There exists DC motors of various voltages (12, 24, 48, etc) and various powers (1/2 hp, 1 hp, etc). They seem to have the necessary power and RPM to operate the blower in a conventional residential furnace. During a power outage a 12 V DC motor would also be easily powered by conventional lead-acid batteries with no conversion loss that you would have with an inverter powering a 120 VAC motor.

My primary question: Would an appropriately-sized DC motor be more efficient than the typical non-ECM 1/3 or 1/2 hp AC motor (PSC or shaded pole 120 VAC)?

My secondary question: Would a DC motor have the same durability and longevity as the AC motor for an HVAC application?

I am thinking of a way to either replace my existing AC motor with a DC motor, or mechanically creating a way to mount both and be able to activate the DC motor from one or more bank of standby lead acid battery during an extended power outage (where natural gas would continue to flow but the furnace would be rendered inoperative due to lack of being able to turn the fan).

  • 7
    \$\begingroup\$ I'd be very concerned about modifying a furnace in any way at all--mess things up and the gas could leak out unburned. \$\endgroup\$
    – Hearth
    Oct 4, 2021 at 2:57
  • 2
    \$\begingroup\$ @Hearth I think that the combustion and the blower are more or less totally separate ... the wall thermostat turns on the gas ... the blower is controlled by a temperature sensor in the fire box ... if the blower fails, then the over temperature sensor shuts off the gas \$\endgroup\$
    – jsotola
    Oct 4, 2021 at 3:35
  • 2
    \$\begingroup\$ Don't forget that you're also going to have to supply 24V to the control electronics in the furnace. \$\endgroup\$
    – brhans
    Oct 4, 2021 at 12:03
  • 3
    \$\begingroup\$ Agreed with @Hearth there are far more ways this could blow up in your face than there are ways to make it actually work. Be very careful. \$\endgroup\$
    – J...
    Oct 4, 2021 at 12:39
  • 2
    \$\begingroup\$ How will the less efficient DC motor be powered from AC during the 99% of the time it's running during normal operation? It seems like that outweighs any conversion losses of the DC->AC inverter used during the 1% of the time. \$\endgroup\$
    – spuck
    Oct 4, 2021 at 17:19

5 Answers 5


Brushed DC motors do not have the longevity of induction motors. They are also not particularly efficient.

You may be able to find a BLDC motor with built-in control that could run from DC and be more efficient than a small induction motor.

There is no motor type which can surpass an induction motor for longevity of service, although a BLDC can come close. There are, of course, cheap induction motors out there that may not last all that long. But all else being equal, the induction machine is the very model of reliability.

It is unfortunate that typical small induction motors are not all that efficient (75 to 85 percent or something like that). This is why BLDC may be worth a look. There are also permanent magnet synchronous motors sold in the same form factor as induction motors. But they require expensive controllers and still need AC mains input.

My suggestion is to stick with AC power and just use an inverter with your battery for backup. This is a lot more simple. Consider LiFePO4 batteries rather than lead acid. Lead acid batteries do have ventilation considerations when used inside a home. The sealed type are least likely to vent fumes or corrosive aerosols into your home.


Conventional DC motors with commutators are still somewhat viable for battery powered use, but they have steadily been losing market share because they are less efficient, less durable and have less longevity than AC motors. They once provided better performance for variable speed applications and had less complicated and more reliable electronic controllers. Newer AC drive technology has taken that advantage away. They no longer have a lower combined motor controller cost either.

Small universal motors may still be competitive for small appliances, but brushless motors are taking over that market also.

The bottom line is that the retrofit described is not a good plan.


Full points for looking for a way to avoid running a generator simply to keep a gas furnace functional.

Frame challenge:

You won't gain enough efficiency to save any money, and the complexity will be an albatross around your neck.

Wouldn't matter. You still have to power the furnace controls.

And those require either 120VAC or 24VAC depending on the section of the furnace they are in. That's 24 volts AC.

A UPS isn't going to be enough here. UPS's are sized and scaled for high current draw for minutes. (i.e. 600 watts of draw for 10 minutes, long enough to flush tables, spin down the database, spin down Apache, and 'shutdown -h now' the server). As such, they are not built to have "long legs"... they just don't have the internal efficiency, because who cares if the inverter uses 50 watts while idling, it only runs 10 minutes.

And you'll want an electronic converter anyway

You could possibly arrange something with a brushed DC motor. However, the best option is a brushless DC motor that is really an AC motor coupled to a variable frequency drive. Which is a type of inverter.

So we've come full circle back to inverters that make AC. Only your way, with a brushless DC, your inverter is only driving the motor, and leaving the rest of the furnace a huge unresolved problem that will require yet more engineering.

So cut out all the complexity, and you have a COTS AC inverter driving a COTS furnace system. Any furnace repairman can maintain the furnace using common parts. Even the inverter is common; 12V and 24V inverters are sold at any truck stop.

Why not just throw battery at it?

Efficient MPPT solar charge control tech is falling out of favor in small solar installations. The reason is simple: the MPPT inverter is pricey, and it's more cost-efficient just to rack more solar panels, because panels are laughably cheap these days.

So, think in a similar vein: For the cost of developing a hacked, one-off, unrecognizable to any furnace repairman custom solution, all to save a few percent of efficiency.... wouldn't it be cheaper just to rack more battery capacity?

And since you are using lead-acid batteries precisely because they are dirt cheap... the answer is unequivocally "yes".

And I'm not at all certain there's more than a few percent of energy to be saved, but that is not the point of this frame challenge.

The best answer of all: non-electric furnaces

Not to replace your primary furnace, but to be a cheap, simple, supportable, turnkey solution to power-out heat. So you can stay with gas primary, or it even gives you the liberty to go heat pump, in which case this becomes your "emergency heat". You can install multiples - they're cheap, and possibly even cheaper than a competent battery/inverter setup, depending on the cost of flue installation.

When AC power goes out here, my house is toasty warm. Because the furnace is one of several models that is gas and does not need electricity.

It has a pilot light, with a thermocouple near it. The thermocouple has hot pilot light on one side, and cool room on the other. That creates electricity in the millivolts. That energizes a gas valve, which opens the gas to the furnace. Whoosh! And convection does the rest. There's an external thermostat called a 'millivolt' thermostat, which is rated for the low voltage yet high current involved. It wires up with standard 18 AWG thermostat wire.

This particular one is a 50,000 BTU, double-sided wall furnace that sits in a standard interior 2x4 wall, and emits heat from both sides. It costs under $1000 (though of course the installation of the vent stack will add costs). However they make a variety of others - single-sided, floor furnaces, baseboard even.

They simply work. If there is gas, they heat.

enter image description here

It is a typical 70%-80% efficiency unit, which is because it is using the gas heat to operate the draft - to push air up the stack. This one draws air from the room (sucking in cold air from outside through the building's numerous leaky windows), but they make direct vent versions also.

I advise buildings with electric-dependent furnaces install at least one of these in a strategic location to protect rooms with water plumbing.

You aren't even locked into a millivolt thermostat. You can use 24V coil relays so a Nest (or any other 24V 'stat) can control the millivolt 'stat.

  • a 10A SPST relay, 24VAC coil, operated by the Nest when it intends to "call for heat" from the wall furnace (e.g. auxiliary or emergency heat). The Nest's 24VAC closes the relay and the big contacts handle the fairly large current in the millivolt system (to which the Nest is totally unsuited).
  • another 10A DPDT relay, 24VAC coil, fed from the same transformer as the Nest, so it is picked up at all times the main furnace has AC power. The NO contacts go to the above relay. The NC contacts go to a genuine millivolt thermostat.

Thus you could set the actual millivolt 'stat higher, so it would run a higher temperature during power outages to help natural convection carry heat around the building (since the forced air system is out of commission).


There are simple single phase BLDC motors for simple use. One I have saw was mounted on Husqvarna automower that propels the rotary knife cutter. They are produced by Emb Papst, I am not aware how exactly do they function, but it runs on Ni-MH battery and only one phase.

See: BG3612


I have no idea (other than reading the other answers) about efficiency of different motor types.

However, I don't think this plan is a good idea because you are adding significant complexity to a functioning system.

  • Additional conversion (AC-DC) for the new motor.
  • Backup power system for the motor, but still need backup power for the controls, igniter, etc.
  • Switchover circuit to switch the motor from utility power to battery backup.
  • Possible bypass and/or modification of sensors that make sure the motor is running to avoid combustion without proper air circulation. ("Possible" because this may be based on airflow or temperature sensors and not on detecting the motor actually running.)

Far better would be to either just put in a backup (battery or generator) for the entire system or install a secondary heating system that does not require any electricity.

  • \$\begingroup\$ The focus on the perceived issues with the fact that this is a furnace is not necessary nor useful for answering this question. I should not have to explain this, but this is a very simple furnace, standing pilot, no electronics, no draft motor, requires only 24VAC for gas valve solenoid. In the event of a power failure, only the blower motor (horribly inefficient 1/2 or 1/3 hp shaded pole) is what prevents furnace operation. \$\endgroup\$ Oct 4, 2021 at 15:16
  • \$\begingroup\$ I was under the impression that BLDC was more frequently known as ECM motors (at least in the HVAC industry). If I'm not mistaken, they are essentially stepper motors, and yes they have double the efficiency of an equivalent shaded pole AC motor, but they are pretty much impossible to obtain as they are not sold as replacement or "drop-in" motors in hardware and industrial stores the way that PSC and shaded pole motors are. \$\endgroup\$ Oct 4, 2021 at 15:19
  • \$\begingroup\$ A typical residential furnace, in my experience, has a lot of complex electronics - and has for many years. Igniter. Solenoid. Sensors for flame, overheating, airflow, etc. Even "24 VAC for gas solenoid" needs to come from somewhere and be controlled by something. \$\endgroup\$ Oct 4, 2021 at 15:46
  • 1
    \$\begingroup\$ We often have "X Y questions" - e.g., "how do I replace an AC motor with a DC motor" without any context would likely result in comments of "What is the motor being used for?" because that can have many ramifications. So knowing it is a furnace is good. Knowing more about the type of furnace (gas, standing pilot, etc.) is very helpful and should be added to the question so that everyone knows the parameters. \$\endgroup\$ Oct 4, 2021 at 16:05
  • 1
    \$\begingroup\$ I think ECM means electronically commutated motor. If so, that is a logical description of a BLDC motor. \$\endgroup\$
    – user57037
    Oct 5, 2021 at 0:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.