RMH smoking back after small changes

Hello there all Permies!

18 months ago I built my first RMH. It was a deviation from the standard 8'' system. The heater and the mass are in different rooms (two rooms of 18sqm each, but there is a bathroom extension where the bell is). I was very unhappy about the performance so I made changes to the house (added more insulation) and also extended the heater's bench. The weather was warming up by the time the work was completed so I left it in the the testing state until October '18. I must add that I sat the barrel on strips of ceramic blanket (as I figured it might work as an air seal) which then was 'just' covered with cob. By January this cob cracked in many places and the bell itself was moved as some one leaned against it, not knowing it wasn't properly fixed.
Regardless of the mock up, the system work great and I could not be happier!! In 4 months I used only a quarter of previous year's firewood consumption and I had never been more comfortable! I have never seen another RMH in live action, but, even though the roar was not scary, it seemed to be working very well. I was so confident with it working correctly, that on a number of occasions I left it burning without supervision. It pulled great and I never had problems with smoke back or starting up cold (BTW, I also have a spinning cowl topping 4.5m chimney).
Earlier this month the weather warmed up slightly, so I decided to remove the barrel, paint it with stove paint and fit it back on, this time for good. Cleaned the barrel, painted with stove paint (rated for 350C, but my barrel never got hotter than that; I would boil water or make pizza on top of it, but the center would rarely, if ever, reach 250C), cleaned ashes from manifold, spread new layer of cob (this time i didn't want to sit in on the ceramic blanket) and replaced the barrel. As I may have to sell this house, I thought I would also replace cracked bridge brick before plastering the whole thing ready. I want to mention that the cracked brick was loose and there were visible gaps between bricks before I attempted any changes.
On firing for the first time after changes, smoke back appeared and I decided to do a better job with rebuilding the feed tunnel. I washed all bricks, dipped in clay and stacked them properly.
After a single firing, I was under impression that the problem was solved. But that was only one firing.
Before going down to plastering, I wanted to make sure everything is right and decided to remove the barrel again and raise it slightly (as it seemd to be lower compared to before). More cob was added and barrel was raised to the minimum of 3'', though I never measured the difference, I was too confident after 4 months of flawless operation.
There was smoke back on subsequent firings but I attributed that to moisture inside the cob. Stains on outside chinmey and water inside the duct seemed to prove the theory.
After a couple of days, I touched the inside of manifold and everything seemed bone dry (no more moisture in the duct too), however, heater would not draw as before. Moreover, about 60-90min into firing the roar would stop, top bricks of feed chamber would become hot (I never remembered them hot before the changes, but I also never paid attention to it... hovewer, I would surely remember burning my fingers touching them, whereas I would always stick my hand right into the feed tunnel to move sticks forward) and smoke back would start. Soon before it happened I would always notice bell radiating almost too much heat for what I was used to.
I have tried many sollutions. I used refractory paste to seal any possible gaps between bricks in the feed tunnel. I started plastering over the manifold to close any possible air leaks (I noticed that a crack which was present when the system was working well, widened considerably now; I opened it, sealed with the refractory paste the plastered over it).
Although I still would not get a draft as nice as previously, I thought it might be because of the warm weather outside (it goes to -1C in the morning but we had an 22C during the day today).
Two firings later, I had a nice start, as the inside was still warm from a tiny firing earlier the same day, but after ca 90 minutes the system started to smoke back and I heard an audible thud: the crack in manifold opened again.
Also, when this happenes (draw slowing down and heat building up inside the feed chamber), I notice temperature drop in the chimney.
I've followed all leads mentioned in Erica's and Ernie's book. Closing the crack seemed to have somehow solved the issue, but then, again, something seemed to have blocked the draw and heat or smoke building up inside manifold caused the crack to open. The barrel I used is wider towards the top, so I could not have used another one for manifold, therefore it was built using clay hollow bricks and earten mortar.
Would there still be moisture in the cob that was used to set the barrel? That was about 15L of cob in total and it has been about 10 days since it was done. Since then, I only added two layers of gipsum, but I doubt it steams into the bell... Chinmey is not blocked  but the cowl seems to be less effective than before the changes.
Although the build was not standard, heater's dimensions where all according to the book. Bell is slightly offset, but so it was before the changes.
The only two things I altered in the meantime: I sealed front door that was drafting and I plastered stones and cob around the duct that leads to the second room. It didn't add any insulation or mass around the bell. It's a thin layer of gipsum over cob and stones. The heater does not sit in basement. The room it is in is 1m lower that the one heat exchanger runs though.
I do not mention other specifications as they seem irrelevant, if the heater worked just fine, before I painted the barrel.
I cleaned ashes from the manifold, but not from the duct. I peeked in and there isn't a lot of them though... Would that be a problem? Easiest way would be to clean them up, but I don't have an ash vacuum cleaner at the moment.
I can write all about the build, if need be.
What is noticable is that the bench does not warm up as fast (and as well) as it used to, even when the chinmey remains hot...
The only (or best) way to warm up the bench is to push all wood, coal and ambers forward, when smoke back starts, and close opening with fire bricks. Then, slowly, the bench starts to warm up. but most of the heat stays in the other room, where the bell is.
I wish I had paid attention to how it all worked when all was fine... I could have some data to compare. However, all was well and I never have any spare time, so I did not do it.

I don't think there is anything unusual in my operation. I do it the same way I used to. Only the heater responds differently. The bell heats up way faster. I used to need about 90min to start baking pizza. Now the 'oven' is ready within 45min. It is way warmer around the bell now.

The wood used are mostly prunings. Perhaps they have not been properly seasoned, but we are in semi-arid region and they have not had any rain on them in at least 8 months. They are only 1-2'' in diameter but I tend to use sticks longer than the feed tunnel. But I used to use them even longer (up to 3 feet) before the changes and I have never ever had any problems.

Any help will be highly appreciated.
Kind regards
J

Now then

Once the heater cooled down, I realized the crack disappeared and thus suggests expansion joint problem (but would that cause no more heat going to the heat exchange and chimney? the chimney was way too cold just 2-3h after the fire was put out).
Straps of ceramic blanket used in the prototyping was meant for the reasons of metal expanding whil hot.
'The RMH builder's guide' only mentions masonry manifold and, instead, focuses on building metal one. However, it is suggested that expansion joint is created by heating up the bell. Wouldn't this expansion joint be a possible air leak? It is also said that wraping refractory insulation around th metal bell before embeding it in cob was not as effective... this puzzles me a bit, as it seems that air leak down there is causing my headaches, because it ruined my otherwise great heater

What I am going to do now is to remove 1/8'' or so of cob around the bell and I will use ceramic blanket and stove gasket instead.

In the meantime, I hope to hear from more experienced users.

I just remembered something.
In the picture above, you can see the first meter of duct rising right by the bell.
for 17 months it was only covered with a ceramic tile and wasn't air tight at all. At some point the tile even broke in half and for a couple of days I didn't even realize it. Inspite of 1/8'' gap created, I didn't notice any change in heater's performance. Bench was warming up as expected, no smoke back was detected, etc. I understand that it is already outside the manifold. My point is that I was so happy with the heater, regardless of all prototyping mock ups.

I've chiseled out 1'' of cob around the barrel, all the way to its rim. Closed the space with ceramic blanket and fired stove. Same thing: 45min in, temp in top of barrel reaches 200C, chimney cools down to <50C (the lower part) and fire starts climbing up the sticks and into the room.

Is there a reason why the gap, which I mentioned disappeared, would be visible again? It is masonry manifold and this time there was no metal movement involved... The only contact points bell had with the manifold were what the rim stood on... this couldn't have caused the crack to widen again...
I am really lost.
I read other topics and books. I understand that in most cases manilofd leaks are pointed the culprits. But how likely is it that a leak developed in the same time i was amending the bell?

Could any one suggest how to go about plastering inside the manifold, please? I imagine, I would need to scrub the bricks and cob clean? Perhaps I never had a good clay to start with, but I feel that cob (or I should say "my mix") has failed me too many times, so I intend to use refractory mortar to plaster the inside and then gipsum to plaster again the outside.
Funny thing is that I inspected top of brick behind the crack over the manifold cleanup and I didn't seem to see it broken or cracked... And yet, this new crack that appeared just yesterday, even though not lined up with the first, was pretty much along the same path...
I would very much appreciate help of someone who has attempted such fix.

I have a feeling (though it may be wrong, as I am looking for faults all over the place) that the bell gets warmer than it used to. I even considered that the radiant heat warms up the feed tunel as well as the air that goes into it. One time I placed a divider to block the heat. It seemed to have worked but it certainly did not work today. As far as offsetting goes, the side of the bell facing firewood fed touches top of the heat riser, just like before the changes. I had to offset it, owing to its slightly conical shape.

Attached is another photo, taken just before removing the bell.
When zooming in, the rim can almost be seen. Cracks in cob are all over the place. As said before, the barrel sat on strips of ca 1'' of ceramic blanket. Even outline of the tiny crack above the cleanup openning can be seen (it is not really seen, but when one knows where to look, a slightly darker line can be seen).
Despite all this 'imperfections' I run this heater with great success and was looking forward to finishing it off...

Another hint that may help pinning the problem for experienced builders:
I have never seen another RMH, but last year, before rebuilding it and extending heat exchanger, I was not happy about smoke and flames swirling and dancing towards the back of feed tunnel, further away from combustion chamber. Though, last year I did not have problems with smoke back, I decided to rebuild it. Can't remember exactly what I did, but I think I rebuilt all brick except for the first layer (pad) and insulated better. Having done that, it run superb (in my opinion, not by comparison).
Now, I rebuilt part of the feed tunnel to replace first bridge brick, then painted and reembeded bell... no alteration to insulation of heat riser whatsoever.

Warm weather outside would not be to blame, would it? Thing is, chimney did get hot today, then cooled down and once temperature of the bell dropped below 100C, chimney started to warm up slightly.
Spinning cowl did not help to maintain draft. It's a fact that my chimney is 3m away from the bell, but it was there during the happy days

What did you make your riser out of?

I left  a metal sleeve in the center of my cast riser and had to pull out chunks through the feed tube as they burned loose and fell down. My first riser was an emergency hurry up double walled ss stove pipe and when I replaced with cast ceramic and perlite, it looked fine from the outside, but the inside metal was ready to fall out.  The stainless steel was also not as good a riser (even though it worked), and when I replaced with the ceramic perlite, I started to get the glowing barrel at times.

I have had to handle the barrel on my own rocket a few times when i changed the risers out, and when I substituted a batch burner for the J tube, and it is a bit tricky sometimes to avoid hitting the riser.  But the point is it sounds like something blocking the riser, possibly you have bumped it during reassembly ( i assume you checked it when you took off the barrel.)

Inspect the part of the tunnel right under the riser, look for anything that might have fallen down, possibly protruding just a bit downward. You may be able to reach in and pull bits out, or if you don't see anything, a good stiff tape measure could  be used to stick up and make sure the riser is perfectly clear.

Hi John;
I moved your post from wood stoves to rocket mass heaters, hopefully we can help you get your rocket roaring again.
I will say, wet cob takes a long while to fully dry. Also atmospheric pressure changes can make your system stall.
Would it be easy for you to try adding an extra pipe on top of your chimney to make it taller? Is your chimney made with insulated pipe where it is outside ? We currently have another poster, their stove works great until an atmospheric change then it will not heat up... this could be your problem , also warmer temperatures can affect draw as well.

I think to start We need your dimentions , length of pipes thru the mass , number of bends in the system. The acual size of your transition manifold.
You mentioned some ash in your piping and not having a vacume avalable to clean it.If you were to remove your spinner cap on top of the chimney , do you have a way to "blow" any ash up and outside ?
Sealing your barrel with ceramic blanket then covering with cob is exactly how my barrel is mounted, that is not your problem.
What materials did you use to build ?  Clay brick ? fire brick, insulated brick ?  What size and material is your mass piping made with?  Smooth pipe inside ? Or is it the "ribbed" pipe ?

Hi, guys, and thank you very much for your input.

My heat riser is made of refractory bricks, like the rest of the system (combustion + feed) and wrapped in 40mm ceramic blanket.
Dimensions are exact copies from the 8'' RMH in the Wisner's book. Ratio 1:2:4, as I can remember correctly. Manifold is a bit larger than standard, as I read it should be minimun 1.5x the system's CSA.
Heat exchanger is as follows (although it would not matter much to pin the problem, as this has not been altered in any way):
90deg
3' vertical
90deg
2' horizontal
90deg
7' horizontal
90deg
5' horizontal
180deg
7' horizontal
90deg
15' vertical
I know this is over the recommended 50', but that's why I installed the spinner, to help with draft. Before extending by adding 10' of horizontal run plus 180 deg, bench was way too hot and room not warm enough. The addition has worked like a charm.

The weather has warmed up a little, compared to the firings before changes, but would that be such a coincidance? Besides, why changes in chimney temperature, once heat riser was cooling down?
Bob, when I removed the bell (though I did it twice and alone, I am sure I did not touch the riser, I did so last year and knew not to make the same mistake), there was some sort of covering on the refractory bricks inside the riser. Almost like thin plaster... I peeled off a little to see what it was. Perhaps a piece is hanging off, slightly blocking the draft... It seems I will look inside tomorrow, once I have the refractory mortar to plaster the manifold...

Thomas, I understand cob takes time to dry. But still, there was only 1'' of it below barrel's rim plus 3-4'' 'band' embedding the bell. There cannot be so much moisture in that little amount, can there? I know it was an issue to start with, as can be seen on stained wall around the chimney, but there is no more water in the duct. BTW, last 7'' of chimney is double wall and it outgrows the roof by over 2'. It could have been better, but was not an issue for over 4 months, so would be too much of a coincidence if it was to blame...
The duct is made of smooth galvanized and vitrified (the first two) pipes. I think, to be sure, I will sacrifice my temperamental vacuum cleaner to see if removing ash from the duct will halp. I mean, this ash was there 3 weeks ago, but since I cleaned the manifold (and there was a fair amount, 2-3''), there may be too much air getting in the manifold but not able to puch on. However, there may be 1/2'' of ash inside the duct. That's in the first 3'. The further up the system, the less ash there is. It will be done tomorrow. Unless the hoover stops working half way through the cleaning
Since I will plaster the manifold, do you guys have any recommendations? Should the mix be dry or wet? How much to scrub  the bricks/cob? How long does it dry before firing?

Just thinking now...
I am sure I didn't bump the riser when lowering the barrel.
But since it almost touches the riser, maybe, somehow I bumped it while leveling the bell.
What would be the drill to diagnose problems? I mean, if no movement can be seen on joints...
I remember touching last course's brick and it was loose. But I guess they all become loose as they were only glued using clay slip, just like in the book.
Frustrating thing is that I have plenty to do but can't stop thinking about the rocket.

the plot thickens, if it is that close to the bell you have likely reduced the clearance too much  

and 8"system needs about 50 sq inches that's a minimum of 2 inch clearance between riser and top of bell, just to be safe I would aim for 3"  

The extra length over 50' and the warmer temps could be your issue.  

Can you easily try removing your spinner and adding a pipe or 2 , just to try ? Leave the spinner off for testing . Pressure changes and temp can play havoc your draw. I assume it was colder outside when it was working well ?

Bob, I said above that mine is ca 3''. I think it over that. I remember measuring this last year. This time the barrel is even higher, as I wanted to try if this would still improve draft...

It is the bell's side that may be touching riser. It had to be this way to avoid bottom of the bell (which is the barrel's top) being too close to the feed.
Ernie mentioned this in a video, I guess. So that the bell doesn't become chimney (or the other way round, I suppose: the feed tunnel doesn't become chimney)

Oh, and I just remembered, according to the pro boards at Donkey's forum,manifolds and transition points where a csa of gas has to pass around an obstruction or at right angles in  narrow spaces, you need to add extra over and above the exact amount An extra course of brick around the bottom of the bell will probably do it without too much hassle.  (but also check for obstructions in the riser)

Thomas, I am afraid I don't have more pipes to test.

Should the length be an issue, why would it work so well for over 4 months and stop just now, when I changed something? I don't know...

That's a fact, it was way cooler 2 weeks ago. I am hoping this is the issue... I think, for the first time in my life, I crave for winter to come back! hahaha

Think about the gas as a liquid flowing down, --now imagine a pipe of a certain diameter just above the bottom of the bucket, and imagine the turbulence that would occur as you move the bottom of the pipe close to the sides of the bucket

Now if it has always been that way and you added height to the bell (didn't you say you removed a CF seal?) then forget I said anything.

bob day wrote:Oh, and I just remembered, according to the pro boards at Donkey's forum,manifolds and transition points where a csa of gas has to pass around an obstruction or at right angles in  narrow spaces, you need to add extra over and above the exact amount An extra course of brick around the bottom of the bell will probably do it without too much hassle.  (but also check for obstructions in the riser)

Barrel sits about 1/2-1'' higher than when it worked fine. I remember that lifting it last year improved the draw. That's why I set it even higher this time. But another course of bricks?

Tell you what, I can't wait until tomorrow.
I am removing the barrel now, unless it is still too warm.

The other thing that occurred to me concerned your seal of the front door.  If your house is pretty air tight you may be creating a partial vacuum inside pulling back on the draft.

simple enough to test, when it starts smoking back, open the door

bob day wrote:Think about the gas as a liquid flowing down, --now imagine a pipe of a certain diameter just above the bottom of the bucket, and imagine the turbulence that would occur as you move the bottom of the pipe close to the sides of the bucket

Now if it has always been that way and you added height to the bell (didn't you say you removed a CF seal?) then forget I said anything.

Yes, it has always been offset like that. I don't want to forget anything you say :-) I want to learn :-)
The truth is, it was so close to the riser that ash built up to 1/3 of riser's hight... When reassembled, I though that ash might have been useful there, as it blocked some heat radiating towards the feed... I tought that was to blame. But it is very unlikely. I always needed minimum of 60-90 min to reach temperatures for baking. Now it is all done in 30 min.
I think your first suggestion might be right. something in the riser...

Unfortunately it is still too hot to remove :-)

bob day wrote:The other thing that occurred to me concerned your seal of the front door.  If your house is pretty air tight you may be creating a partial vacuum inside pulling back on the draft. .

I mentioned sealing the door, as I myself thought it could be it. I would not say the house is very air tight. But I did open the door on two occasions. I didn't pay too much attention if it helped or didn't.

I'd like to believe I am quite an observer :-) When something is out of ordinary, I tend to notice it. And I can't help thinking that there is too much heat radiating of the bell. I used to be able to sit there for 20min, feeding tiny branches to fill up space in the feed tube. I did sweat over the past day doing so...
As mentioned, it is warmer, inside and out. But something blocking riser is most appealing to me :-)

I need to go for a jog or to do sth productive. And in 2-3h I will look dragon in the mouth.

Once again, I really appreciate help in solving the riddle!

I hope you figure it out

I hope so too!
And I hope tomorrow is going to be a better day, otherwise I want to give it all up! One thing I have learned through this whole project (I don't mean just the heater, the house and many other things) is that caring about the planet costs me my sanity I don't know any more what is more important hahaha so I wonder: have I lost it already?

Bob, there was nothing blocking riser.
I've found some cracks in earthen plaster inside manifold. But I couldn't tell if those would cause any leaks. I started scraping it off and tomorrow I will plaster again with refractory mortar. One thing I need to take care of will be to water down all bricks. Cob I used to embed the bell did not stick to rest of the manifold very well. Came off in large chunks in fact. That might have created so problems...

Ah!
I almost forgot.
I did measure the distance up on top of the riser. Turned out that I lied to you. It was 2.5''. Which is weird, as i must have been 2'', then, all that time it was functioning very well.
Tomorrow, I am going to lay another course of bricks and will once again set the bell on pieces of ceramic blanket
I don't care if it is unstable. I'd rather have a functioning stove!

I will report back my findings. Though with days of 20C (and 8 at night) it may be hard to have a reliable result.
In the meantime, I'd appreciate any advice on plastering with refractory mortar, as this is going to be my first time.

I resist refractory cement as much as possible, and stick with the clay I dig around here. The advantage (as I see it) is that if there are surface cracks, I wet a brush, soak the area to repair and add a little sifted clay that I brush into the cracks. sometimes it takes a few applications, sometimes it  never really ends, and I just make it routine to seal everything periodically. and when I change the design I throw it all in a bucket of water and start over!!!

Perhaps when i really settle on a final design I will look harder at refractory cement. I bought two 25# bags but at 60$+ per each I hate to open the bag.

But here's how I would do it   wire brush all surfaces , and dust them off good. wet the surface with water, apply the cement.

There is also a new technique/ materials out there called geo polymers, the ingredients are cheaper than refractory, at least as temperature resistant and quite insulative when applied in layers.  It is definitely interesting stuff

http://donkey32.proboards.com/thread/2372/simple-straight-away-ltgs-binder  If you're going to be playing with rocket stoves this is a great site to know about

This is what I'm working on   my build

Have disappeared for a while
Been quite busy with the heater and other jobs.
Bob, that's a very interesting project you have there. I wish I wasn't so tired of reinventing the wheel hahaha
I feel like I need a break... It's one of reasons for being so impateint with my rocket. Been burning too many candles at the same time. And on both end
Refractory materials are pricey. And they cost up to 50% more in my area (southern Europe). I mean, all is more expensive here. I'm 80 miles away from any important builder's chain and local merchants go crazy with prices.
Nevertheless, I got only 5kg of mortar and that didn't break my bank. Was just enough to render manifold. Of course, I did a scrach coat with regular cement.
I restacked part of feed tube.
Now just to vacuum duct tubes and ready to launch.

Funny, now that I Have fallen out with cob (or, I should say, my mixes or my subsoil), I came across this thread about feed tube warming up. Which is exactly what happened in my case, but only after changing heater. Perhaps there was nothing wrong at the end... just pressure issues, like Thomas suggested.
I am too restless and it all seemed too coincidental

Can't wait to fire the thing, hahaha
But prefer to wait for mortar to set a little.
I used ceramic blanket all around, topped with stove gasket. Seems this time I don't need to fire for expansion joint

I take it you haven't had a chance to do the open door test. and also that there was no problem with the riser

Nope
Never tested in outdoors.
When this whole project started 4 years ago, my planning was too ambitious, hahaha. I'm still paying the price for that mistake. Never have time for anything.

It's about 7-8C (45F) in the morning.
That's the coldest it gets. And it looks like winter is not coming back. Tomorrow morning I will test it. Some wind is forcast which will certainly help. Fingers crossed.
I've tested draw with incense sticks. Even with no fire and cool mass, cool combustion and wet manifold, smoke was sucked in. I don't know how it would have worked before. I need to start testing things even when they are at their prime

I meant opening the door you sealed if it starts smoking again to see if there is a back pressure on it

Sorry
hehe

The riser was fine. Unfortunately, lol. As it would have been much easier
I really hope it was air leak in manifold. render cracked all the way down to bricks. Brick was intact, but there were gaps where cement mortar touched cob. Except for the lone crack that split render in half, cement render seemed to be fine. That's why I chose not to use 'my cob' again. Initially I thought it was good, this clay and mixes. Later on I realised it wasn't so much so...

I forgot to add that I did add another course of bricks under bell and clearance over heat riser is close to 4'' now... I did take your advice, Bob. Hope it won't be too much.
With close to 20C outside, I don't wan't to stake my motivation and will wait for more favourable conditions first thing in the morning.

good luck

It wasn't a full scale test

There was a nice draw. I could tell, because flames were swept from far side of feed tube too. No swirling or charcoal there.
Most importantly, chimney responce was correct. It was too hot to touch, while bell didn't get as hot as over the past days. And I wouldn't say that room temperature made any difference (of course, when it is cooler, bell cools down faster; however, I had 18C when starting fire this morning, whereas it would be 20C before most previous, unfortunate attempts).
After 60 min, temp in the room where heat exchange is reached 25C. There is no radiant heat traveling there, but heat travels there easily, as it is 4' higher up.
I don't know what was bench's contribution to that. After 4 days of no use, while heater was open, it cooled down completely.
Inspite of that, I could tell there was something right going on like it is warming up, as opposite to most recent firings.
Top bricks in feed tube stayed even cooler than my hand throughout burn, but did get too hot when covered to reduce air flow. I wish I had observed how it behaved during the happy days
All seemed okay.
Opening front door did not show visible change to draft.
Adding another brick under barrel has a positive effect (at least theoretical). The hotter part of bell is further away from feed. Insulation added there also helps in this respect. I only hope it will not affect downward movement of gases.

One thing happened though.
A visible crack appeared very close to where the previous one was. I don't know if I packed mortar too tightly and 'expansion seal' was squeezed too much or there is a weak point in manifold. This crack did not seem to affect rocket. Feed temperature did rise, as stated above (and around the same time), but there was no indication of smoke back  (like draw stopping, bell heating and chimney cooling)

I did not vacuum ducts at the end. So this was not to blame, because draw seemed improved.
I am not very pleased, because I was unable to pinpoint what went wrong...
I have a feeling it was the manifold, as is suggested in the Builder's guide. Too much steam from wet cob might have exposed weak spots in manifold, due to higher pressure... I really don't know.

Should any one have same issues, my symptoms were as follows: chimney not hot enough, bell too hot (even if thermometer didn't show very high temp, one could tell radiant heat was unbearable) and, in turn, bench not responding correctly. In other words, it could feel like there is heat trapped inside bell (that's why Bob's suggestions about an obstacle in heat riser was appealing to me).

Hope there will be no need to come back to this thread at least not for another 7 months... March is starting warmer than unusually with 25C. If any one can suggest a Rocket Mass Cooler, I'd be happy to experiment, hahaha

Hopefully you'll get a chance to run a full burn before winter is truly gone just to make sure everything is ready for next fall.

Actually John,  Your rocket mass heater is a rocket mass cooler as long as it is not in the sun.  

Even though you don't know what you fixed … at least it seems to be working properly.  Maybe there will be a cold rainy day soon and you can fire it up again before next winter.

bob day wrote:The other thing that occurred to me concerned your seal of the front door.  If your house is pretty air tight you may be creating a partial vacuum inside pulling back on the draft.

This is a very important and often overlooked point. Any building that has an exhausting appliance works on a vacuum system. There are some locations that require
constant building vacuum systems for things like radon. Your mass heater has to pull against that draw to perform. Your building may have several air drawing items working at once. Consider if you have things like a gas hot water tank, gas furnace, dryer running, attic fan running, stove hood, bathroom fan or another wood stove or fireplace. All of these things will draw air out of a building and for every cubic foot of air that goes out the chimney, a cubic foot is drawn in to the building from somewhere. Your heater WILL have to deal with this.
Brad  

Brad, thanks for the thought.
Well, there is absolutely nothing of that sort It is a very very simple (but comfortable) house. Electricity supplied by 175W of sollar array (which in practice gives only ca 11Amp at 12V). No fans, no driers.
If anything, it might have been my eagerness blocking air flow, as I was standing there, almost convincing rocket to give up the pranks
To test draw, I place my palm over feed tunnel. When all works as expected, I feel cool draft. When I start feeling rising heat, it means something goes wrong. On all occasions, heater would start normaly and then refuse to continue.
That's why, even though I don't claim to know it for sure, It seemed that outside factors could be ignored and something was wrong inside...

However, it is a good point. One thing I can do next time is to close door between two rooms. I never close it, and since the second room is higher up by 3-4 feet, it may steal some draft, but air has little way out from there, unless door or window are open... I have an opening left for evaporative aircon, but it is tapped in winter.

Funny thing, stuff always comes to me in pairs. Be it trouble or luck. Yesterday I got job offers (one during an on-the-spot interview, second after weeks of waiting).
I am almost certain that same thing can be applied to my rocket. I was fixing one and another broke.

Sure thing, if I get to test heater again, I will report my findings.

More tests and more disappointment

When fired at the end of Feb, it really did work well. There was some wind, but still...
Now that the new render on manifold has dried, crack appears 60 mins into firing. Funny because there was a hairline crack before taking half of the manifold apart. Though it never moved. The current one closes when system cools.
However, I am not too sure anymore if this is the problem, because, at least, there should be strong draw before the gap opens.

Since I have new possible symptoms, I wonder how likely it is that the issues lies in heat riser, even though I don't recall bumping it.
The draw certainly is an issue here. There is a pronounced charcoal build-up. I am thinking if the former causes the latter or the other way round.
Second thing I noticed today is smoke. It is not so bad, but I don't think I have seen this much smoke in 18 months. There would be some, but no longer than 20-30 min into firing. Today it was there all the time.
Combined with other indicators : hotter than before barrel and cooler than normally chimney they may lead to some explanation... But I am clueless.
On first firing after painting, the paint gave off a lot of smoke. I wonder how possible it is that it continues to do so inside...

During the test on 28 Feb, even though the system was cold, bench responded nicely.
Today, even though it was primed 24h ago, it refuses to warm up. I mean, of course it takes time. But I know what it used to do...

It is a pitty that I will probably never find out what was the problem, for the sake of other builders.
Seems it will be way less frustrating and time consuming to rebuild the engine. This time using custom bell and manifold.

EDIT:
In another thread I also mentioned that during last two firings, the draw was breaking up, sounding almost like steam train, rather than rocket.
Does this hint anything to anyone?
Anyone had same issue?

I am reading posts on donkey32 forums and wondering if the pulsation of draw is not caused by the enlarged gap over heat riser.
However, there isn't a stronger draw than before, to start with. So, even if more gases are allowed there, I doubt they do take advantage of that, if you know what I mean...

John Goodbody wrote:In another thread I also mentioned that during last two firings, the draw was breaking up, sounding almost like steam train, rather than rocket.
Does this hint anything to anyone?
Anyone had same issue?

A slow pulse, like a steam engine, 5 times a second or so points to a restriction somewhere in the system. The expanding gases would slow down, accelerate again, slow down, et cetera.
Most causes has been mentioned here I'd think, but I'll post those again, it might be helpful in some way.
Construction faults: Top gap between riser and barrel, manifold not enough space, too many bends, cramped transition to chimney, cold outside chimney.
Places where the hot gases need to change direction there should be more space. A 90º bend 150% of system csa, 180º bend 200%, so the space above the riser should be twice as spacious as the riser's csa suggests, in my own opinion. The same goes for manifold and bends further down the system. Sometimes there's a build mistake that causes a large  piece of cob fall down later on and block the system partially.
All these spots can act as a restriction although it doesn't look like that.

Restrictions that occur through time: ashes buildup in the tunnel/riser transition, on top of riser, in manifold, in elbows in a piped bench system. Straight single bell systems rarely clog up, floor area of most designs is large enough to house vast amounts of ashes. One of my designs showed signs of restriction down the line after ten years of use, cleaning the bell's floor fixed that issue.

Hi, Peter, and thanks for your input.

As described earlier, the system worked perfectly prior to painting and replacing barrel as well as replacing one bridge brick (to do that I had to remove and replace again a few more fire bricks).
All winter it worked like a charm. I didn't have to watch it. Easy to fire, no picky about amount of air even with empty feed.
As said, bell wasn't even embeded properly (sat on a layer of ceramic blanket which was then cobbed over with 1'' of earth).

In fact, I may have a transition issue, since my manifold exit is 180deg, rather than usual 90. But this was NOT changed and had worked great.

The only things altered in the past weeks should have improved draw (gap over riser increased to 4'', properly embeded bell, replastered manifold - I removed the old cob and replastered with mortar so the size should not be different), meanwhile, every next change seems to cause more problems....
Actually, when fired 2 weeks ago, system seemed to be working fine. This morning, before firing (and it was primed, as mentioned), I vacuumed ducts. Manifold was cleaned when replacing barrel...
Moreover, I opened front door to let in more air.

Because I have done all suggested and that can come to my mind, I feel hopeless because I am running out of options.
Reading donkey32 this morning, I figured a restriction somewhere. But people usually mention pulsating in the beginning. Mine starts when system reaches max temps...

If the flow is restricted somewhere, it must be in the manifold. Though all worked fine up to Feb, I know I didn't do great job there. It is not very round and can see why using metal or cast iron will make more sense.

In any case, will RMH behave like this if there are problems with heat risers or leaky manifolds?
Before raising barrel to 4'' clearance, it never pulsated like this. It would draw sort of okay and then stall once hot. Now this pulsating is added.

Hi John, sorry to hear you're still having problems,  I don't have anything to add, but will continue to point (with others) at obstructions from the manifold onward.   Peter mentioned the idea that your csa formula needs to increase wherever the gas turns a corner.

I'm going to say most likely, that if all other elements are truly free, increasing the space at the top of the riser would not be a problem.  If there is a restriction further on it might allow a larger flow of gas , and that larger flow/ push could then meet the partial obstruction and oscillate as Peter mentioned.  

I doubt the crack you see is doing anything other than being a crack, at that point in the system the most likely thing would be a little smoke escaping, If the crack is an indication of an internal failure and something collapsed at the manifold after it was sealed, that would be another story

It's a little difficult to picture the full exhaust route through your extended benches and such, perhaps a diagram of the overall path (with dimensions) of the exhaust would help us understand better.

Hello again!

As mentioned in previous posts, bell is offset for two reasons:
- to be as far away as possible from feed
- to accomodate as much room as possible for the 180deg transition

As can be seen in the clip below, parts of bell were not even embeded at all. I'd be surprised if it was air tight. That's why I became (perhaps too much) very confident about the design and stopped being precise, etc.

I start to think if too much moisture could cause some damage to other parts of the build (duct or chimney), since the very first firing after embeding painted bell produced lots of moisture that spattered chimney and outside wall. Started wondering if that could affect the operation of chimney's spinner but it seems to be working fine and it is not clogged with anything.
Vacuuming ducts was my last hope. Did so today but problem persists.

Hey John,

It sounded like you were saying the exhaust was smoking as if there was incomplete combustion as well

It's a little difficult to hear a pulse, and the rockety effect sounded real good, how would you say it was performing when you shot the video?

was it smoking,  smoking back?  with all the symptoms (cool exhaust, etc?)

Looking at your diagram, (wildly out of scale :-))  the 180 you mentioned is not going around a corner on the level, rather it is going straight down from the riser and immediately straight back up, and almost as quickly turning a 90degree on the horizontal.

If you're sure the rest of the run and up the chimney is perfectly clear, and you have a nice big manifold opening, I would suspect the issue is with those close together 90 degree angles

Looking at the top picture in the post it looks like the vertical piece next to the bell has a cleanout T, rather than a smooth 90 degree radius turn.

You might try and replace that t with a simple 90 and see if that smooths out the exit path for the exhaust.

I know that answer seems very far fetched, but it's all I can think of short of extreme renovation, and mostly it would be the section right there where you have so many twists and turns in such a confined space.

One answer might be to take the exhaust straight out the top of the bell eliminating most of the turns. this will reduce the heat in the bell and increase the heat in the bench

The 180 is manifold exit into 90deg elbow rather than a strainght piece.
3' vertical tube
Then the T.

Well, the T was meant for the bottom, to be used as cleanout too. Due to space restrictions, cleanout had to go in the front and thus I used an elbow there instead. Hence the T where it is seen. To tell the truth, for most of winter, this T was only covered with a piece of loose tile. One day this tile cracked in half without me noticing it for days, because it didn't cause any issues.
Clip in the previous post is from before, when draw was strong. I posted it to show that number of turns as well as that T didn't stop it from working correctly.

Yes, that's correct. Today the exhaust was smoking, chimney again became just warm and then the pulsating started. But I didn't make a video of that. It looks just like in this one, though not every time as bad. Only happens once hot


As for charcoal:
Before, the bricks would warm up on putting fire out, with air flow restricted. Only like 2-3h into firing. Now, the brick that covers it gets too hot to be touched. I understad draw was stronger and cool air pulled in not only cooled feed tube but at the same time made sure combustion was complete.
I just want to figure out why it is not drawing as it used to, because this causes the rest (charcoal smoulders, warms up feed tube and fire smokes back)

Bob
Do you think that air leak somewhere in the duct or chimney could slow the draft? I mean, the syphon is important in J-tube, but further down?
The only other explanation I may have is that I plastered stones around the vertical piece that runs parallel to the bell, before it becomes T seen in the first photo (first post). I mean, it doesn't touch bell or anything. Just may reflect back the heat, impeding cooling of the bell (and this side receives most gases travelling down to manifold). I know it may sound crazy, but what else have I got?

Your last suggestion is to eliminate the manifold? Well, at least how it is now...? And exit into the other room directly from heat riser? Is that what you are saying?

All of the last posts in this thread imply a restriction.
And it makes most sense to me too.
But I can't figure out where it may be!
Ashes were cleaned out. Heat riser checked and clean. Nothing in chimney. Spinner turning.
Because of all these, I figured it must have been due to manifold leaks, hence plastered again and rebuilt part or manifold...

I think we are running out of ideas. Which is pitty actually.
I will have to go away for a while, then come back with fresh energy and rebuild the thing. The engine, I mean. Not the mass.