# Almost Plausible Solar Takeover Plan

• Featured
Mentor

## Main Question or Discussion Point

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.

NTL2009, hmmm27, bhobba and 5 others

## Answers and Replies

Related Electrical Engineering News on Phys.org
Svein
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.

256bits
Gold Member
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.

Mentor
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.
That's true. I was thinking about the lower 48 states in the USA. Obviously, it won't work near or above the arctic circle. I mentioned that in #1.

The required overcapacity would be less in southern states than northern states.

The upfront and operating costs for renewables continues to fall and despite the 20+ years of naysayers claiming it will destroy the economy and is a fools errand.

However - whenever I see an "argument" regarding the economic feasibility of renewables that does not account for environment cost of the status-quo, then I simply have to assume the motivation of discussion is purposely ignoring the driving issue.

You can agree or disagree with my opinion (based on educated experts) that we are in an environmental crisis that is due to human activity. To address this we need to make many changes in order to reduce the amount of carbon converted to CO2, by a staggering amount, and to do this we need to shift our source of energy, but also use (consume) less.

No scientist, study, or solution is perfect - and constantly knocking them for being imperfect (sniping), meanwhile the alternative being "let's do nothing until we have a perfect solution" is what? Irresponsable.

I feel like we are driving on the highway towards an accident, and refusing to slow down because : "There is no accident", "I don't feel the accident.", "That accident will slow me down - I can't stop now" "There is no way around that accident, so why slowdown."

If you went to see 100 doctors and, 97 of them say you have cancer - guess what, you get the {expletive] chemo / surgery / radiation...or what ever it takes to fight it.

Sorry for the rant - but we NEED solar today because it is an effective part of trying to fix a serious problem that will not just go away "like a miracle" tomorrow. Yes there are many challenges and nothing can do is perfect, all we can do is work on the problems and keep fighting. If you do not believe we have a crysis and that we need to change, then please, just say that, so we can see where you are coming from.

vela and DrClaude
Mentor
Sorry for the rant -
What in this thread are you ranting about? It is pro solar.

Or do you mean today literally, only 13 hours left?

russ_watters
Mentor
I haven't read it yet, but I will say from an aesthetic point of view I like the take. Too many solar advocates ignore/gloss-over the capacity factor issue, so it's refreshing to see one plow throw it head-on. The conclusion is a surprising one, but at least I know they've gone to the effort to attack that difficult problem and find a potential solution.

My concern that I'll need addressed when I get a chance to read is is how they'll deal with the marginal cost vs output of the installations. It's one thing to say that the average cost gets paid back in production in X years, but if the marginal payback of the last few solar plants is 50x, it won't make sense to do it. As an academic exercise it might be an interesting scenario, but I expect it would still be cheaper to run the baseload on something else.

NTL2009, bhobba and trurle
mfb
Mentor
$8T over 20 years (by that time you probably start replacing cells) is$400B per year. Divided by 4 trillion kWh we get $10 cent/kWh. That's quite expensive for the raw production, it doesn't include additional storage cost that comes with this option, it doesn't include maintenance, and it needs$1/W solar, a factor 3 improvement over today, at a time where installation costs (=people working on site) are already significant. 4-10 days of storage is going to cost a lot, even with hydrogen (the low conversion efficiency is probably acceptable with the big overproduction).
Where do you get 1 TW from? Average demand is about half of that.

Mentor
As an academic exercise it might be an interesting scenario, but I expect it would still be cheaper to run the baseload on something else.
We use marginal costs to allocate between competing sources of power. If there is only one kind of source, there is no competition. More important, I think the reliability engineers would argue the case for solar on simplicity and reliability rather than strictly cost. Here are some of the non-cost factors:
• There are few things in this world more complex than a steam power plant or as simple as a solar panel. Simplicity and reliability go hand in hand.
• The solar panels would be more thoroughly distributed and geographically closer to the loads than central power pants are. That reduces the MW*miles of power transmission. That has a powerful benefit to reliability.
• Solar panels require little or no smarts or Internet connections, thus offering immunity to cyber attacks. Considering cybersecurity, it would be wise to resist the temptation to add any kind of digital monitoring or control.
• Panels could self-dispatch using only the local information of grid frequency. That's the way we controlled distributed power grids since the 1880s, long before remote communication and central controls. See this Insight. No remote communications needed. No economics or costs considered, only frequency.
• A solar farm should exhibit incremental degradation rather than total failure when components start failing.
• On the minus side, the magnitude of panel snow/ice/dust removal efforts is hard to imagine. If done by robotics or automation, we would sacrifice some of the simplicity.

bhobba
Mentor
Where do you get 1 TW from? Average demand is about half of that.
Generating capacity is determined by peak demand, not average. EIA.gov says that the installed generating capacity of the USA is 1050 GW. (I don't know if that include Alaska and Hawaii), but they're small anyhow.

and it needs $1/W solar, a factor 3 improvement over today, The numbers change rapidly. See this news from last week.$1B for 690 MW including storage. That's $1.44/watt including storage. https://www.smithsonianmag.com/smart-news/solar-energy-project-nevada-will-be-biggest-us-180974862/ My concern that I'll need addressed when I get a chance to read is is how they'll deal with the marginal cost vs output of the installations. It's one thing to say that the average cost gets paid back in production in X years, but if the marginal payback of the last few solar plants is 50x, it won't make sense to do it. As an academic exercise it might be an interesting scenario, but I expect it would still be cheaper to run the baseload on something else. Current state of solar generation is "cheap, but not cheap enough". Business (at least in Japan) is ready to operate solar plants for 0.165$/kWh, while baseline costs for fossils is 0.1$/kWh. Any solar over-capacity requirement will be devastating for solar power generation with the current economics. In brief, we simply still have too much cheap oil. https://www.power-technology.com/comment/japan-solar-pv-subsidies/ Sorry for the rant - but we NEED solar today because it is an effective part of trying to fix a serious problem that will not just go away "like a miracle" tomorrow. Yes there are many challenges and nothing can do is perfect, all we can do is work on the problems and keep fighting. If you do not believe we have a crysis and that we need to change, then please, just say that, so we can see where you are coming from. Then the "accident" will be avoided if perceived future cost of the carbon volatilization will translate into something like "carbon tax". Current risk perception is$6-$60 per ton of carbon, or 0.007-0.07$/kWh for oil burning or 0.018-0.18$/kWh for coal. The upper end of perceived cost range provide just enough incentive to switch from oil to solar. The coal is already in the process of being phased out as mainstream power source though. Last edited: russ_watters mfb Mentor Generating capacity is determined by peak demand, not average. EIA.gov says that the installed generating capacity of the USA is 1050 GW. (I don't know if that include Alaska and Hawaii), but they're small anyhow. I assumed the x% safety factor was above the average load, as the plan is to average across a day or multiple days anyway. The numbers change rapidly. See this news from last week.$1B for 690 MW including storage. That's \$1.44/watt including storage.
Wait, are the numbers installed power now? If you take installed power then 400% "overcapacity" is barely covering the average, and will let you freeze in winter. Well, if installed capacity is 400% of the peak demand then it might work somewhat if demand is lower in winter (->US, not Europe).
And the 1440 MWh storage won't last long, that's not even enough to cover the night.

Svein
Some random thoughts:
• Solar panels are high technology. It needs a highly sophisticated manufacturing process. From where do you get the power to manufacture the panels?
• Solar power - almost all the power we use is solar power when you look behind the curtain. Hydroelectric power is solar power. Fossil fuels are stored solar power. Trees are a result of solar power. Atomic power is - not.
• To get very physical: One of the laws of thermodynamics (https://en.wikipedia.org/wiki/Laws_of_thermodynamics) states that you need a hot source and a cooler "drain" in order to get energy out of a thermodynamic system.
• "Heat death": There are about 7.5E9 human beings on this planet. A human being produces about 600W when sitting still. That is 4.5TW of "heat pollution".

bhobba and mfb
The argument made here earlier about the low solar capacity due to climate in far north , I think by @Svein doesn't seem valid to me , one of the obvious reasons being that in the North there is very little sun but there are also very few people living there and very low electricity demand.
Scandinavia might be an exception but Russian far north is so scarcely populated that if you ran out of fuel there on a road you might die before finding the nearest home.
I don't know about Alaska but I think Alaska is also mostly empty land.

My summary being that, humans tend to live where the sun also shines , alot.

Mentor
I assumed the x% safety factor was above the average load, as the plan is to average across a day or multiple days anyway.
We are more sophisticated than that. We have lots of software to predict the absolute highest peak load. The public demands that there is enough power to meet demand always, not almost always. On top of that, we typically require 20% reserves. That's the safety factor. But for installed capacity, it is 20% of peak, not average. Average doesn't enter the calculations anywhere. Also, for installed capacity, there's an additional factor for the percent that might be out of service for maintenance, repair, or upgrade.

Well, if installed capacity is 400% of the peak demand then it might work somewhat if demand is lower in winter (->US, not Europe).
Yes, the idea is 400% of peak demand. It can also be expressed more simply if we use a different definition of the rating of a solar panel. Instead of rating it by how much power it makes in the best case instantaneously, rate it by the power it makes in the worst case 24 hours. A conservative rating rather than an optimistic rating. Then using that conservative rating, you need to satisfy 100% of the peak demand.

Using the conservative definition, today's panel advertised to have a rating of 1000w, would need to be redefined as rated at perhaps 100w. Of course that has practical problems because of latitude differences, but it makes the idea simpler.

From where do you get the power to manufacture the panels?
From the same sources that you manufacture anything. Why is that important? We are not discussing free energy, or climate change, or the end of energy angst, but rather a way to spend money and resources to satisfy our electric demand.
---
I should emphasize that the operating costs of 100% solar are not zero, despite the fact that they do not consume fuel. Snow/ice/dust removal is one thing. Rodent damage, random failures, and vandalism are all significant. Technological obsolescence will motivate upgrades once in a while. Hurricane/tornado/earthquake/landslide/fire/volcano will destroy some percentage of the panels every year, and force replacements. We even need to think about the wartime vulnerability. Estimating all that stuff is one of the major tasks of the study.

Just to hazard a wild guess, 10% replacement per year is credible. If capital costs are spread over a number of years, then operating costs will overtake capital costs at some point. Accurate forecasts of operating costs are non-trivial. In #1, I guessed 50 man-years to do the engineering study.

Svein
My summary being that, humans tend to live where the sun also shines , alot.
Well, I prefer some chilly weather to heat-related fatalities (every year media tells us that hundreds of people die in the south of Europe due to "exceptional" high temperatures).

And if people prefer living "where the sun shines a lot", why the large demand for air conditioners and the power need to run them?

Mentor
My summary being that, humans tend to live where the sun also shines , alot.
I should have put it in the title, that my thoughts and numbers cites are specific to the lower 48 US states. I prefer not to expand the discussion to global scope.

eq1
We have a natural experiment in California that’s testing this thesis. In 2017 it was big news that we had to pay neighboring states to take our power because our solar farms generated so much of it. That incident convinced me that the political problem is way harder than the engineering problem.

https://www.latimes.com/projects/la-fi-electricity-solar/

A human being produces about 600W when sitting still.
Surely that isn't correct. That would mean sitting still for 12 hours would be 7.2kwh = 6.2 kcal. I'd be skin and bones if that was true.

russ_watters
The analysis modeled meeting current uses of electricity, based on projected technology costs for 2030.
But by 2030 electric cars, busses and trucks should be a significant factor. So any semi-realistic plan needs to factor in that.

anorlunda
Substantial battery capacity would be needed in regions with the highest percentage of solar generation. In the Southeast, batteries would store up to 36% of annual generation for delivery after the sun goes down.
I'm confused. They say annual generation, but isn't the main issue with batteries their capacity? If they mean production, then 36% of annual production seems substantial in light of the Hornsdale Power Reserve.

Mentor
I'm confused. They say annual generation, but isn't the main issue with batteries their capacity? If they mean production, then 36% of annual production seems substantial in light of the Hornsdale Power Reserve.
You have a good eye. I reacted to the same sentence when I read the article. My conclusion is that it is a plain error. They should have said "daily".

bhobba
Mentor
We have all sorts of claims and counter claims floating around in Australia. Some people say Australia has an inconsistent policy, personally I would describe it as a 'robust' debate. But anyway one state, Victoria, is going gung ho on renewables, and what results will, IMHO likely, in the medium term, determine its viability:
https://www.theage.com.au/politics/...-victoria-pulls-the-plug-20200217-p541nj.html

Long term I am unsure what will happen as I think eventually Fusion power may happen - but that has been promised for a long time.

Thanks
Bill

So I wonder if they also worked out the land area we'd have to destroy for this environmentally friendly endeavor?

Topaz (https://en.wikipedia.org/wiki/Topaz_Solar_Farm) is 550MW peak. It covers an area of about 19km^2.

If we use that as a reference, 19km^2 for 550MWpk, then to produce 8TW peak results in about 276000km^2.

The entire US has an area of about 10million km^2 (including water).

Are you really suggesting that covering about 2.7% of the entire united states with solar panels is somehow a good idea? For the "environment"?!

Then, a slight issue: TPES.

In the US, quick google didn't give me recent numbers, but even old numbers tell a tale.

TPES, per capita in USA, = 9200W (https://en.wikipedia.org/wiki/List_of_countries_by_energy_consumption_per_capita)
Electricity consumption per capita in the US is 1300W (https://en.wikipedia.org/wiki/List_of_countries_by_electricity_consumption)

Which is to say, covering 2.7% of the entire USA in solar panels only solves 13% of the problem, so to really get "environmentally friendly" we'd only have to cover about 20% of the US in solar panels.

Sounds totally reasonable to me.

So then, in 2013 (again sorry about old numbers, but it gives you a reference), USA consumed 91000PJ TPES (PJ = peta joule), or, given there are about 3.154e7 seconds in a year, 3TW continuously.

That 8 terra dollars could off course completely make the USA carbon free for all of its energy without having to cover 20% of the country in solar panels, but instead, lets all continue on this wind and solar feel good loony train.

Keep in mind, only about 3% of the total land area of the US is "developed"....