Glenn Herbert

I agree with moving the raised beds off of the concrete. Are there trees close to where you were standing to take the photo? If not, the grassy foreground would probably be a good location for at least some of the beds. I would consider placing some or all of them just at the edge of the concrete at the far side; this would minimize shading from any trees behind you in the foreground. Add some boards or rocks to level the far side of the beds where the ground slopes away, with the near side just touching the concrete so you don't have to worry about weeds there.

"a very crude stone foundation that is bound to heave all over the place in the winter"

This is the place to focus your work on. No matter how good the cob or earthen structure, an unevenly shifting foundation will destroy it. You need to make your foundation deep and wide enough to be below the frostline. If you have expansive clay soils that swell when wet and shrink when dry, there is probably nothing that you can do to make a rigidish structure last for the long term. In that case, if you have to build there, I would go with something like a timber frame that can flex a bit, and infill with mass/insulating material in smaller panels that can shift independently a bit.

If your foundation is shallow, you may be able to use a Scandinavian-developed technique of digging down a foot or so outside, and running insulation down the wall and horizontally the same distance you would ordinarily go vertically.
Depending on your situation, I might consider raising the planned floor level, adding enough rubble to support that, and filling around it with earth sloped up to protect the final depth from frost.

If you will be ordering new glass, you should be able to get stock sizes that come out even with the overall wall width, or adjust the walls by a few inches before construction layout starts. If getting recycled glass from a Restore or the like, it should be possible to do some figuring on the spot to select items that will fit the overall space available.

This concept is much less "architectural" but much more practical, and I think better for function. The pretty sloped glass greenhouse look would as others have mentioned have been a bake oven all summer and much of the spring and fall, whenever it was sunny, at least without additional thermal mass and ventilation expenditures.

One thing that concerns me is the notion of center bolted glass... unless it is extra strong, drilling glass for bolt-through mounting will create an extreme stress concentrator and weak point. It would also mean only single-pane glass... in Montana. Is there any glass resource already identified? Will it be bought new? Recycled from other sources? If not ordered new, the glass will need to be on hand before finalizing the window dimensions. Letting glass run past the edge of the wall framing is asking for it to be bumped and broken. I would be concerned about that in any circumstances; in a busy multiple-use area, it would only be a matter of time. I think this would be a good place for a concept from one of my professors in architecture school 40 years ago, bypassing systems: each part of the building operates on its own module, and may "float" relative to other systems. The roundwood framing has an optimal layout and fixed overall width. Let the window wall, above the roundwood base, be set a few inches in front of the posts with its frame dimensioned according to the glass actually used. With thoughtful arrangement of the final materials this could look attractive and interesting. Horizontal roundwood at the base of the window wall would naturally go with the windows centered above the logs and not pressed back to the posts leaving a big snow-collecting shelf outside. (I note that the sketches appear to show something like the positioning I suggest.)

The roof as sketched would be quite low slope, and require some modern technology to be waterproof. Aside from commercial rubber roofing or an asphalt built-up roof, I think the best solution might be a green roof, using the billboard material as done on the wofatis with a thinnish layer of soil and plantings to hide and protect it, and give a taste of the uncommon building methods and materials employed on site. It could not follow full Oehler design, but could be a viable option for locations where that is not practical. For the structure size contemplated, making it strong enough would be easy. It would mitigate summer heating in the space, and while not giving much insulation for winter, could probably be combined with another insulation method... scope for thinking about leading up to construction. The structure can be designed with a certain number of inches between ceiling and roof surface to be filled with whatever is decided upon.

A quick sketch shows that four octagon rooms fit neatly together giving a tiny square center room. This would have a very jagged outer perimeter, and would in practice morph into a nipped square overall shape. The center would be too small to be functional unless the outer rooms were quite large. So I think the outer rooms would have to be some version of trapezoids.

I would group the windows so that two outer rooms shared a window zone, putting the windows at one corner of each room. This would make only four exposed window sections, and longer stretches of uninterrupted wall which could be very thickly bermed. To reduce hard-to-use corners of trapezoids, the corners would be further clipped to give 16 sides to the outer wall.

I would expect the roof to be earth covered for camouflage from the air, with membranes like the wofatis, so steep slopes not required. In fact, steep slopes would be problematic if earth covered; there is a tendency for earth to slide off unless the underlying slope is gentle. 1:12 is the recommended maximum slope for thinner earth coverings (green roofs) to stay in place. Thick coverings which merge into berms at the base may be able to be steeper.

That is just an ordinary strap hinge and pintle, with the strap bent so that the pivot is centered on the gate frame.

You could cut a slot in the gate so the strap fits into it without bending, for an even stronger hinge structure.

A really simple way of spanning a 12+ to 16" space is to corbel in the walls until a standard firebrick can span them. 9" long firebricks can be turned to project 3 or 4 inches into the space, and if the ends are 8" apart, 9" firebricks will close the top safely. This gives a clear interior width of 14" or 16". Add another course of firebrick continuing the original walls, to weight the corbelled bricks and strengthen the construction.

Since the water should never be in contact with the combustion core (so as not to steal heat from the fire and cool it and make it less efficient), the same setup I suggested before would work with an insulated firebrick core.

You didn't mention before that you plan to use exclusively woodchips as fuel. This makes your feed tube sensible, and also changes the parameters for how your firebox should be designed. Other people on Permies have experimented with woodchip feeders, and you might benefit from researching their efforts. An exposed metal feed which would tend to shed excess heat easily may be good, while a firebrick burn tunnel would let the heat build up for efficient combustion. I think you would want a smaller air supply opening. Reading about batch box primary and secondary air supply at batchrocket.eu should help you figure out a good system, even though it is geared to burning solid wood.

Given that it appears that you only want to heat water, not have a solid mass, I think any coil around the "barrel" is going to miss a large fraction of the heat harvesting. What I would suggest is a larger barrel with a hole cut in its bottom and welded or sealed around the lower edge of your existing barrel. Fill the outer barrel with water up to above the top of the inner barrel, and you will have a "ring" of hot water that you can pipe directly to your spaces for radiation, or submerge a coil in to heat the circulating water.

I note that you have copied the typical youtube rocket stoves made of metal. These are inherently less efficient than a properly built rocket mass heater which has to be made of masonry or other non-metallic refractory material, and the combustion core highly insulated. If you insulate your metal rocket, the heat will destroy the metal probably in a year or so. If you don't insulate it, you will get a less than optimal burn. The lower heat range of a metal core may contribute to the lower than desired draft which is a factor in pulsing. An uninsulated riser inside the barrel will let heat radiate freely and equalize the temperatures in riser and barrel, which negates the "push" you get from the barrel cooling the exhaust gases and helping with the draft.


Another thing - I understand that corrugated exhaust pipe as you show for a chimney may have significantly more friction than smooth stovepipe, and contribute to lower draft and thus pulsing.

A flat roof requires waterproof material like concrete (including perhaps some type of coating). If you give the roof some pitch so it drains quickly, you have many more options for material as long as you have water-shedding surfaces on top. Brick arched vaulting with clay tile on top is a commonly done example. How much rain do you get? How often, and for how long at a time, does it rain? Is there a rainy season? The less heavy, extended rain you get, the wider the options for natural, non-concrete mass structure.


Your adobe house was designed backwards for your purposes. The tin roof would have largely shaded the adobe walls, so that the mass would never get hot. If you have enough rain that a tin roof makes the most sense, putting it over an earthen masonry vault, either in contact or with a little air space, might transfer enough heat from tin to vault by radiation to warm the roof mass. I would want to test a small sample on a shed or something to see how that worked.