Feeding 100% of Europe electricity use with solar panels in North Africa would required many years (!) of the world’s current aluminium production, just to build the transmission cables.
Do the calculation, you’ll see.
Long distance power lines do not work to transmit massive amount if electricity on a global scale.
1) HVDC designs I’ve seen are copper (towers use aluminium because it’s light, IIRC)
2) http://www.necplink.com/docs/Champlain_VT_electronic/04%20L....
Gives 2500mm^2 cross section for a 1 GW cable
3) WolframAlpha says Europe’s electricity production is 410 GW: https://www.wolframalpha.com/input/?i=total+Europe+electrici...
Which means the total conductor cross section needed is ~1 million mm^2 = 1m^2. Ok, this sounds like it’s going to be a lot.
4) Lets put a line across the Sahara to connect all the panels plus connections to the existing EU grid in Gibraltar, Athens, and Milan.
It’s about 3700km from Casablanca to the middle of Egypt: https://www.wolframalpha.com/input/?i=casablanca+to+egypt
Likewise 350km for Gibraltar, 1000 km Awjilah to Athens: https://www.wolframalpha.com/input/?i=Awjilah+to+athens
This gives a total length of about 5000km, if I spec the cable for 100% of EU power going through each cable, which is excessive as I was trying to suggest this as an adjunct to batteries and local PV rather than a total replacement for either: any combination (including none) of transmission and storage only has to cover lower nighttime/seasonal averages).
This gives me a total volume of 5000km * 1m^2 = 5e6 m^3.
5) This is copper, worldwide production of copper is 14.6e6 tons/year, given the density this is indeed 1.4e6 m^3/year and therefore multiple years at current mining.
6) Global aluminium production is 82.6 million tons/year: https://www.wolframalpha.com/input/?i=worldwide+aluminium+pr...
Aluminium is 60% the conductivity of copper; I assume that means I need the conductor to be 1/0.6 times the cross section? Not my field. Assuming that, I want 8.3e6 m^3 aluminium, given the density that’s 22 million tons, so 3 months.
Edit: I forgot Milan!
7) Tataouine to Milan is about 1500 km: https://www.wolframalpha.com/input/?i=Tataouine+to+Milan+
Therefore multiply my mass estimates by 1.3
Edit 2:
410 GW is also 2-2.5 times current global PV installation: https://en.wikipedia.org/wiki/Growth_of_photovoltaics
If you did put enough PV for all of Europe on top of/along side the Casablanca-Egypt line, the PV would need to be about 550m wide: http://www.wolframalpha.com/input/?i=%28410GW%2F%281kW%2Fm%5...
(I only need to care about peak power in this case, not average, hence only the 20% efficiency factor and not including the additional 25% duty factor).
Couple examples: You put a line to cities in the south of Europe. But existing lines from there are nowhere large enough to take all that power, you will have to build lines from Milan and others all the way inside Europe. You can’t build a line to Gibraltar and call it a day, it has to go all the way up to Norway (albeit with tapering)
You also overestimate the efficiency of aluminium power lines, don’t take into account the amount of towers and/or plastic material that would be needed to wrap the lines if you make them underground.
We are talking about years of production, and that’s just for Europe, if you want to replicate this to other continents you quickly get a supply challenge, even if you spread this over 30years, at which point it will be too late, regarding global warming. Not to mention that m the CO2 emissions associated with building such massive lines would take decades (if ever) to be offset with the gains from the use of Solar, which was the point of the thing in the first place...
But in the end my point was not that this is impossible, but that efficiency is not the largest limiting factor here. Material supply is the largest problem. It is not insurmountable but it is a real challenge.
> But existing lines from there are nowhere large enough to take all that power, you will have to build lines from Milan and others all the way inside Europe.
Really? Okay. All I can do is look at maps like this one:
https://www.researchgate.net/figure/7-Modelled-AC-transmissi...
Which look, to my completely non-expert eye, like a decent size grid already exists.
I know that picture doesn’t contain enough info even though this isn’t my domain, but it represents the limited level I’m coming from.