Moving heat from a low temperature to a high temperature requires energy by all laws of thermodynamics so this definitely isn't "free." Even the wikipedia article states:
"The magnetocaloric refrigeration system is composed of pumps, electric motors, secondary fluids, heat exchangers of different types, magnets and magnetic materials. These processes are greatly affected by irreversibilities and should be adequately considered."
Irreversibility is "the cause" of poor efficiency so we can infer it's not even that efficient right now. And right now the technology only works to cool a system already at near-cryogenic temperatures to one much colder.
If you want to look for future refrigeration technologies our best efforts are probably placed in thermoelectric modules which use the Peltier effect. They can also heat and generate power all using the exact same device. Right now they are 1/4 to 1/5 as efficient as standard refrigeration but they are solid-state and hampered primarily by ohmic resistance in the thermocouple pellets. They, however, can really be improved with some new material improvements and DARPA has just the projects out there, investing in "high Z materials." Most of their uses are in environments were extreme reliability is needed - I.E. deep space power generation for probes from heat expelled from a radioactive source - and cooling photonic equipment due to the precise temperatures needed.
|