|
Post by pyrapodpaul on Oct 26, 2015 9:02:08 GMT -8
Hi all, I am about to go ahead and build a rocket-mass-heater with a difference and would welcome any input from those who know the subject well.
It is to heat a well-insulated square wooden cabin I have built, which is one open space. The idea is that it will be in the centre of the space and will heat a cob floor which surrounds it, the intention being that people will sit in a circular arrangement around the central heat source, much as people do in teepees, but without the smoke (hopefully!).
The cob floor will be octagonal and bordered by a wood floor running out to the walls of the building. My initial design is here: pyrapod.wordpress.com/rocket-mass-heater/ (I tried to attach this image to this post, but it didn't want to for some reason ) So, if you look at the drawing you can see that the idea is to construct the flue-gas passages using cob, capped with paving slabs or tiles, and then cobbed over the top. Kind of a hybrid between a standard RMH ducting system and a hypocaust/ondol system. I want to avoid using lots of metal ducting because this is a very low-tech project. So, is this idea practicable and safe?
Also, given the floor area of the space (350sqft overall, of which 83sqft is cob), could the system be scaled down to use a smaller barrel?
Any thoughts most welcome. I haven't done anything like this before... I am confident working with wood, but on this I am in uncharted territory.
Many thanks,
Paul (in UK)
|
|
|
Post by patamos on Oct 26, 2015 17:22:30 GMT -8
If you can get your hands on a bunch of old clay bricks (for hypocaust pillars and general shaping of tubes/chambers) that will save you a lot of mixing time.
Also, good to factor in the gasses having a mind of their own over that wide a surface area. Like an old river in a wide valley. The forces acting on their direction are wide open. The gasses may at times very well only want to head to one side. Pyrophile posted some russian articles about this earlier this year (or maybe last). Worth looking at.
For that and preferred heating considerations, i'd suggest putting adjustable baffles at the intersections… my 2 bits
pat
|
|
|
Post by pyrapodpaul on Oct 26, 2015 23:55:16 GMT -8
Thanks Pat. The clay bricks idea sounds good. To control gas flow, adjustable baffles could be good, I suppose I'd have to set it all up 'dry' and run the system to test it before I cob over it all. My main concern apart from whether it will actually draw well is safety. Is a 3-4" cob floor on slabs over the passageways reliable at containing the flue gases? I appreciate you mightn't know, but if anyone has any insight into this point it would be most welcome. Thanks again, Paul
|
|
|
Post by patamos on Oct 27, 2015 21:55:52 GMT -8
I think it depends on the fibre content, temperature and width of the hypocaust chambers. When portland-cement-based concrete is subjected to a temperature increase of more than 5c per minute it can spall explosively. Exactly where this quick heating zone ends is hard to say, but i'd suggest that the first few feet around the heater core had best be done with slabs of clay tile. Side wall of old flue liners maybe. Same thermal issue applies to organic fibre. Once above a certain temperature it will carbonize and no longer perform its integrating function. If you don't mind a bit of a synthetic compromise, adding little 2cm threads of fiberglass to the cob can be a functional alternative in those high heat areas. Concrete suppliers usually sell it by the bag for not much $. The greater the span of the hypocaust chambers the more a point load in the middle of the bridge can collapse it. But 3-4" of fibre rich cob is very resilient and ought to do the trick for any span up to a foot. The cob will also contain the flue gasses so long as your ratios are right and you do not have too much shrink cracking. For details on that check out Donkey's posts in the reference library. The challenge you face in getting the gasses to split evenly is not to be underestimated. In fact, i am quite certain that the gasses will most often prefer one side or the other, as they are prone to positive feedback dynamics. Perhaps you would be better off having a single direction spiral? thoughts for now
|
|
|
Post by pyrapodpaul on Oct 28, 2015 15:57:02 GMT -8
Hi Pat, and thanks again for your thoughtful and helpful responses. So yes, on your first point, I think I'll use fire-brick for the beginning of the passageways near the barrel. Re the fibre for binding the cob, I would be reluctant to use fibreglass but if it's necessary I'd have to. The cob above the cover-slabs though may be protected enough from the intense heat to not carbonize...? Thanks for the pointer to Donkey's posts on cob mixes and cracking, I will check them out. Meanwhile, and hopefully as a good solution to your point about directing the gases effectively, check out this modified design - I'm calling it the 'double-damper', further down the web-page I linked to before - pyrapod.wordpress.com/rocket-mass-heater/The passageways are 4" wide, separated by cob 'walls' which could be cobbed bricks as you suggested before. 2 gaps in the walls have a damper operating on them, moved by a rod running inside the exit pipe (no risk of gases escaping into the living space through a damper-lever-hole). The dampers would be open when starting the stove, allowing the flue gases can escape directly through the exit pipe. Then, once the stove is really going, the dampers would be closed and the flue-gases would flow through all the passages to get to the exit pipe. I am really trying to keep everything pretty simple. Do you reckon this would work ok? I don't really know if dampers work like I have imagined here for example... Once again, I appreciate any input on this, it's already been great for me to communicate with another enthusiast! Hopefully through doing all this I will in time gain useful experience which I can in turn pass on. Paul
|
|
|
Post by patamos on Nov 1, 2015 20:35:25 GMT -8
One of the important factors is the amount of insulted chimney stack you have. This provides much of the draw. Other than that, making sure the flue runs are smooth and with wider CSA in the turns.
Having a bypass/start up damper near the fire chamber is usually a good idea in my opinion. Beyond that one and another in the exit flue (to reduce air flow after the fire is out) i cannot say how your plan to use more dampers will work out...
I tend to cast fiberless cob around 26 gauge galvanized furnace ducting and usually custom build an extra long damper axle. The hole coming through to the surface is less than 1/4" and i have not had problems with gas escaping. That said, the more you are asking of the system the less negative pressure to pull gasses along...
|
|
|
Post by DCish on Nov 3, 2015 6:14:08 GMT -8
I'm wondering how you plan to access these passages for cleaning purposes. As elegant as the circular system is, it may be hard to clean, while long strait flue runs with a door at the end are quite easy to clean.
Although I love using bells, your design is low and wide, so a bell may not be applicable. That said, wider channels with longer dwell time are quite effective at heat harvest, so don't feel obligated to keep the channels narrow (though narrower is stronger from a construction perspective).
You may want to consider thicker layers of cob closer in and thinner farther out to avoid overly hot zones.
I'm with Patamos on the concern about trying to get two different passages to draw consistently, and would be much more likely to use a single channel.
Good luck, I look forward to seeing how it comes out!
|
|
|
Post by pyrapodpaul on Nov 8, 2015 15:31:30 GMT -8
Hello again, and thanks Pat and DCish for your replies which have led to a lot of thought, and the evolution of a new concept... I am now thinking of a design which is essentially a series of 4 bell-chambers. There is a drawing here - at the bottom of the page: pyrapod.wordpress.com/rocket-mass-heater/So, to explain this a little: Surrounding the stove & barrel are 4 v-shaped trenches arranged in a square. These would be lined with an appropriate cob-mix. Along the bottom of each trench is a gulley, or 'open flue', which is the essential route of the flue gases. They also all have an access-hatch at one end so the ash can be periodically raked out. [This would be below the wooden section of floor - easy enough to make a removable section] The top of the v-shaped 'banks' would be bridged across by thick slabs (paving slabs or similar - not shown in drawing), over which would be 3-4 inches of cob as before. This v-shaped space would be divided into 4 chambers by 4 brick walls, each with a hole at the bottom. [Presumably this hole would have to correspond to the system size (?) - so perhaps a 6-inch circular hole.] The gases would get to the 1st chamber via a short section of flue-pipe running downwards from the bottom of the barrel. Another short section of pipe would be fitted between the 1st and 4th chamber to house a damper to help with starting the burn. (thanks Pat for reassuring me this will be ok). With the burner going and the damper closed the gases should move around through the series of chambers and out to the exit-flue from the 4th chamber, and the octagon of cob on top of it all should get nice and warm! So question 1 is: Do you guys think this would work? Have I made any false assumptions, or silly errors...? And question 2: Can I cob directly onto the sub-soil, or do I need an insulation layer first? Otherwise, does anyone have any general comments on this? Cheers, Paul
|
|
|
Post by DCish on Nov 10, 2015 5:04:35 GMT -8
Nice! Some general comments:
1) Trenches: I'd be disinclined to require flue gasses to go any lower than the level of the fire itself. Once it's going strongly you'd probably be able to pull it off, but the start and possibly the end of the burn might be tricky to get flow going. You have a j-tube design, so burying it in the floor gets the fire low enough to run the flue through the floor, but I wouldn't want to go deeper than that.
2) V-shape: You should be fine with shallow, wide ducts. With that you won't get stratification, so you'll want to do something to ensure you don't get a stream of flow straight through the void. Maybe a brick on long edge 6" in from the inlet tube on each chamber. That would force gasses upward and to the sides to disperse better without complex channeling. You could leave it loose so that you can tune it as needed, and push it out of the way when cleaning.
3) Thickness of paving slab chamber tops: I don't remember the rate of travel of heat through cob, but for brick it is about an inch an hour. My two-brick thick bell starts to become perceptibly warm after about 5 hours of steady firing. Consider heat travel times and expected use of the space when calculating final thickness.
4) Heat tuning of the surfaces of the four chambers: The first chamber will be the hottest. If you want more even temps, maybe put 3/4" of a cob-heavy cob-perlite mix over the top of the first chamber before the final layer of cob, and reduce that layer by 1/4" for each successive chamber, having no perlite over the final chamber. Having the cob-perlite mix be cob-heavy means that it would be only partially insulating, thus you'd still get heat infusion, but at a slightly lower rate, compensating for the fact that the initial gasses are far hotter in the first chamber than in the last.
5) I know that Peterberg doesn't calculate the floor of a bell when calculating ISA because nominal heat is lost. However, I do notice that warming at the bottom of my bell does occur. A half-inch or inch of cob-perlite insulation on the floor of the entire installation (non-load-bearing portions only, I would think) both would ensure that the vast majority of the heat goes up, and isolate your warmed area from the intrusion of the cold from the earth below. Others may know better if this will truly make a difference. Good luck!
|
|
|
Post by patamos on Nov 11, 2015 11:21:38 GMT -8
I agree with DCish, If anything i'd be inclined to simplify the layout. Essentially you are creating a hybrid flue run/bell/hypocaust. The more you ask of it to harvest the heat the more potential for the heat distribution to be uneven under different firing conditions, existing mass temperatures and atmospheric conditions. So adjustability is of value. But so is simplicity
I think that many bells and gates and changes in shape could further exacerbate the variant behaviours in some surprising ways. So my gut feeling is to suggest an organic approach
If we think of a river and how rate of flow varies depending on altitude drop and valley shape. The faster the gasses move through an area the less time they have to drop heat. So I'd say create a single spiral with narrow flue at first. Then gradually widen as you get closer to the perimeter. The further you go downstream the more it looks like a hypocaust with pillars and perhaps a few adjustable gates/dampers adjustable from the top down near the end to evoke a bell effect. When building set some of the slabs and bricks loose at first. Then fire it up under a various conditions and notice where the warm and cool areas are and adjust to balance them out.
At the same time, there is something to be aid for uneven heat distribution. On cold days especially, you may want to sit closer to the hearth for a rapid recharge. Other times you may want only a little (e.g. sleeping). So variety is valuable. If a steady decrease in intensity as you near the perimeter is desirable, then a gradually narrowing spiral flue run would be good to maintain rate of flow…
a very interesting project you are undertaking. I have been thinking about a bell-hypocaust floor in my workshop for a while now, so this is fun to ponder with you all.
p
|
|
|
Post by pyrapodpaul on Nov 29, 2015 14:17:37 GMT -8
Hello again, and thanks DCish and Pat. I have taken on board your advice, but am struggling somewhat to come up with a design which ticks all the boxes... On the one hand I would like to go with the suggestion of a single, possibly gradually widening, spiral, and yet I want to be able to clean out all passages...tricky. I have drawn up various ideas to solve this, basically variations on my 4-bell-chamber design with narrower trenches/passages, in a sequence running twice or 3 times around a square, and each straight section having an access hatch at its end, most of which run under/through other passages. It's all got a bit complicated and so I think I will go with your advice Pat and just experiment with bricks and slabs and see what works... I will keep you posted on where this goes, it's about to get more real! Thanks again guys, Paul
|
|