Controlling ventless millivolt ignition fireplace from 24 volt smart thermostat

Question

I've seen questions similar to mine being asked before but every solution appears to be a "hack" that is likely to fail shortly after installation. I'm hoping that someone knows of an "industry approved" solution that has eluded me so far.

The goal is to have a ventless natural gas fireplace with millivolt ignition as emergency heat and as a source of humidity, and continue to use my current 24VAC HVAC furnace as primary heating. I doubt it changes anything about the solutions that I intend to humidify my house with a ventless fireplace but I thought I should mention it out of completeness.

Like many of the solutions I've seen with a similar intent I'd want a "dumb" thermostat in parallel with the primary "smart" thermostat so that if there's a power outage while I'm away or asleep that the house doesn't turn into an icebox. The question is then how to attach a millivolt fireplace to a 24VAC thermostat.

The most common solution I've seen is to get a common and inexpensive fan control relay to close the millivolt circuit on the fireplace. While this certainly works for some definition of "work" I'm concerned about running DC through relay contacts rated only for AC. I've seen this done before and the usual outcome is the contacts get burned out fairly quickly from the DC arcing.

Another popular solution that concerns me is seeing people use 24VAC dry contacts in a thermostat to control the millivolt ignition circuit on a fireplace. There's a reason that not all thermostats are rated for millivot DC, DC can burn out AC switch contacts or overheat a solid state circuit.

I have two smart thermostats already, one that provides a 24VAC output for humidity control, and the other has 24VAC rated dry contacts. I should be able to use one or the other for this project, in addition to a millivolt "dumb" thermostat to fall back on in case of an extended power outage.

Should I expect a fan control relay to work fine to operate a millivolt fireplace? Are there relays built for my intended application that I simply have not found yet? If so is there a specific nomenclature for these devices I should be using to find them more easily? I'm not seeking product recommendations but rather what I can expect these relays to be called so I can search for them at a hardware store.

I'm quite certain that the dry contacts on the one thermostat I have will not react kindly to a millivolt fireplace. I'll have a long run from the fireplace to the thermostat and that will call for 14AWG wires to minimize voltage drop, and the contacts on the thermostat don't appear large enough to accommodate that size of wire. I've been trying to find out how much current flows from a typical thermopile but haven't seen a useful number, but if they want 14 or 16 AWG wire then it must be significantly higher than that seen on 24 VAC systems where 18 or 20 AWG is the norm.

In trying to work the problem backwards I thought I'd see how much power a gas valve takes since that should tell me how to size up the relay.

I found this as something I believe to be representative of a common fireplace gas valve: https://images.thdstatic.com/catalog/pdfImages/25/2582b4e1-4efa-4fcc-a7c7-7001b6dba1c1.pdf

This table is from that spec sheet but I added the power column.

ELECTRICAL RATINGS		Power
24 Volt Models	12 VDC - 0.18 amps	2.16 W
	24 VDC - 0.2 amps	4.8 W
Millivolt Models	250 MV to 750 MV	???
Line Voltage Models	120 VAC - .034 amps	4.08 W
	240 VAC - .017 amps	4.08 W

If I assume the voltage drop to the valve results in 250 mV at the valve and it's consuming about 4 watts then that's 16 amps. If I assume about 750 mV at the valve and it consumes 2 watts then that's about 2.5 amps. Given that I've seen people claim they've been using fan control relays rated for 240VAC 6 amps and not have them immediately melt down then it seems the numbers are at least close.

I can keep doing the math but what I'm really searching for is a spec sheet on a few thermopiles that list the current they produce in addition to the voltage. It seems odd to me that such a fundamental detail isn't shown on spec sheets so that installers can select the proper size wire and such for an install.

Answer 1

Simply pick a relay with the correct ratings

The answer to this question is that "if the fan control relay doesn't fit the bill, you can just pick a different relay." An AP&C PAM-1 or Functional Devices RIBMNU1C (the latter uses 2.75" snaptrack aka MT212 for mounting) will do the job nicely here -- both relays have DC rated contacts, coils that can accept 24VAC, and full UL listings for field wiring use.

Answer 2

Let's talk about relay contacts

As you noted, relays are available with a wide range of ratings. One reason for this is the physical effects that occur when the contacts are opened -- or closed. Arcing occurs virtually always when a relay opens, and can occur when it closes, too, and the terminals inside the relay are designed for this. We expect some damage (pitting) to occur from the arcing -- but did you know arcing helps to clean the contacts, too? There are a variety of base and plating materials used in relay contacts. Some are great at handling higher currents, or more resistant to pitting damage, but perhaps are more susceptible to oxidation. Arcs can help clean the oxide layer. A relay operated with too little current will eventually fail due to oxidized (electrically insulating) surfaces on the contacts.

What's millivolt?

A thermocouple is a device that converts heat energy to electrical energy. A thermopile is a whole bunch, of thermocouples (tens to hundreds) joined, usually in series, to produce the whopping voltage of typically 0.25 to 0.75 volts - or, in scientific terms, 250 to 750 millivolts.

A great question is "okay, but how many amps are we talking?" The answer seems quite elusive, but an article on All About Circuits provides a clue: (emphasis mine)

The voltage produced by thermocouple junctions is strictly dependent upon temperature. Any current in a thermocouple circuit is a function of circuit resistance in opposition to this voltage (I=E/R). In other words, the relationship between temperature and Seebeck voltage is fixed, while the relationship between temperature and current is variable, depending on the total resistance of the circuit. With heavy enough thermocouple conductors, currents upwards of hundreds of amps can be generated from a single pair of thermocouple junctions! (I’ve actually seen this in a laboratory experiment, using heavy bars of copper and copper/nickel alloy to form the junctions and the circuit conductors.)

If that's so, then perhaps the only guide as to current we have is resistance of the manufacturer's recommended wiring. The gas fireplace manufacturer Montigo mentions in their FAQ:

The supplied 18/2 wire will work up to 25-ft in length. For longer lengths, an increase in wire gauge is necessary to overcome the voltage drop.

Make some inferences

A table at hyperphysics indicates 6.3 ohms per 1000 feet of 18 AWG solid wire, so Montigo's 25-foot length (50 feet of conductor) has a resistance of 0.32 ohms.

It's interesting to note that the wire-sizing table in the Robertshaw installation data you linked is comparable. Note that I'm doubling the lengths to account for the "there-and-back" resistance of the entire circuit. For 200 feet of 14 AWG about 0.51 ohms; 128 feet of 16 AWG about 0.51 ohms; all the way along to 32 feet of 22 AWG at 0.52 ohms.

Maybe the best we can do for estimation is to suppose a high-performing thermopile produces 750 mV (open circuit) and somehow sustains that voltage while delivering power into a circuit consisting solely of a 0.32 ohm wire. Ohm's law says 2.3 amps will flow. Recognize that we've ignored the internal resistance of the thermopile as well as the internal resistance of the gas valve so actual current should be less than this.

Montigo's recommendation

Incidentally, at that same web site, Montigo offers the suggestion below for controlling their fireplace with a timer or WiFi switch. They don't appear particularly concerned about the precise choice of relay. Indeed, it seems your worst-case outcome is that the solution eventually quits working due to oxidation of the relay contacts. Folks report the same problem when their millivolt fireplace is controlled by an ordinary mechanical light switch instead of a special millivolt switch. Burn-out or other catastrophic failure seems to not be a thing with millivolt appliance controls.

Answer 3

This is straightforward. You leave the real millivolt thermostat in place (and set lower, so it will kick on even if AC power is out, to prevent the building and water pipes freezing), and parallel it to a relay with a 24 volt coil. To fire the fireplace, send 24V to the relay, and any smart thermostat can figure out how to do that.

Wire and contacts must be heavy because voltage drop is absolutely devastating when you only have 600-750mV to start with. Current is certainly under 2 amps. (For instance, a 25' thermostat lead with common #18 thermostat wire has 320 milliohms. 1 amp of current would drop 320mV, devastating to a 600-750mV source.)

For a 24V supply transformer if needed, any plug-in 24VAC transformer will suffice.

For the relay, there's no kill like overkill, because millivolt circuits are VERY susceptible to voltage drop. There are plenty of heater control and HVAC relays and contactors made for controlling quite high current AC loads. And they're CHEAP due to economies of scale, and are instantly recognizable to anyone in the HVAC business. Some are 2-pole and feel free to parallel the poles.