the fossil fuel energy generators. But it is not the increasing capacity of renewables that is the problem, it is the reduced need for the fossil fuels. The problem is going to be crossing that economic bridge. When that happens, the market will be turned on it's head and the fossil fuel generators will become fossils themselves. The only hope is to economically tie the installed renewables to the fossils. At some point subsidies to renewables will need to reverse, to subsidise the fossils during their transition to extinction.

 

[hmmm27]

But it is not the increasing capacity of renewables that is the problem, it is the reduced need for the fossil fuels.

Depends : is it cheaper to strip the H's out of water, or out of petroleum ?

 

[russ_watters]

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

So...is there a paper associated with this model, or is that article and the graphical presentation of results on the website all there is? Unfortunately, without the details of the basis for the model, there's really nothing to evaluate (unless we guess what they did, which I'd rather not).

The premise is that if you start from scratch with only the hydro we currently have, what would the lowest-cost all-"renewable" grid look like. That's a little bit "out there" for my taste and I think the "almost" in your title is a generous characterization.

 

[Baluncore]

Depends : is it cheaper to strip the H's out of water, or out of petroleum ?

Cheaper economically or environmentally?

 

[hmmm27]

Cheaper economically or environmentally?

yes.

 

[russ_watters]

yes.

Economically: fossil fuels (mostly methane)
Environmentally: water

Stripping hydrogen from fossil fuels does not substantially change their climate impact because the combustion products remain the same whether you do it in one step or two.

 

[hmmm27]

Stripping hydrogen from fossil fuels does not substantially change their climate impact because the combustion products remain the same whether you do it in one step or two.

Specifically, I was thinking of having C - not CO2 - as a (by)product.

 

[russ_watters]

Specifically, I was thinking of having C - not CO2 - as a (by)product.

As far as I know, there is no process for that, and if there was it would necessarily be more expensive. The goals are opposites of each other because of the innate chemistry of the processes.

 

[Baluncore]

Specifically, I was thinking of having C - not CO2 - as a (by)product.

Water is cheaper and excess energy is free.
You are missing the top-down point of the article by getting bogged in the myriad possible bottom-up details. Just how energy will be stored is variable in place, time and economy. The cheapest path now will almost always cost more in the longer term.

 

[hmmm27]

You are missing the top-down point of the article by getting bogged in the myriad possible bottom-up details.

The segue was reacting to "the fossil fuel industry is toast"(paraphrased) ; I'd happily argue that for the next century or so, the industry will still be going strong : the "trick" would be making it clean and ghg-free.

 

[Svein]

In his novel "Friday", Heinlein introduces the "Shipstone", a very efficient rechargeable (and high-power) battery. It is recharged by solar power in the deserts. Are we looking for a realizable version of a "Shipstone"?

 

[anorlunda]

So...is there a paper associated with this model, or is that article and the graphical presentation of results on the website all there is? Unfortunately, without the details of the basis for the model, there's really nothing to evaluate

I was attracted to the article for the same reason you were in #7 --- a novel way to approach the capacity factor problem. But in #26, I give a reason why it would be rejected as a national plan.

Like many engineers, I am interested in limiting cases. They help us to clarify our thinking. I was hoping to make a thread where I forced the limiting case by assuming 100% of one source. Then we could explore that. I did not intend for it to expand into CC, or environment, or the "best" energy future, or national energy policy, or EV, or solar versus other power sources. Alas, that didn't work. Some public discussion topics are too "hot button" to be contained.

 

[Lord Crc]

As I mentioned, 100% solar seems to require a bit different usage patterns than normal. You'll want to shift load to be peaky, like the solar panels, and avoid night time load. Instead of people charging their EV at home at night, it seems they should charge during work.

On the other hand, with V2G they can then use the car to help out with their cooking etc once they get home.

Alternatively one would have to install quite the battery backup, if people are to use the grid more or less in the same pattern as now. I don't know what is cheapest, but I assume it is easier to go for huge batteries, rather than substantially upgrade infrastructure in cities.

Nuclear plants will have to be decommissioned at non-trivial cost, though I'm not sure how many are expected to provide service until 2050 or beyond (2030 + 20 years).

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

 

[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?