David Mackay's "Sustainable Energy - Without the Hot Air" has some ballpark numbers:
https://www.withouthotair.com/c10/page_62.shtmlTo create 48 kWh per day of offshore wind per person in the UK would require 60 million tons of concrete and steel – one ton per person. Annual
world steel production is about 1200 million tons, which is 0.2 tons per
person in the world. During the second world war, American shipyards
built 2751 Liberty ships, each containing 7000 tons of steel – that’s a total
of 19 million tons of steel, or 0.1 tons per American. So the building of 60
million tons of wind turbines is not off the scale of achievability; but don’t
kid yourself into thinking that it’s easy. Making this many windmills is as
big a feat as building the Liberty ships.
For comparison, to make 48 kWh per day of nuclear power per person
in the UK would require 8 million tons of steel and 0.14 million tons of
concrete. We can also compare the 60 million tons of offshore wind hardware
that we’re trying to imagine with the existing fossil-fuel hardware
already sitting in and around the North Sea (figure 10.4). In 1997, 200
installations and 7000 km of pipelines in the UK waters of the North Sea
contained 8 million tons of steel and concrete. The newly built Langeled
gas pipeline from Norway to Britain, which will convey gas with a power
of 25 GW (10 kWh/d/p), used another 1 million tons of steel and 1 million
tons of concrete (figure 10.5).
Elsewhere in the book he shows the embodied energy and CO2 costs of various construction materials.
I've read similar comparisons of relative amounts of steel and concrete required for various types of renewable electricity generators vs. fossil and nuclear plants by Vaclav Smil (one or more of his books: "Still the Iron Age: Iron and Steel in the Modern World"; "Power Density: A Key to Understanding Energy Sources and Uses"; and "Energy Transitions: Global and National Perspectives, 2nd Edition" has the details) and others, and they all reach the same conclusions. IIRR Smil does cradle to grave analyses.