Energy viability

Every methane-cracking method invented so far is a net loss of energy. Meaning that it takes more energy to crack the methane than you ultimately get from burning the hydrogen. But in theory, this need not be true:

CH4 -> C + 2 H2 (endothermic: 74.850 kJ/mol)
2 H2 + O2 -> 2 H2O (exothermic: 285.820 kJ/mol)

Hydrogen production efficiency comparison

Type Method Energy out / energy in
Methane cracking Theoretical best case 382% see chemistry equation above
Not methane cracking Electrolysis of water 80%
Methane cracking Best technologies so far
Not methane cracking Heat -> turbine -> electrolysis of water 25%

There have been some green initiatives to use non-fossil-fuel energy to crack methane. However, this is only worthwhile if the efficiency is better than non-methane-cracking ways to make hydrogen (see table).

Above 100%, you can keep the system going without any non-fossil-fuel energy inputs. Just burn some of the hydrogen itself to crack the methane.

In the ideal case, fossil fuels could power the entire world with zero carbon emissions. At least, until oil and gas reserves run out.