Glenn Herbert

If I were heating an IBC with an external hot air source, I would run the flue in horizontally near one corner and around the walls to maybe the opposite corner and then up. Maybe around three walls, if the exhaust after a two-side run was still quite hot. Holding the flue about a foot off the bottom would allow feet to stretch underneath for more roominess. A bench built a few inches above the flue would protect skin and give a needed function of seating. I would give the flue a slight constant slope with a weep hole outside the IBC so any water that gets in has a place to drain. I would make sure the flue leaving the riser top slopes down an inch before meeting the IBC so that water cannot get back into the core.

For best draft and general function, I would try to have the J-tube sunken relative to the IBC, so that the exhaust could go (ideally) horizontally or else down a bit and then horizontally from the riser top to the IBC inlet. A hillside location, digging down a bit for the J-tube, or raising the IBC on a platform, or any combination, would ensure easiest function.

I'm sorry to have to tell you that a 20# propane tank is not going to be tall enough to enclose a functional riser. Angled feedtubes leading directly to the base of the riser eliminate one of the features of a rocket which is the wood gases encountering sharp bends which promote mixing and better combustion. If the wood is burning in the feed before it reaches the bottom, the flames will need to be sucked down and you will need an even taller riser to keep the draft going the right way. Starting such a beast will also be problematic - you may never be able to get a fire that wants to go in the right direction. You would need to insulate the feed at least near the bottom, or the water around it will steal too much heat from the fire.

Recommended proportions for a J-tube core are in the range of 1:2:3 or 1:2:4 or 1:1.5:3 (feed tube to burn tunnel to riser), and in any case the riser must be much taller than the height of the feed opening from the base, or you constantly risk the fire burning backwards. A 4" system will consume its fuel so fast that you will have to feed it every ten minutes at the least. Can you have someone sitting outside feeding the fire while someone else enjoys the tub?

Given the size of an IBC, I don't think you can fit an effective heater in it and have room for people. I think the way to go would be to make a standard J-tube core outside, heavily insulated, and run its exhaust in and up through the tub. It could go through a propane tank for a place for gases to slow and transfer heat. You would need a metal plate at the penetration to avoid overheating the plastic, but with water cooling the plate, a couple of inches clear around the flue should work.

Small systems might be trickier to start, that would depend on the details, but principally, you run into surface-to-volume effects. The smaller a system is, the more surface there is to create drag and take heat from the combustion stream, relative to the amount of heat generated. Volume is relative to the square of the surface, so combustion quantity is less and generates significantly less heat. At the same time a smaller system has less energy to flow, the surface drag increases. A smaller combustion volume will not be able to get as hot as a larger one because heat is bleeding out of it faster than it can build up, and you may not be able to get to complete combustion temperatures, leaving unburned gases in the exhaust which will condense to creosote.

You know what a regular red brick feels like, right? Hard firebricks feel distinctly heavier than that, and soft insulating firebrick feel considerably lighter. Insulating firebricks can be scratched with a fingernail, while hard bricks take work to cut with a masonry blade on a saw. Insulating firebrick would be ideal for a core, as long as you can be gentle with feeding wood. The feed tube probably wants to be lined with hard firebrick splits, or at least the top courses made of hard firebrick. You do need to mortar to seal all cracks so no cold air gets in where it is not supposed to. The fireclay mortar will work fine as long as you can keep the assembly from vibration or shocks and shifting. It is a sealant, not a cement or glue. A layer of straw-reinforced cob around the core will keep things in place even if you have IFB and additional insulation is not needed.

The 180 bends in a single 5" flue are definitely problematic. It is generally said that parallel flue paths are a bad idea, because they can never be identical and one will always end up with more flow than the other. However, short parallel flues may work fine in your case, as the friction in them will be much less than the friction in the connections before and after. Running three or even four 5' lengths of 5" duct will probably work fine, with a plenum at each end for gases to equalize. A few inches of cob over the assembly and you should get good results. You don't want more than a few inches of cover, as that (like 6" or so) would take hours to conduct heat to the outside and start warming your shop, and keep warming it far into the night after you leave.

A 6" firebrick J-tube can be tricky depending on your material. Insulating firebrick can easily be cut to whatever dimensions you like with a hacksaw, but hard firebricks are constrained by available sizes. If you cannot get a square cross section in the feed and burn tunnel, it is recommended that it is better to have it taller than wide. Standard firebrick dimensions will let you have 4 1/2", 5", or 7" height in various combinations of orientation. Using splits (4 1/2" x 9" x 1 1/4" thick) in the mix, you can get 5 3/4" high which would be close to perfect.

Addressing your overall situation, I think you would be much better off, both in reliability of function and in warmth given, going to a 6" system. If you have some more firebricks, you can make a J-tube core that is not much larger externally than your 4" core. A 6" J-tube can be reachable for cleanout without special ports. For workshop heating, you probably want more instant radiator than storage mass.

A batch box core will give more heat with less frequent loading, but that is a considerably more technical build. I would do a J-tube and use it for a season or so before trying something more ambitious.

For the bell, if you want to use it as a bench, you will need some mass so the surface does not get too hot. Two or three inches of cob, or a layer of old red bricks or any other masonry, should take care of that, and even out the heat release without delaying it too much. A length of rectangular HVAC duct would work as the liner/base of the bell, ideally 12 to 16" wide by 8 to 12" high by whatever length you have.

To be able to keep the barrel you have, you would need a thinner insulation layer on the riser. Ideal but costing some money would be a "5-minute riser" made from a length of 8" stovepipe lined with 1" thick ceramic fiber blanket.

Welcome to Permies, and to rocket science!

What you show is essentially a 4" diameter J-tube system, which is at the extreme small end of what can work for a rocket mass heater. Most who are not experienced in RMH construction have difficulty making such a small system work.

I appreciate that you are using masonry for the feed and burn tunnel; given the small size of the system, you will probably not get the very high temperatures that quickly destroy metal in the riser and other parts of the core. What kind of firebrick are you using? Heavy dense ones, or soft light ones? Hard firebrick will pull heat out of a small system as fast as it is generated, and you may have difficulty getting a strong fire going. The core needs to be insulated as much as possible so the firebrick can build up heat.

A 4" J-tube is going to be difficult to reach in to clean ash. I think you need to add an access port at the base of the feed tube so you can reach in with a tool to rake the floor of the burn tunnel.

Considering the system size, the 2" gap between riser and barrel is probably sufficient. The 15' of duct is problematic, though. I don't think it will allow a strong enough draft to work well. I would advise switching to a "half-barrel bell" covered in cob or other mass. This will have essentially zero drag, and allow the best possible airflow.

Jay is not on permies any more. If you ask good questions with supporting information and possibly photos, someone will probably be able to help.

I think if you want to get more heat to the workshop side, filling the tile cores all the way to the floor would be helpful, so there would be no more air spaces to insulate the wall.
If you only want to prevent heat rising in the cores, you could stuff the cores with rags or something... umm, maybe something noncombustible ... and only need a small amount of grout above the level of the heater top.

I think it would be fairly easy to cut into the tops of those hollow tiles and pour thin grout into the cavities to fill them. The question is whether that would prevent the wall from getting hot above the heater level. I am not sure of the answer.

I have figured out the second IR photo, with the dark silhouettes of pipes running across the top. The 146.9C reading is on the surface of the wall just below the beam. The exposed beam is considerably cooler, but the back of the beam in contact with the wall would be as hot as the wall. I think quick action is advisable.

By the way, I'm sure you know this, but hollow tiles I have seen have multiple rows of chambers, and you would need to drill through all of the partitions to allow the closest chamber to the heater to be vented, along with the rest of them.

Rereading, you did specify "nearly all the way through", so you and anyone following the thread for advice will be good to go.