[idiotsecant]

I am an electrical engineer that designs control systems for renewable power production for a living.

Curious what makes you think that the overhead and inefficiencies inherent to a million small solar installs is somehow better than a single managed facility benefiting from economy of scale both for maintenance and design.

Additionally, curious how you plan to address the problem of adding additional generation to existing way overloaded distribution systems to accomplish this. If this massive hypothetical solar install is all non-grid tied then fine, I guess, but you're losing a substantial amount of the power that's made that way.

Distribution systems don't come for free and have many of the same problems as 'last mile' internet. Not terribly complex but expensive en masse, particularly in areas that are not densely populated (which is a lot of the US).

There is a reason we spend a lot of money on transmission. Spending a lot of money on distribution helps a very small part of your network. Spending a lot of money on transmission helps a huge part of your network.

[quags]

Not OP but I suspect some comes from what I would view as the ability to first add solar to where there is existing infrastructure like parking lots, roofs, telephone poles etc. These areas are already built out and adding solar on top of it doesn’t take over an untouched eco system. Deserts are not voids of nothing ness and there is already a vast impact on the environment already.

[walrus01]

In my opinion the greatest benefits for large numbers of discrete small PV installs (with battery) is in places where it's uneconomical to extend the grid.

I'll use some hard to reach parts of WA and BC and OR and ID for example. You might be able to build a nice house/cabin on a piece of rural land and find that setting the poles and running lines to bring basic 100A or 200A service to that house will cost $40,000.

For 40k you can build quite a large off grid PV system that will have a reasonable ROI on it to serve the same loads, vs. spending 40k one time on construction costs for grid and then $50 to $200 monthly electric bills recurring for a long time after that.

As far as grid tied decentralized power systems do I agree with you 100%. It is VERY COSTLY in labor and complications to do something like cover the roof of a Home Depot or similar warehouse-sized structure in grid feeding PV, as compared to doing medium-sized to massive scale ground mount PV on empty land somewhere.

[oceanplexian]

Unfortunately, banks don’t finance anything “off grid” with conventional financing. There are some banks that do, but you’re going to pay a premium compared to a standard Fannie/Freddie backed mortgage.

That’s the biggest issue with this plan, which can be fine for a certain kind of person but not 90%+ of the market of potential people who inevitably have to finance a home.

[walrus01]

It can sometimes be "good enough" to check the box for bank financing that electrical service is available at the county maintained road which is the mailbox/street address location for the property... Something with a long or difficult driveway could easily be 30-40k to extend grid service from road to house.

[seanmcdirmid]

My uncle in Newport WA, close to the Id border and pretty isolated (although the house has electrical service now) used a solar shower for a long time. But there wasn’t actually enough sun to be off grid all that much, and unless you have running water to build a water wheel, some grid access is needed. You could probably do more with solar in scrub areas, but you might need some kind of diesel generator back up for the dark winter.

[walrus01]

Yeah one of the big challenges with off grid at that latitude is that your december and january kWh cumulative production (per month) will really suck. It's not as bad as it used to be with panels available at $0.40 per STC watt rating.

If you go to the NREL pvwatts calculator and plug in a theoretical 20kW STC rated system at his latitude/longitude you'll find that the Dec. and Jan. production will not be very much at all.

Sometimes it can be more affordable to significantly over-size a system to product enough kwH per month in mid winter vs. spending $4000+ on a generator.

If you can get 360W 72-cell panels at $150 a piece in whole-pallet quantities, 4000 bucks buys a lot of panels... Not counting the mounting and cabling and junction box cost.

As a budgetary figure, can easily spend $4000 on a small Generac propane generator and wiring/transfer switch.

At other times it can totally make sense to have a small generator to run 2-3 hours a day in mid winter, feeding a charger that can add a constant 1500W into a battery system to keep up the battery string voltage.

[PaulDavisThe1st]

> For 40k you can build quite a large off grid PV system that will have a reasonable ROI on it to serve the same loads, vs. spending 40k one time on construction costs for grid and then $50 to $200 monthly electric bills recurring for a long time after that.

Why would such an off-grid system have any monthly electric bills? Are you just pre-amortizing the cost of replacement batteries?

EDIT: I'm an idiot. The bills are for the grid-tied option.

[bombcar]

They’re saying that it’s 40k once for PV or 40k once for a grid connection, but then you need to pay the grid operators monthly.

[PaulDavisThe1st]

Ah, thanks. I didn't read carefully enough.

[debo_]

You are not (necessarily) an idiot. Misreading something doesn't make you stupid!

[PaulDavisThe1st]

I said I was an idiot, not a stupid person :)

[debo_]

I knew that was coming.

[epistasis]

Interconnection queues are overloaded. But most of all:

> There is a reason we spend a lot of money on transmission

This should be in the past tense. Transmission was the only option, because generation needed to be centralized to be anywhere near cost effective. And transmission is suuuuuper expensive, by far the biggest part of my solar bill.

Further, it's used ineffectively, often with capacity factors far below 30% because it's sized for peak, not for average use.

As the cost of generation gets lower and lower, the cost of transmission and distribution becomes an ever higher fraction of power costs, and the next thing to minimize.

Big transmission across the continent makes the most sense. But it's also the hardest to build. We are likely stuck repowering existing transmission to higher capactities.

Sure, we will need a lot more transmission capacity, but the less we can get away with will probably lower costs.

(And transmission is only geographic arbitrage, we will soon have time arbitrage with batteries that will rival nearly all transmission costs. Transition is not getting cheaper, but storage solar and wind are.)

[nonethewiser]

> Curious what makes you think that the overhead and inefficiencies inherent to a million small solar installs is somehow better than a single managed facility benefiting from economy of scale both for maintenance and design.

Location. Deserts are far from people. 80% of people in the US live on the east half of the country (and most of these on the eastern half of that). But even the midpoint is too far from deserts to use energy from it.

[maxerickson]

My town built a solar install on undeveloped land adjacent to the airport. That's lower impact than desert and cheaper than smaller installs would have been. I'm sure there's lots of situations like that, even in areas denser than the rural Midwest.

[s1artibartfast]

Transmission losses are tiny. Think single digit percent.

Meanwhile urbal solar is many times more expensive, and vastly less efficient per panel.

The total us grid transmission losses are 5%

https://www.nrdc.org/bio/jennifer-chen/lost-transmission-wor....

[unusualmonkey]

> Deserts are far from people

So... you've not heard of LA or Arizona or Nevada etc?

[titzer]

I lived in LA. It is a desert. It'd be just fine to cover parking lots, highways, and rooftops with solar. There are at least 100 square miles of parking lots in LA.

[seanmcdirmid]

LA actually isn’t a desert, although there is desert in the LA area. Most of LA is classified as Mediterranean, and gets enough rain (the closer to the coast, the less deserty it is). Lots of mountains make for interesting climate zones.

[titzer]

Fine, to be pedantic, it's right on the limit of being desert. LA averages about 15 inches of rainfall per year, whereas the classification for desert is less than 10 inches. With the amount of irrigation in the area, this is not generally acknowledged. Funny, I used to have to hop over a giant puddle in the street every morning to catch the bus to UCLA because the sprinkler systems for someone's 4ft wide lawn were just that badly misconfigured. I always thought that inane.

[seanmcdirmid]

I spent all my time in LA living in Westwood, and saw some interesting rain storms while I was there. Interesting tidbit: Santa Monica (just next to the coast) actually gets less rain a year than LA (13.5 inches vs. 14.7 inches). I'm guessing they get more moisture in fog than in rain.

[ben_w]

You could, with current technology, put the PV on your antipode with sufficiently small losses it's basically fine.

Actually building one is harder than designing one, so I'm not suggesting breaking ground on a global grid immediately, but if you in the USA can get past the three way split of {east, west, Texas}, that'll do a lot all by itself.

[caeril]

Transmission losses are a concern, but this is fairly irrelevant to the OP's original idea: that of producing carbon fuels in the desert and then transporting them. Most populated places in the US are pretty far from Erath, Louisiana or Cushing, Oklahoma, and we get by just fine under this arrangement.

[tobias2014]

High voltage direct current is the keyword. It is possible. Unfortunately the plans to build a giant solar farm in the Sahara and connect it to Europe never took off (for now). Losses are about 10% per 3000km. See also what the DESERTEC foundation does.

[chiefalchemist]

> Distribution systems don't come for free and have many of the same problems as 'last mile' internet.

Nor does centralized. The grid needs to be upgraded to handle the significant increase in demand. But more importantly, over-centralization will mean we're putting all our energy production in fewer baskets (than we have now). Decentralized is a form of a redundancy, a form of backup. It's also a form of independence.

There's no single silver bullet. We'd be wise to blend, and blend wisely. And yes, they might have some added financial costs, but not doing it will surely come at other costs (e.g., blackouts).

[noiceyeha]

Engineering mitigate signal attentuation and paying all the middle men from Desert->a home thousands of miles away ain’t cheap

[mycall]

What is your opinion of distributed solar having less or more chance of causing forest fires than a few managed facilities?

[CyberDildonics]

Where are you getting the idea that solar panels cause forest fires?

[mycall]

No, the opposite. High power long distance power lines do. Solar panels are much less likely.

[CyberDildonics]

What does solar have to do with it at all? Why mention 'forest fires' in the first place?