Glenn Herbert

"It depends"... on what the load on each beam is, and how they are attached. How tall is the building, and how much floor area does each beam support? Do you have any idea what kind of wood the original beams are? Softwood or hardwood? How is the structure above connected to the beams, and how will you support that to replace the beams?

Also note that Peter's short article links to a comprehensive MHA report with lots of process photos.
2015 MHA Annual Meeting - Batch fired Rocket Bell

My feeling is that, if the foam insulation is used on all sides of the frame for a tight fit, there will be little air movement and little condensation. This of course depends on the specifics of the climate...

Here is a report from the (US) Masonry Heater Association with process photos: Kuznetsov stove
A pretty complex project with not much discussion/direction in the report.

A workshop by Igor Kuznetsov and Alex Chernov with process photos and drawings.
Complete layer-by-layer drawings, but a complex design with lots of precise brick cutting needed. You probably need to be experienced in masonry to follow it without handholding.

Bell with dead-end benches MHA workshop by Peter van den Berg with photos and drawings.
A simple design with the few critical details explained, and firebox design and construction covered thoroughly elsewhere in the same website.

"Gotta get a good scooper made."
I found that a tuna can tweaked to have a flat edge, and its folded-up lid taped to a short stick, 8" or so, made a fine scooper and took maybe five minutes to make. I got a sardine can to make a proper scooper, but the makeshift worked so well that I have used it for three years now.

Beautiful dam! I wish I had some areas that were wet enough and flat enough to do that. All I have are well-drained fields and hillsides with deep ravines that need big rock to begin to stabilize them. Oh, and a creek that can move 80 foot white pines right through my property in major floods, and tumble 2-ton rocks if they are not well bedded.

I don't think it will be an issue on your small watershed, but my impression is that many of the sticks in a beaver dam are laid parallel with the water flow, which would resist high flow better than if they were crossways to the flow.

How hilly is your property? That strongly affects the most appropriate options.

The character of the stone is also important, as in how well it stacks, if it is flat or round, etc.

The riser inside a bell or barrel does not depend at all on heat coming in from the bell; it should be insulated and preferably lightweight so that the immediate heat of combustion warms the inner surface very quickly, ideally coming up to full operating temperature within 5 to 10 minutes. Heat will slowly pass through the riser wall, and will warm the bell some but hopefully not much beyond the hot gases passing down around it.

If the downdraft tube is similar to the riser, hot gases will move quickly through it. It will all get very hot, since the gases will have moved on before transferring much of their heat to the tube. There will be more heat left for the following duct or channel.

A bell, being very large relative to the riser, will let the gases move slowly down through it, giving up heat to the much larger internal surface, and leaving less for the following channels. This is the same for a barrel or a masonry bell, the difference being that the barrel lets the heat pass quickly on to the room while the bell passes the heat through layers of masonry on its way.  Gases do not move as a plug leaving the perimeter undisturbed; they generally move as a whole, with some degree of turbulent mixing depending on the design.

I would never build a masonry bell without including an access panel for inspection, cleaning, and if necessary, riser repair or maintenance. This panel if sized and located appropriately would serve some instant heating function.

We don't know how well that system really works, or if it depends on assistance from the down channel at all. Being all one brick thick, that system would shed heat fairly rapidly from all surfaces, and might get some down channel cooling, at the cost of less heat retention in the riser and less complete combustion.

18" x 18" would be easy to cover with a corbeled brick arrangement, similar to the one in the link I gave. Or you could corbel in the top of the chamber by a couple of steps to say 9" x 9" and cover that with a plate, for easy inspection and maintenance.