This is not a finished project—just a rough draft i threw together today to see if the idea holds water. i’m sharing it early to get feedback on feasibility before sinking more time/money into it. here’s the gist:
core idea
use immersion-cooled bitcoin miners (antminers s21/s19/s17/s9) to generate waste heat, then repurpose that heat for home heating , cooling , and maybe laundry.
power via solar + battery storage to keep it self-sustaining.
thermal recovery (the sketchy part)
waste heat from the immersion tank would route through:
Radiators (heating when temps drop below 19°c).
Absorption chiller (silent cooling when temps rise above 25°c, though these things are expensive , finicky , and less efficient than heat pumps still, the idea of using miner waste heat to cool your home feels cool on paper).
clothes dryer (redirect hot water during laundry cycles, but only if immersion tank temp stays under 75°c. if it spikes, the system dumps heat outdoors via an exhaust fan).
note: this part is highly theoretical. i haven’t tested the absorption chiller integration or thermal loop dynamics yet.
control systems (the “i hope this works” part)
temperature state machine
a plc would comtrol load and toggle between modes:
Heating : send hot water to radiant floor / home radiators.
Cooling : activate absorption chiller.
Dry_clothes : redirect heat to dryer.
Exhaust : trigger outdoor fan at 75°c.
load balancing
priority order:
Home loads (lighting, appliances).
Exhaust radiator (if immersion tank >75°c).
Battery charging (only if above minimum reserve)
4–7 miners enabled sequentially as solar allows.
Sell excess to grid (lowest priority).
btc mining is higher up than grid sales, but still behind critical loads. makes sense when mining rewards > dumping power for pennies.
current status
this is all just draw.io diagrams and logic flow right now. no hardware built yet.
big unknowns:
how absorption chillers handle variable thermal input from miners.
whether the staggered load control can prevent immersion tank overheating without constant manual intervention.
safety risks (e.g., dielectric fluid leaks, electrical load limits.
Why share this early?
I'm curious:
Is this even feasible, or am i missing obvious flaws?
Are absorption chillers worth the hassle, or should i pivot to a simpler heat pump-based system?
How would you prioritize thermal recovery vs. mining uptime?
This is what it should look like
https://drive.google.com/file/d/1qhC2RxMpXX_FEjwHvynYdoDqLZo...
waste heat from the immersion tank would route through:
Radiators (heating when temps drop below 19°c). Absorption chiller (silent cooling when temps rise above 25°c, though these things are expensive , finicky , and less efficient than heat pumps still, the idea of using miner waste heat to cool your home feels cool on paper). clothes dryer (redirect hot water during laundry cycles, but only if immersion tank temp stays under 75°c. if it spikes, the system dumps heat outdoors via an exhaust fan).
note: this part is highly theoretical. i haven’t tested the absorption chiller integration or thermal loop dynamics yet. control systems (the “i hope this works” part) temperature state machine
a plc would comtrol load and toggle between modes:
Heating : send hot water to radiant floor / home radiators. Cooling : activate absorption chiller. Dry_clothes : redirect heat to dryer. Exhaust : trigger outdoor fan at 75°c.
load balancing
priority order:
Home loads (lighting, appliances). Exhaust radiator (if immersion tank >75°c). Battery charging (only if above minimum reserve) 4–7 miners enabled sequentially as solar allows. Sell excess to grid (lowest priority).
btc mining is higher up than grid sales, but still behind critical loads. makes sense when mining rewards > dumping power for pennies.
current status
this is all just draw.io diagrams and logic flow right now. no hardware built yet.
big unknowns: how absorption chillers handle variable thermal input from miners. whether the staggered load control can prevent immersion tank overheating without constant manual intervention. safety risks (e.g., dielectric fluid leaks, electrical load limits.
Why share this early?
I'm curious:
Is this even feasible, or am i missing obvious flaws? Are absorption chillers worth the hassle, or should i pivot to a simpler heat pump-based system? How would you prioritize thermal recovery vs. mining uptime?
thoughts?