Crypter-Heating a Colorado Mountain Home
@May 8, 2026
A Bitmain Antminer L7 has been running in our Colorado crawl space for about three years, hashing Scrypt and recycling ~3.4 kW of heat into the home's ductwork. The natural gas furnace stays as primary heat when needed; the miner is a continuous baseline contribution. Here's what worked and what I'd tell anyone trying the same thing.
The setup
- L7 sits on horse stall mats over a pallet in the crawl space — no slab underneath, just the mats absorbing fan vibration.
- Insulated flex duct ties the exhaust into the home's supply trunk via inline boost fans.
- Inlet: two standard furnace filters mounted on a cardboard box. Cheap, replaceable, low friction.
- Carpet cushion draped over the miner and inline fans as an airborne-noise blanket.
Noise
It's loud, but manageable. Three things made the difference:
- Inline boost fans shifted the dominant frequency from a high server whine to a lower hum. Total volume didn't drop much, but low frequencies are far less irritating and carry less through floors.
- Duct geometry matters. Biggest practical diameter, shortest run, no choke points or sharp 90s. S-curves dampen noise but add friction — use sparingly.
- Mechanical isolation with the mats handles cabinet vibration; the carpet cushion handles residual airborne noise.
Use thermal-rated insulation on duct runs through living space.
Sharing ducts with a furnace
Outlet shared, inlet not shared — separate inlets prevent the miner from re-ingesting its own exhaust.
The furnace blower out-pressures the miner's inline fans, so when the furnace fires it back-pressures hot air into the miner's duct. A simple deflector at the shared outlet keeps furnace flow heading the right direction without a powered damper.
In the temperature logs, when the furnace cycles on around 07:00, duct temps rise slightly and the L7's fans spin up to compensate. Predictable and harmless.
Dust
Visible dust on the inline fan blades after one year despite the inlet filter. Plan on a wipe-down annually; the miner's intake screens as often as necessary. Mine have lasted three years and still look good. This will largely depend on your air-environment.
Economics
The numbers from three Colorado winters:
- Gas bill: down ~$100/month in winter.
- Electricity bill: up ~$250/month.
- Mining revenue at instant sell: ~$150/month.
Net: roughly break-even in winter, with the heat as a free byproduct. Summers run slightly negative — the miner's still consuming power but the heat is a liability, not a benefit. In Colorado that's a fine trade because the winters are long; in a milder climate the math gets ugly fast.
The constant multi-kW draw is also an unusually friendly load profile for grid-tied solar — no storage needed when the load is always there during sun hours.
Summer low-power mode pulls ~⅔ power for ~⅔ hash. Efficiency in J/MH barely changes; the real win is cooler chips and quieter fans when you don't need the heat.
The hardware recouped its purchase cost over the three-year run, but only because coins were held and sold opportunistically rather than dumped at break-even.
Failure mode
At ~3 years, one of three hash boards started flaking — intermittent at first, recoverable with a reboot, eventually permanently dark. The other two still run. Budget for one board failure inside three years.
Things worth knowing before you start
- Circuit sizing. 3.4 kW continuous wants a dedicated 240 V circuit. Don't share with anything else; don't run it on a tired 15 A 120 V receptacle.
- CO and smoke detectors in the crawl space. Unattended multi-kW gear next to a gas appliance — cheap insurance.
Verdict
Worth it if your winters are long, your electricity rate is reasonable, and you can hold coins long enough for a favorable market.
