|
Post by invention1 on Sept 16, 2018 17:58:32 GMT -8
Well I've done an experimental burn, with some mixed results. First, I chopped off the P-Channel per Peter's suggestions. Secondary air is injected just below the top of the port. Then I made a run with pellets as fuel. 5-7 LBS (estimated, not measured) hardwood pellets in pellet basket, lit with a splash of fuel alcohol and a propane torch to get things moving. Very few pellets dripped through the basket, although the spacing in this version is more like 1/4" instead of the recommended 3/16". Smoke from the riser was absolutely clear from the beginning of the run till about 30 minutes into the burn. I don't have a Testo, but I've got airflow meters and temperature gauges! So that's what I measured. I can only estimate the quality of the burn by observing smoke and ram's horn pattern. Once it developed the ram';s horn pattern of flame was present and well developed. For the first ten minutes I could not observe it or take a pic of it, the stuff coming out of the riser was just too hot. In the middle of this run measured air consumption a the primary air inlet as 8.6 CFM/.0041 Cubic Meter/Sec. For comparison, my old 8" Fisher woodstove uses about 42 CFM, on a strong chimney. THis primary airflow is suspiciously low - I probably have some leaks that need to be addressed. Temperature in the middle of the top end of the riser was 603°C. Also turned the stainless steel end of my type K thermocouple probe blue. Type K is good to over 1200C, and this is a lab quality gauge so that's probably pretty accurate. At about 20 minutes into the burn, the ram's horn pattern became unstable - flickering on and off, you could hear a pulsation from inside the stove as things ignited and then went out. at about 30 minutes into the burn flames no longer reach the riser From 30 minutes to about 1 hour there was some white smoke, no ram's horn, pellets coaling, at most 6" flames at the pellets. At 2 hours it was essentially out. I was able to get all the ash to fall just by lifting the pellet basket and setting it back down. Easy to clean. There were no broken ceramics. But there was a key problem: ZERO airflow into the secondary air inlet. No, it's not blocked. I observed this several times, both with pellets and later with a cordwood run. No airflow into the secondary air inlet. Now about this airflow meter: It is probably older than I am, they were throwing it away at a lab I worked at, so I took it home. It is so precise, it will measure the airflow of a *cough*. Yes, a cough. It measured .06 CFM. I'll describe the cordwood burn in the next post, but let's focus on the secondary air for a bit. I'm pretty sure I've got the overhead P-Channel placed too far from the port. There should be a strong low pressure right at the port, which should pull in airflow here. But airflow into the secondary air inlet was nil. I left the airflow meter taped to the secondary inlet for quite a while to see if it would ever change, but it did not. Every so often the vanes would move a bit because of a passing breeze, although I had them shielded with a paper tube. But if there was any airflow above or equal to a puff of air, this meter would have seen it. Here are some dimensions and ratios: Riser Diameter: 6" CSA 28.27 sq in Base dimension 4.34 (72% of riser) Riser height: spec: 43" actual 48" Firebox width: Spec 8.6" Actual 9" at the bottom, 8" at the top (this was for practical reasons of constructing the roof out of 9" brick- next build I may cast a roof or just use ceramic fiber sheets.) Firebox height: Spec 13" Actual 13" Firebox depth Spec: 23.8 actual 24" Port height spec: 9.5" actual 9.5" Port width spec 2.17" actual 2.25" Port depth spec: 2" actual 2" Primary air spec 20% of riser CSA 2.37X2.37 actual 20% 2.37X2.37 (I found a lucky tube size that was perfect) Primary air spec: coming from the door of the stove. Actual:coming from about 1/3 to 1/2 way from the door to the port end through the floor/pellet basket Secondary air final tube (vertical tube) Spec: 5% of riser CSA 1.41 sq in. Actual 4.76% of riser CSA 1.34 sq in Secondary air final tube (vertical) dimensions spec: 2.17W X .65D Actual: 1.77W X .76D (inside measurements) Secondary air horizontal part: Spec 50% to 100% larger CSA than final tube Actual: 59% larger CSA Distance from secondary air outlet to port (back of the firebox): About 3" horizontally. Air outlet is just below the top of the port in the vertical dimension. I see recommendations from 2" to 3.25" for this dimension (I can't tell if these apply to bottom channel or ceiling channel designs). I'm thinking that the ceiling channel air outlet has to be much closer than this to the top of the port in order to experience the lower air pressure at that spot. How close should the air outlet of a ceiling P-Channel be to the back face of the firebox/top of the port?
|
|
|
Post by invention1 on Sept 16, 2018 19:18:49 GMT -8
Alright now about the cordwood run. I think it was less successful on cordwood overall.
Airflow was still measured at zero at the secondary air inlet through the whole run. There was one time when I knocked my replacable P-Channel out of place (later it will be pinned) and did measure some airflow, but this was definitely not in the right spot in the stove, it only proves that if there was airflow into the secondary tube, I'd have seen it.
Early in the burn, there was little flame in the riser, no ram's horn pattern established, then the fire essentially went out at 18 minutes, just smoldering, white smoke. I put the P-Channel back in place, and relit.
At 30 minutes into the burn, cordwood fully alight, but no ram's horn pattern. Smoke was clear at this point.
46 minutes into the burn, there was a bit of flame near where the primary air comes in, but wood was coaling, but wood not used up.
56 minutes wood was just smoldering, wood not used up, not really coaling anymore just sitting there. A bit later the wood was 100% charred, but essentially had gone out, hot air coming out of the riser. I added a few handfuls of pellets and hit the charred cordwood with a propane torch to relight. This was the second time it needed relighting.
ten minutes later, it was clearly acting like a rocket stove - gases exiting the riser were glowing red (did not measure temperature for fear I'd ruin my probe) good rams horn pattern, clear smoke, sounds like a rocket stove should sound.
40 minutes later, cordwood all coaling, no flames.
So in summary it was a poor run on cordwood, better run on pellets. Pellets are perfect fuel, this cordwood might not be all the way dry I don't know, but it didn't burn well and had to be relit. I may not be loading the stove right, too mush air space, too many big pieces of cordwood, not enough kindling (or it should be all kindling) Once it was all charred, it went up like a 4th of july firecracker (or Guy Fawkes effigy, depending on where you might live).
I'm going to fix the secondary air outlet before another run, check the pellet basket for cracks, and see how easy it might be to clean after a cordwood run.
|
|
|
Post by peterberg on Sept 17, 2018 0:23:23 GMT -8
Distance from secondary air outlet to port (back of the firebox): About 3" horizontally. Air outlet is just below the top of the port in the vertical dimension. How close should the air outlet of a ceiling P-Channel be to the back face of the firebox/top of the port? As far as know there's no distance mentioned other than zero. So, the p-channel should be dead against the back wall of the firebox. As such it restricts the port somewhat which is good because the lowest pressure in the venturi is just behind the narrowest point. When you think about it you'll see this lowest pressure point is exactly where the opening of the p-channel is situated. See this chapter at the batchrocket site. All SketchUp drawings where a p-channel is drawn shows the thing flat gainst the back wall. With a small overhang in order to create the narrowest spot in exactly the right place. Now it should be clear that at the height of the overhang the channel should only be cut out at the back. That recommendation of 3" in front of the port is clearly wrong, I would appreciate a hint where on the site it is so it can be corrected.
|
|
|
Post by invention1 on Sept 21, 2018 9:03:26 GMT -8
Once one gets a wrongheaded idea, it is very difficult to dislodge it before the train actually flies off the end of the incomplete bridge. Peter, I think I was confusing recommendations for floor channels, which do need to be spaced back from the port, with ceiling channels and that sent me down a long wild goose chase. This sort of information is obvious when you know where to look, but really hard to find if you start off with a wrong idea. Thanks for being patient. Images of floor channels are clearly spaced a few cm in front of the port. I started building this prototype with a floor channel. Later I realized that a floor channel would be impractical in this design and thought I could just flip it over on the ceiling. This was the first mistake. Also, I don't want a stove with a consumable part, and after watching Matt Walker say his pre-port tubes die twice a season, but folks with overhead P-Channels say they last and last, I was convinced. In this post, it is stated (for a floor channel) that the distance from the port wall to the secondary air outlet is 0.5 to 0.75 times base figure, and so I went off assuming that a floor and ceiling channel would work about the same (wrongly). donkey32.proboards.com/post/24506From batchrocket.eu: "This secondary air needs to be added to the stream in advance of the strong turbulence in the port and behind. Supplying air in the riser itself seems obvious but it won't work, mixing isn't powerful enough this way." The words " in advance" sent me off on a wild goose chase to find how far " in advance" or "ahead of", when the answer was actually " immediately in advance". The images there don't really detail the distance (zero) between the P-Channel and the back of the firebox. I don't think that paragraph actually comes right out and says that the P-Channel is right smack up against the back wall of the firebox so that's probably a good thing to add. Maybe change "in advance" to 'immediately in advance, touching the back wall of the firebox" so that dunderheads like me won't misunderstand. Or just show the picture below. Some great pictures show up in other places on batchrocket, but I hadn't noticed them yet. Finally after a long period of having a completely wrongheaded idea of what I was building, I ran across this image on Permies. It was time to smack my head and say 'D'oh!" like Homer Simson. If this image originated from you, then bravo, it is a great detailed drawing. Now I've built a new overhead P-Channel to my new best understanding of the requirements. I've also fixed another detail - the firebox I built slopes in a bit so that a 9" brick will comfortably span the top. It's really meant to be a bit over 8" at the top but it was more like 6". This would reduce the change in velocity into the port near the top, and reduce the Benoulli effect, also pulling in less secondary air. I'm reconstructing it now a lot closer to spec, and I'll check the airflow again to make sure it is working. I'll inspect for leaks too, I think I have a bunch in the door and the ash drawer. I started out intending to build 3 prototypes - this will be version 3.1, but the mistakes I'm making here will be fixed when it goes in the living room. This is why we build prototypes!
|
|
|
Post by peterberg on Sept 21, 2018 11:12:15 GMT -8
OK, you've got the right idea now. Some pictures weren't clear enough for some people and this is one of those. They placed the p-channel against the back wall but the overhang consisted of only one "lip". I'd think it's clear the Bernoulli effect is weaker this way, with the sides of the overhang being cut out. What you see above is an older design sporting a recess in the back wall which I won't recommend anymore so I did another for the site.
|
|
|
Post by invention1 on Sept 23, 2018 18:06:08 GMT -8
Fixed up some problems - fixed the secondary air inlt so it is to spec, plugged up some leaks, added a gasket on the ash door.
I did three runs this evening.
First run with pellets. Results terrible. Lots of thick smoke, never developed a ram's horn in the riser. Had to restart it after 15 minutes. Flames were rather tepid at best. Not much flame visible in riser. No ram's horn. Made measurements of primary air (always about 12-13CFM after the stove was running). After 30 minutes, stirred and relit, smoke cleared up but no flames in the riser. I'd describe it as a lazy fire.
I was also measuring secondary airflow. It was zero every time I measured it. It I licked my finger, I thought I could feel a bit of air, but this lab quality meter could not measure it. As I said before, it can measure the airflow from saying the word "Ahem" while clearing your throat.
I have a teeny hole at the secondary air port where I can actually measure the pressure inside the secondary air tube, right up near the outlet. I have a small metal tube in there, and was generally measuring about -2.6 Pascals at that point. That is a really small pressure. .00037 PSI for those familiar with that measurement.
Cordwood Run:
Once the pellets were coaling, I stacked cordwood on them, hit it with a torch for a bit to encourage it, and it started up. Had to leave the door cracked open to get the smoke to clear up, then it ran well. Good well developed rams horn. Glowing stream of gasses at the top of the riser. Primary airflow about 12 CFM. Secondary airflow still zero. Negative pressure at the secondary air pipe outlet still about -2.6 Pa. Temperature inside secondary air pipe (I can also stick my thermocouple probe in there) about 377C. This is how a rocket stove should behave. It was better than my previous cordwood run, I think I've learned how to stack it and fire it.
Another Pellet Run:
So then when the cordwood was all coaling, no more flames visible, I raked it all into the pellet basket area, and dumped in a bucket (about a gallon) of pellets. Didn't use the torch, just stirred it a little. Left the door cracked for a bit then shut it. Smoke clear. After about 15 minutes there was a good well developed rams horn, clear smoke, primary airflow about 13 CFM, apparently a really good burn. Secondary airflow was zero. Still about -2.5 or -2.6 Pa pressure in the secondary air tube. Flames not really filling the firebox, but strong enough to go out the port and swirl up in the riser.
[EDIT] it looks like, although the airflow meter I am using can measure a very small amount of flow, it has a minimum velocity it can measure. Airflow in the secondary air tube is below that minimum air velocity, so no measurement of flow. The negative pressure in the secondary tube is about the same as the negative stack effect pressure in a common hot water heater flue, although small it should be enough to move airflow. Next run I'll check flow with a smoke puffer and not worry about the airflow meter.
In summary: A terrible run on pellets, a great run on cordwood, and then a great run on pellets. So what to conclude? Is this thing working? or not? Why was the first run with pellets terrible, the second pellet run good?
Pellet fires seem weak and not large enough to fill up the firebox. Sometimes they are tepid fires. Is the 6" stove just too big for the amount of pellets I am using? Is my pellet basket holding a 4" stove amount of fuel in a 6" firebox?
All of the primary air comes in through the pellet basket. There's plenty of cross sectional area there, if it isn't choked with ashes. Free open area of the pellet basket is about 18 square inches (Primary air is 5.6 square inches, Riser CSA is 28 square inches for comparison) Pellets don't impede airflow much unburnt. That primary air is doing a great job with cordwood, no problems. Am I firing it right? Am a starved for primary air? or starved for air at startup until it gets to drawing? Dumping in too many pellets at once? or not enough? Should I start a few pellets then feed them in slowly once the first batch is going?
It seems like this thing isn't ready for prime time just yet, but I'm struggling to figure out what's not working right on it.
|
|
|
Post by Vortex on Sept 24, 2018 1:45:05 GMT -8
A cold stove is a very different beast to a hot one, a hot stove is a lot more forgiving of design faults. A small fire in a large cold firebox can't raise the firebox temp up to the point where it can burn clean before it runs out of fuel.
|
|
|
Post by satamax on Sept 24, 2018 1:57:55 GMT -8
Invention1.
Two tricks to visualize if there is airflow, Kirk's one, light some incence, and wave it in front of the ports. Or mine, do it with a lighter or candle. If it pulls flame, it's drawing.
|
|
|
Post by invention1 on Sept 24, 2018 5:02:29 GMT -8
A cold stove is a very different beast to a hot one, a hot stove is a lot more forgiving of design faults. A small fire in a large cold firebox can't raise the firebox temp up to the point where it can burn clean before it runs out of fuel. Yeah, this makes sense. Thinking this through, it seems my firebox is too big for the amount of pellets I'm putting in there. The size of the pellet basket is a highly scientific calculation from Professor SWAG (S*****y Wild-Ass Guess) so that is not unexpected. Adding more pellets just makes a big smoldering pile of pellets. Pellets don't burn unless there is airflow through them. So it isn't easy to upsize the pellet fuel. I think I'm near the limit now.
|
|
|
Post by invention1 on Sept 24, 2018 5:11:56 GMT -8
Invention1. Two tricks to visualize if there is airflow, Kirk's one, light some incence, and wave it in front of the ports. Or mine, do it with a lighter or candle. If it pulls flame, it's drawing. I have a bunch of professional equipment for this from my gig as a home energy auditor (which is why I have lab grade airflow meters and temperature gauges) one of them being a smoke puffer that's used in proving draft for gas fired equipment. I'll employ this on the next burn. HOWEVER if the airflow into the secondary tube is so low as to not be measurable, but visible with a wisp of smoke, then I would argue the secondary air tube isn't accomplishing anything. [EDIT] it looks like, although the airflow meter I am using can measure a very small amount of flow, it has a minimum velocity it can measure. Airflow in the secondary air tube is below that minimum air velocity, so no measurement of flow. The negative pressure in the secondary tube is about the same as the negative stack effect pressure in a common hot water heater flue, although small it should be enough to move airflow. Next run I'll check flow with a smoke puffer and not worry about the airflow meter.
|
|
|
Post by invention1 on Sept 24, 2018 6:53:59 GMT -8
So here's a mental exercise that I did last night:
Let's say I can show that my firebox, being derived from a cordwood rocket stove, is really optimized for cordwood. It's a 6" RMH, pretty much built to spec. This box is too big for the pellets to work. If you add more pellets, they just smother. I'm wondering if the pellets are starved for primary air, yet the pile is too big to let enough air through for proper combustion? More pellets won't fix it, they just won't burn in a big pile.
Lots of hacked and commercial pellets stoves (the Wiseway, for example, which has been copied by a lot of hackers) are based on a 4" model. They usually use 4" steel tubing as a basic construction material. We know better here, and try to use refractory materials and insulation to achieve peak temperatures along with calibrated airflow. But I haven't seen a hacked DIY pellet heater that is bigger than 4". Commercial pellet heaters are 3" usually.
Could I somehow combine a section optimized for cordwood with a different firebox optimized for pellets? Is doing everything in the same firebox part of the problem - it is not optimized for either?
Answer: NO! NO! AND HELL NO! You can't put two stoves on a flue!
Question: Why not?
Answer: BECAUSE I SAID NO IN ALL-CAPS!
Actually digging into the physics of this it isn't impossible. The first flue I built (before I knew where to ask if this would work) had two wood stoves back-to-back across a wall. It worked great with one, or both stoves fired up. The idle stove we'd just shut off all the air inlets. Never any backdrafting that I experienced. That flue is still in use 40 odd years later, I know the folks who live there now. Two stoves on a flue, quite successful AFAIK.
If two stoves are on separate floors, it can create a world of trouble: the upstairs stove can let out smoke and CO, or the downstairs stove can backdraft, also letting out CO. But on the same level, with a strong and properly sized chimney, they work quite well, and thousands of old houses did this.
Answer: It's against code!
Actually, IRC 1003.14 allows multiple appliances to be connected to a flue as long as the flue is properly sized. So, no, it is not against "code". It is against most people's prejudices. This is why the usual responses ("Hell No!") quoted above are always in all caps and do not include any actual reasoning. Google this question for several hilarious examples.
The bell for a 6" stove needs about 57 sq ft of inside surface area. I have planned to achieve this with a steel/masonry bell of 31.5 sq ft, and an additional deadend bench of 25.4 ISA.
The bell needed for a 4" stove is 39.8 sq ft. Hmmm ... that's about the same as the masonry/steel bell I've planned to make. If I had no deadend bench, the bell would be just about right for a 4" stove. 39.8 ISA bell with a 17.2 ISA bench. Small bench, but doable.
What if you can switch off the deadend bench?
So let's do a thought experiment: Two stoves, side by side (actually within the same overall masonry enclosure) and insulated from each other/the world with ceramic fiber, both lined with firebrick.
The 6" stove is a standard cordwood RMH with a 6" riser in a bell about 20" diameter. The bell is sized for the 4" stove, with a deadend bench making up the difference in ISA to accommodate the 6" stove.
The 4" pellet stove also has a riser in the same bell. Both stoves are designed so they can be shut tight when not in use. There is also a damper between the base of the bell and a deadend bench. For using the 4" stove, one shuts the damper. For the 6" stove, one opens it. So the system has a bell that can do double duty - the bell/deadend bench is sized right for either system.
The 4" stove works a lot more like a J-rocket than an RMH. Most DIY pellet hacks seem to work this way - 4" air inlet, 4" firebox, 4" flue - sounds J-Rocket-ish to me. The Wiseway pellet stove has a secondary air inlet, that is basically the only control on the stove. But we're still getting the advantages of heat exchange and storage from the bell, compared to a standard commercial pellet heater. Although 4" square opening is way too big for primary air for an RMH, that big primary air opening may be needed to get full airflow through the pile of pellets, which is obstructing airflow some. J-Rockets suffer because one has to keep feeding them - that problem is solved with pellets, which can be automatically fed.
I'm really happy with my ceramic pellet basket experiment. I've done 6 runs now, with cordwood and with pellets, not breaking a ceramic element. I've taken no special care, the ceramics are just out of the way of the cordwood. There is no metal exposed to flame. Most pellet heaters burn out the pellet basket every 6 months or 1 year and the ceramic pellet basket solves this problem.
Pellet stoves have several features that optimize them for pellets. First, all of the primary air goes through the pellet basket. The draw forces air through the minor airflow obstruction of the pellets. Some actually use a fan. The Wiseway and many DIY hacks use a gravity feed pellet system - the thing just feeds itself, and as long as the pellet box is airtight, there is no back-firing into the fuel supply. Others use a pellet auger, and I've developed one that works pretty well, but augers and motors are noisy. If the auger ends up in the system, it'll be on 12V and will work fine for weeks without utility power.
Under the pellet basket is an ash cleanout. This can be a drawer or just a door, but as the pellets burn, smaller particles fall through into the pit. The ash pit needs to be airtight or else built as a secondary burn chamber.
Now an ash pit is a fine feature to add to an RMH, and after working with one I'm certainly going to consider incorporating this feature.
So the basic idea is two stoves, one optimized for pellets one optimized for cordwood, side by side, each stove with a tight shutoff, and a damper in the system that switches the bell size. Shut down the stove you are not using right now. Turn the damper for the small one or the big one as needed.
It certainly takes a little skill to use such a dual device, as there are two different settings and some fiddling with dampers. I'm sure it would work, just not as well, if one forgot to change the damper to the right position. What happens if the bell is too small? Mostly you'd lose some heat exchange, air going up the flue will be too hot. What if the bell is too big? The bell would not get as warm, and you'd lose some heat exchange. What if someone ran both stoves at once? It would probably work, given a strongly drafting chimney. What if you leave the idle stove open? I'm not sure what problem this would cause. If the idle stove doesn't smoke and backdraft, maybe no problem at all. The draw in an RMH comes from the riser, not the flue. Maybe the extra air from the idle stove would cool the bell and make it not exchange heat as well. Unless the idle stove backdrafts, all I can see this would do is make the system exchange heat with the house poorly. I don't see a showstopper problem with having both appliances in the same bell, it just won't work optimally if one does not set it up correctly. I will certainly examine this with lab grade CO detectors (not just an alarm) before putting my blessing on it, along with a permanent CO alarm nearby.
I have a really strong draft on my flue. It's about 26 ft to the top, and when I've measured the draft it's been at the top or a bit above the recommended ranges. I'm not worried about the problem of insufficient draft, in fact the opposite.
I haven't decided to do this, just decided to think about it for a while. Anyone else see a showstopper problem with two stoves on one properly sized, switchable bell? Besides, of course, the problem that this idea is completely insane. Other than that.
|
|
|
Post by invention1 on Sept 26, 2018 18:32:27 GMT -8
I have a hypothesis that the pellets are wanting a much smaller firebox than cordwood, for the size pellet basket I am using (which is arbitrarily picked anyway). I dry-stacked some thin firebrick in the firebox to simulate a much smaller firebox chamber, fired it with pellets, and indeed by the crude indicators I am using it was a better burn. The pellet firebox is about 6" wide X 4.5" tall, about 16" long, with a pellet basket grate area of about 8X9", 37.5%A free area. Total free area of the basket is about 27 sq inches, which is about the same CSA as a 4.5 X 6" firebox. Primary air is smaller, about 2.3X2.3 or 5.6 square inches.
Pellets tend to burn about 20-30 minutes, versus the 45 minutes typical with cordwood. One of the advantages of pellets is that you could theoretically add them continually during a burn, extending the burn time indefinitely, but the disadvantage is you probably need to do that to get everything up to temperature and running right. Also to add up to the KW of fuel energy that a batch of cordwood represents, one would have to feed pellets for longer.
Several times during the burn I was able to get flames to travel into the riser and form a ram's horn. It is about 9" from the edge of the pellet basket to the port, often the flames didn't reach this far. In this case the port is now 2.25" wide, and about 4-1/2" tall, (versus 2.25X9.5" for a spec 6" port) because my quick-and-dirty dry stacked firebox is 4-1/2" tall. The riser is still a 6" diameter riser 48" tall (a little longer than spec).
Adding pellets tended to create heavy smoke for 2-3 minutes. Sometimes opening the door would help it burst into flame from heavy smoke. I'm wondering if this means it is starved for primary air, or just that's the way it needs to start up, with more air and the door open. I started the pellets with a pint cardboard box of alcohol-soaked pellets - this starts the fire quickly and gets things going fast. Starting pellets with the torch works, but they tend to smoulder once the propane torch is removed, I think alcohol gel or alcohol soaked pellets does a better job of getting them going fast and getting the firebox up to temperature. I'm going to make up a batch of alcohol gel soon.
I was able to use a smoke puffer and verify that, indeed, there is plenty of secondary airflow - I was earlier fooled by the minimum airflow capability of my meters, indicating zero airflow. I can't quantify it, bu the air was moving into the secondary air intake quite rapidly verifying plenty of secondary airflow. Most likely, all that secondary airflow just went up the chimney, because currently the secondary air inlet isn't down where the pellet flames are located. Eventually it'll need to be relocated to where it can do more good.
I was able to get flames and a ram's horn into the riser, although not particularly consistently, and it was better toward the end of the run after things had come up to temperature. The extreme temperatures usual in a cordwood burn were not achieved (I didn't measure anything just estimating this time). The lazy burn pattern I noticed in the last burn in a fullsize firebox was gone.
Primary air intake is the same one that's been there, sized correctly for a 6" system size. It seems to work OK with a 4X6 pellet system, and a pellet basket with a grate about 8" X 9". In the pellet configuration, this is essentially a J-rocket, as I suspected earlier. I say that because the J-Rocket has a consistent system size throughout the firebox and riser, whereas a standard RMH has a larger firebox, a smaller port and a riser all different CSA to achieve turbulence. I think incorporating a port into this design is still a good idea, getting that turbulent mixing going if I can ever get flames out that far, or get the pellets close enough to flame into the port.
This experiment is telling me that, for the size pellet basket I have, I need a much smaller firebox, and shorter as well. The pellet basket/firebox could be almost right up in the port.
Currently I don't have a pellet feeder in place, I'm adding pellets by hand occasionally. I think I could get the fire hotter and more consistent by feeding pellets slowly once the fire was hot and flaming, currently adding handfuls of pellets is putting the flames out, leaving coals to reignite after 2-3 minutes.
So far the experiment is pointing toward the dual-firebox cordwood-pellet idea isn't impossible, even on a single riser. I'm now imagining two parallel fireboxes side by side, one optimized for wood and one for pellets, but on a common riser with a common port. Or maybe they are both sidewinders with two different size ports on a common riser, and getting air from a common primary air inlet. It may work out well to use a floor channel after all, as it might be able to deliver secondary air to the cordwood firebox as well as down low where the pellet firebox can use it too. If the pellet basket is moved off to the side to it's own little firebox, then the floor channel (or S-portal or Walker Pre-port, if that's how it ends up configured) won't have any obstructions to constructing it that way. I'm beginning to think both fireboxes could use a common primary and secondary air system, eliminating any concerns about backdrafting and so forth. There may need to be a damper or other means to shut off the nonfunctioning firebox. A blank piece of steel over the pellet basket would work, as that area won't get hot when there is no fire. Same for the regular cordwood side, just blank off the primary air inlet. It would also be cool if there was an actual switch that would cover one or the other on a hinged piece of steel.
I don't have a testo, thus no way to really calibrate any of these sizes or try varying them to optimize performance. However, if I can achieve the metrics of clear smoke and a ram's horn in the riser, then that indicates it is certainly in the right direction. Unless I can get better instruments that's the best I'll be able to manage.
|
|
|
Post by peterberg on Sept 26, 2018 22:17:18 GMT -8
In my opinion, the distance from the pellet basket's edge to the port should be much less, 2" or smaller would be more like it. In that case the gases escaping the basket are still hot enough while reaching the port. Also, with flames closer to the port the p-channel would heat up quite quickly, delivering hot air at that point. It's all a question of balance.
|
|
|
Post by invention1 on Sept 28, 2018 7:22:42 GMT -8
In my opinion, the distance from the pellet basket's edge to the port should be much less, 2" or smaller would be more like it. In that case the gases escaping the basket are still hot enough while reaching the port. Also, with flames closer to the port the p-channel would heat up quite quickly, delivering hot air at that point. It's all a question of balance. I'd agree, the pellets need to be much closer to the port. But one more epiphany that struck me recently - I believe the pellets want a smaller firebox, or else the firebox wants a bigger pile of pellets! Now, I think that a bigger pile of pellet fuel will just smoulder. So I tried a quick-and-dirty smaller pellet firebox. I arranged some thin firebricks in a dry stacked firebox right around the pellet basket. The volume of this pellet firebox is about 872 cu in/ 14300 cc, whereas the conventional cordwood firebox volume is about 2652 cu in / 43500 cc. Pic and drawing below will illustrate it better than words. The results? Instead of a lazy, not very vigorous pellet fire, and cool riser temperatures, there was a roaring fire, flames reaching into the port, at the peak the temperature at the top of the riser reached 450c. Smoke absolutely clear. Went about 20 minutes like that then started to slack off, I added more pellets and it got to that state again. A very successful run. Same fuel, much smaller firebox. I was inspired by Peter's Batch Box Rocket 17, where the riser is on top of the firebox instead of behind it. In this case the conventional RMH firebox becomes part of the riser when burning pellets. And oh about the port: The conventional port in the cordwood firebox isn't where the mixing and turbulence happens. I arranged a little restriction at the outlet of the pellet firebox. This isn't calibrated yet, but for the pellets, that restriction is the port. Once I perfect this, there will be a secondary air inlet at the end of the pellet firebox, and a port at the end of the firebox. The conventional firebox for the cordwood is just part of the chimney at that point. The conventional port at the end of the cordwood firebox doesn't do anything - fuel is already mostly burnt by the time it gets there. System size? The conventional firebox is a standard 6" RMH. The pellet heater is a J-Rocket. The system size for a J-rocket is the same as the air inlet. The air inlet for this stove is the same as the primary air inlet for the 6" stove. System size for the pellet J-rocket is 3"-ish. The floor of the standard 6" RMH is the roof of the pellet firebox. The port in the top of the pellet firebox is also a convenient ash cleanout in the floor of the RMH. The drawing below isn't a practical stove yet - it is just a hack. However here is the idea at this point: A conventional 6" RMH, arranged over-and-under with a pellet firebox below. If the cordwood firebox is in use, the pellet area is just the primary air inlet. If the pellet firebox is in use, the cordwood firebox is just another part of the riser. Both at the same time would starve the system for primary air and thus not recommended, but would not cause any backdrafting or CO problems. The bell will be sized for a standard 6" RMH. The pellets can burn less fuel at a time, but longer times, thus creating an equivalent amount of energy to dissipate in the bell. This isn't physics yet, but it's how I'm going to approach it jsut to keep it simple. It's a crazy enough idea as it is.
Photo of dry-stacked pellet firebox
Drawing of what the system might eventually look like:
|
|
|
Post by invention1 on Sept 28, 2018 18:02:51 GMT -8
While people are chewing on the two-firebox idea (and no doubt rounding up men with straightjackets because it is admittedly crazy) I'm playing with how to deliver pellets to this beast. I've made some pellet augers that work pretty well. They are noisy, as such things usually are and involve motors and moving parts, a point of failure. The advantage of a pellet auger would be one can modulate the flow of fuel. But in a gravity feeder, the system works at full capacity, feeding pellets as fast as they burn up. Unless you shut a guillotine grate on the pellet supply, it will keep going until your pellet hopper is empty. I still am not sure which one will end up winning. Several hackers and Wiseway stoves have been successful with gravity pellet feeders. This guy at Bigelow Farms has done several mods on a very low tech pellet rocket stove, and he's found a 45 degree pellet feeder works OK, but others have generally gone more than 45. Nonetheless his system is pretty well developed. Check the diagram at 2:49. The first key is that the feeder pipe has to be steeper than 45 degrees. The key to using one in a rocket stove will be to keep metal as far out of the burn chamber as possible. One might also use a smaller pellet hopper in a gravity feeder, using the filling of the hopper to regulate fuel supply. But an auger could have a huge hopper, regulated somehow by electronics, timers, or thermostats. I'm still not sure if gravity or augers will prove out to be better overall. To this end I built a wood and PVC gravity feeder model. The wood part is approximately the dimensions of the pellet basket in the real stove. I started with a 2" diameter PVC tube as this was laying around handy. In this stove, the pellets will have to feed from the side, although most pellet burners feed from the top. Will it fill up the basket evenly or all pile up on one side? Will the pellets burn back up the feed tube a long way, due to the extreme heat of the rocket stove model? Will the 2" tube be too small? I took some pics of a pretty successful run. The key seems to be that the 50% mark of the opening has to line up with where you want the top of the pellet pile to be. The feeder tube is much steeper than 45 degrees (I haven't measured it yet) and parts outside the stove will be vertical. If there is a feeder hopper, it must be sealed airtight or you will get burnback. Some people just have a larger tube and fill it with enough pellets to go for a few hours. My goal was to have the pellet basket mostly full of pellets, with the pellets backed up the feed tube, after pouring them into the end of the tube. Removing a few pellets should make more fall down the tube. See what you think:
Wooden pellet basket and feeder model. Pellets stack up in the feeder pipe after filling the basket. As the pellets in the firebox burn up, more fall down after the ashes crash. Some will fall through the pellet grate, some will require a shake or a cleaning every so often to clear. That's OK, pellet grate needs to be easy to clear, and will likely need this twice a day. So far merely lifting the grate and giving it a tap has cleaned it completely, but that's only after 3-4 burns. Remains to be seen how it stands up over time. If it doesn't last I'll just make it out of stainless and plan on making a new one twice a season. Pellets have stacked up relatively evenly in the pellet basket. Although higher on one side, it is not excessively so. As you can see the feeder pipe ends at what will be the edge of the masory, hopefully protecting the steel from erosion. I should make this part replaceable as well as I don't know how long it will last. Or maybe the last few inches will be made out of brick, a steep ramp, or cast out of refractory.
|
|