Metallurgical coal is going to be the very hardest thing to get away from. Improvements in the coking process (carbon capture) are plausible, but the chemistry basically demands burning a butt-load of carbon.
Thermal coal is replaceable by conservation, nuclear power, and to some extent wind/solar plus storage, and if I were in charge I'd be pushing to ban new thermal coal development. But metallurgical coal is going to be almost physically (or chemically) impossible to replace.
It's a a big use, 14% of global coal use from memory based on a recent query.
BP's annual energy review 2014 gives total global coal use as 3,881.4 Mtoe (million tons of oil equivalent), which works out to 5.54 billion tonnes of coal. 14% of that is still a large number, 776 million tons.
I happened to look up total forestry production in the US recently, let's see if I can't find that ....
Total US lumber production is equivalent to 928 Mboe. So, yeah, applying 84% of US timber production to coking activities would supply the world with steel.
US steel production is just under 100 megatonnes (2007). Global production is 1,600 megatonnes. So US production is 6.25% of global. The US could provide for its own steel manufacture somewhat more feasibly.
Incidentally, energy use in producing construction materials is among the reasons I see a possible future not for biofuels but for biomaterials -- utilizing plants or organic processes to directly provide structural materials (or replacements for them) that we use.
Carbon separation from seawater seems another possible approach, though I haven't specced out that for coking purposes:
Thermal coal is replaceable by conservation, nuclear power, and to some extent wind/solar plus storage, and if I were in charge I'd be pushing to ban new thermal coal development. But metallurgical coal is going to be almost physically (or chemically) impossible to replace.