Did I do the math right? I'm using full precision but write only 2 sig figs.
a D + D → T (1.01 MeV) + p (3.02 MeV) b D + D → 3-He (0.82 MeV) + n (2.45 MeV) c D + 3-He → 4-He (3.6 MeV) + p (14.7 MeV)
We do reaction a and b, remove T from the plasma and let it decay to 3-He. So we have 2 3-He and do 2 c, which uses up another 2 D and end up with:
d 6 D → 2 He + 3 p + n + 43.9 MeV
I live in Austria and food grows fine here, so I picked the third shade of blue which is 130 W/m2 insolation. Since plants don't need IR I pick 100 W/m2 as LED based "insolation".
Assuming 33% efficient LEDs we need 300 W/m2 electric (there, plenty of IR :).
Two 32 km long and 8 km diameter O'Neill cylinders have 1.6E9 m2 mantle area, so we need 4.8E11 W or 1.5E23 J in 10,000 years.
Now we need to make fusion reaction d
1.5E23/43.9 MeV=2.1E31 times
to get that energy. Assuming 75% electrical efficiency (we convert the charged fusion products directly to electricity, and only the waste heat goes to the steam turbine), so let's make that 2.1E31/0.75=2.9E31 times.
This gives us
2.9E31/6.022E23=4.8E7 moles of this reaction which needs 3 D2O per reaction and thus we need 4.8E7*3*(2*2+1*16)/1000=2.9E6 kg D2O.
That is 2.9E6/1107 kg/m3=2596 m3 D2O.
If we cover the mantle of the O'Neill cylinders with D2O we'd need a 2596 m3/1.6E9 m2=1.6E-6 m=1.6 um thick film.
Bugger, last time I did the calculation (when I didn't factor in LED and fusion power plant efficiency) I got 0.4 mm.
Is 1.6 um right?