Embedding a heating cable in cob

Subject says it all. Has anyone done this? It seems like such a simple way of getting some "low and slow" heat into a room, especially if the space is well insulated and not a candidate for a small RMH. It's also a potential way of using free or cheap power from solar panels or time-of-day metering to warm a mass of cob, which then releases the heat gradually into the living space.

This has possibilities! I have long mused about multiple heat sources feeding a single thermal mass.

As I see it, the challenge is making and keeping a thermal bond between the wire and the mass, for efficient heat transfer.

If using something inexpensive like nichrome wire, will there be unexpected chemical interactions with the clay? An oxide layer forming that insulates against thermal transfer?

I thought about plain old Nichrome for about a minute and then dismissed it because of the possibility of mineral reactivity and potential short circuits. I guess there might be a way to string it through a series of ceramic tubes (sort of like the internet) that fit properly but keep the cob mix from fouling the wire.

What I'm looking at right now is a length of insulated parallel core wire like these. Take about 20 m of that and run it back and forth as I build up the wall, then embed a thermostat to monitor the cob temps as well as one for the air in the room.

Of course, since the application I'm building right now is a cob panel in a stud frame, I have an imperative to not let the mass heat up too much. That means it needs to be self-limiting and not reliant on a device like a thermostat and/or software to prevent thermal runaway, so maybe some of this stuff.

Hey Phil,
I like the idea of using overproduction from PV stored in a thermal “battery”.
I see a couple of challenges though.

In a RMH the cob heats up to a much higher temperature, especially around the ducts. So I wonder how much output a warm wall would actually have if the inside is only heated up to 30 C or something.

The self regulating wire will cap at way too low temperatures from what I know. It’s mostly for frost protection. Also the output is only 30W/m at 10 C. So as your wall slowly warms up that will decrease even further.
With the 20m you mentioned that means 600W in the beginning, then dropping. That’s not a whole lot.

I know it's not a lot of energy, Benjamin, but it's a small and well-insulated space (will be, anyhow...it's under construction right now). 22 square meters, and for comparison I can warm my office of about 15 m2 in just a few minutes with a tiny forced-air heater on its 500W setting on a cold morning. What I'm hoping is that the self-limiting cable can get the mass up to 40-50 degrees and just hold there, and that will warm the room to 18-20. What I need to do is sit down and do some calculations, possibly substituting higher wattage cable values to pump more heat before they start ramping down. There are many on Alibaba that put out 60W/m...what i need is the curve that shows how much the resistance goes up at the temps I want them to work at.

Yes, those diagrams would be helpful. I don't know if the ones on Alibaba provide that though. I'm sure the traditional company would.

How much surface area is that wall compared to the 22 m² floor?

It's a super interesting idea, please keep us posted.

As it happens, I was just chatting with a helpful sales rep who sent me a chart for their medium heat range cables:



She quoted $1.40 US per meter and I think I would use 20 m of the 30 W/m version. At 80 degrees its output is halved, so that would be a 300 W heater and all I need to do is figure out if the dissipation from the surface would be anywhere near equilibrium with that level of input, since I don't want it getting much hotter than that.

The radiant wall section is 2 m2 and it's a stub that divides one part of the room from the other, like cutting into the angle of the letter L. It's roughly in the middle and open on both sides (which is what got me thinking).

It should work. My concern is to make sure you use a good heat wire that will last, because repair will be rather destructive.

There is a company that does that commercially with forced air through ductwork embedded in concrete (think rmh bench fed from an electric heater). It’s a closed loop system so the heater keeps adding heat to the air to get the mass way warmer than you would think possible. They are heating 2000 square foot houses in Canada with 1500 watts. They are doing it PV, too. I think it is a Swiss or Nordic company so the price and amount of concrete used is more than most permies would like, but the ideas could be applied to cob benches and floors. Run 3 or 4” drier duct loop through the bench and you can feed it with a small ceramic heater from excess pv. The part I like is anything that might break is accessible and replaceable in the future.

Ok, that got me thinking again...I'd try to do it without the plastic, so maybe a thin-walled metal conduit would do the trick (better thermal transfer, too). The first segment of the loop could be ceramic and have a Nichrome element inside. Off to look up materials.

[UPDATE] I found a 25m roll of 20mm flexi galvanised conduit...for $40! This reel would go for over $600 new. The plan is coming together: embed a length of conduit in the cob wall, run a heat wire through it, hook up a basic analog thermostat, and see how it flies. If the wire goes bad (25 year warranty), use it to fish a replacement line through the coil. All up, I expect to spend under $100 on this rig.