Reverse engineering Riley pellet burner

[rectifier]

Ran another, more extended test today while I happened to be out working on my van. Turns out the 1hr on the initial test was only due to lack of pellets keeping it from getting really hot. With the hopper full and throttle open, it can eat 5.7lbs of pellets in 50min, estimated 65kBTU on pine. That's a lot of heat for a small stove and the casing was glowing red. Unfortunately, that's pretty much a theoretical max as 99% of that heat is just going to drive the draft.

Under real world conditions with some form of heat exchanger it should be running at quite a bit less draft.

I then pushed the throttle in to about 50% opening and was able to burn another 5.7lbs for 2 hours in the same configuration - less than 30kBTU, quite a bit of adjustability available, and a more reasonable heat output for say, a large tent. It is nice that it doesn't smoke at all at reduced power since the mixture doesn't change, only the volume of the burn chamber. The exhaust doesn't even smell of wood, it smells like light diesel exhaust once it is up to temp - a consequence of gasification as a burn method and a sign of very complete combustion.

I also have the goal to build this burner so you can slide it into a RMH feed as well, or build it in at a tee to the burn chamber, giving you an option to sustain a continuous burn. I agree that welding is not an option for many people, but a heavy welded firebox should outlast sheet steel by quite a lot - so I am considering offering these burners for sale once the kinks are worked out, I think many people should be able to make use of them. The 4" box corresponds to ~5" system size, which of course... nobody has. But at 65kBTU it should put out enough fire to heat up a 6" system, no problem. As it is, with a ridiculous amount of draft it is still not enough to pull the flame off, so it shouldn't be an issue. I wish I knew someone with a RMH nearby to test it with - here in the big city they have banned all wood heating technologies in new construction for a decade...

I've decided my first build using the burner is going to be a rocket wok burner for outdoor stir-fry and large boiling tasks like beer brewing. It should be a blast, stay tuned! Next will be a classical rocket barrel heat exchanger, without the thermal mass, unfortunately. I think with the long, slow, super-clean burn available, there is a new option to do away with the mass and use an efficient heat exchanger to continuously suck out the heat post-rocket chamber with liquid cooling instead.

I am interested to see how your sheet design comes out, it should be great for experimenters without the ability to weld, as well as potentially being very low-cost. The Clarry is a good example of the level of performance you should be able to achieve. Watch the rivets, I found the aluminum ones tend to melt once a stove gets going, then flaming stuff goes everywhere! Don't ask me how I know

[rectifier]

Slight shift in plans and some tweaks. A proper rocket burner required some insulating materials that weren't at hand, and I'm busy.
So, I added a flange and bolted it on to this motor housing to produce a small stove. That fat stovepipe is actually a jacket around a 3" galvanized pipe that is 2' long, to give it extra draw since it is short and skinny.
For this initial test, the actual stove firebox is far too small for the burner and has no baffles of any sort - efficiency is terrible, with flames occasionally flying out of the chimney at full power, even after removing the jacket. However, it's more stable and a more practical testbed than burning straight into a chimney.

The stove body came up to about 350C on the top and 250C on the bottom of the barrel. It was hot to stand next to. All the paint burned off, a minimum of smoke was produced in the exhaust, the burner firebox hit 400C.

After careful inspection of the Riley photos, the weld holding the hopper to the firebox was cut at the front of the hopper to create a gap. This gap prevents conduction of heat up into the hopper! Also, it sucks any woodgas created in the hopper down and burns it. One tack was left in the middle for support. The hopper was sealed tightly at the top with a plastic lid to prevent any drafting of gas up it.

The Riley stove has a max throttle that is not 100% - the throttle body sticks into the hopper opening a bit as it is drawn. This time I did not open it to 100%, to mimic the Riley stove. It seemed to prevent coals from smoldering up the slope and into the hopper.

Edit: see the mess on the front of the hopper? That was an attempt to add a cooling jacket to stop the gasification of hopper pellets. It worked, but wasn't practical, just proved the point that the hopper needs to be kept from getting so hot. The gap seems to do the job, I may fill it with stove cement to keep oxy out of the hopper while still being insulated.

Attachments:

[rectifier]

The stove also has a port on the other end to look directly down the firebox. It was interesting to watch how it starts, but after it came up to temp, it basically always looked like this - too bright to see the firebox, and blasting straight from the firebox to the stack. As you can see the mixture appears to be pretty good with not much excess air. The actual draft pulled by the firebox is quite lazy.

There is a ton of turbulence created by the square pipe, under lower power it rolls out in a pretty pair of long spirals, but I didn't get a picture of that.

Any ideas on slowing this 2' horizontal flame inside the firebox to get more of the heat out? I'm just thinking of forcing it down around a baffle or two for a start

Attachments:

[rectifier]

Reversed the chimney to above the firebox and added a baffle to spread the flame and force it to travel the length of the firebox.
New temps: after 1hr, 225C on top, 250C on the sides, 210C on the bottom of the barrel, 180C on the chimney base, 170C on the stovepipe drip cap (closest I can get to exit temp with just an IR thermometer)

Smoked a bit, but it was fine soot, flame scraping doesn't help but this also may have been the paint on the baffle slowly burning off (scrap metal...) as it cleaned up towards the end of the burn.

No time today attempt to reduce the power, it looked like temps were still increasing. Still a much better result. This is actually doing great for a stove that has a 2' chimney.

Attachments:

[gooch]

Could you post some pics of the guts for your stove.
Thanks

[rectifier]

Guts look much like the Riley pics but with thicker walled tubing, so everything is a bit tighter together. Next time I have it out I will snap a photo or two.
Furnace wasn't running at the shop today so I burned the stove all day to keep the chill off, cure the stove paint I put on, and give it a long term test. Everything worked fine until the last pellets were being burnt. At the transition from "pellets coming straight down the pipe" to "pellets sitting at an angle in the bottom of the pipe", the woodgas/air mix that was in the hopper was exposed to flame, and it blew the lid off with quite a bang! Going to have to do something about that!

The cast iron housing is too heavy/too much thermal mass to work well. It doesn't get hot enough to radiate well and smokes except at lowest power levels. I'm looking at buying a real cheap surplus stove to add this burner to, firebox slung underneath to make room for a rocket preburn chamber probably.

[gooch]

Sounds good! I look forward to the pics and the outcome. I'm sure that must have been some surprise when the lid blew off and glad you weren't hurt.

[rectifier]

No cheap stoves at the surplus shop, so reconfigured the stove to a rocket today, added perlite and took photos - however I left the phone at the shop so no pics today, sorry. I've been using the phone since it takes dinky photos that are good for this site and some of its dialup users. After the design is finalized I'll take some real photos and post them on a storage site.

If anyone is following my temperatures I'm switching over to F since that seems to be a standard for wood stove-related temperatures.

Anyways, the "pellet rocket" is a full-on success, with the top plate of the stove coming up to 650F in the direct blast of the rocket jet. Meanwhile the initial stovepipe temperature is below 300F and the exhaust (6' of 3" pipe away) less than 250F. During warmup a fine particulate is visible in the exhaust, but only if you shine a laser through it. After the stove is hot it is perfectly clear.

Warm-up lasted longer than it should have, then took off with a sudden rush. Over 30 min, the temperature crept up to about 300F on the top plate. Then I looked over 10 minutes later because some grease on top was smoking, it had come up to almost 700F in almost no time. I think this was due to the cast iron stove housing quenching the secondary draft of the chimney until it heats up enough - for a long time the initial stovepipe temperature was around 150F, forcing the rocket chamber to do the work of pushing the exhaust out.

The rocket effect is finicky with respects to the fire power. At full throttle it roars out of the inspection port if you open it. Back off the throttle a bit and the flame drops out of the insulated chimney and it idles gently, maintaining the stove temperature with much less fuel. Back off much more and the feeding fails due to weak draw. So it basically only has 2 power levels. I think this is due to the positive feedback of the stack effect, once the flame reaches the chamber it builds draft until it hits an equilibrium with the 3" exhaust volume.

Can't complain too much about the rocket behaviour as this is a 5" system size pushing a 3" exhaust - surprising that it even works.

I blocked the hopper slots with stove cement - stopped the boom, but at the cost of more woodgas in the hopper at an earlier time. I think I need a check valve to keep the hopper purged, thinking of a rubber bung at the top with a PCV valve in it

[rectifier]

I'm back with photos as promised. First, the internals sitting next to the firebox in the approximate position for wide-open throttle operation.

Edit: There was a question about how I sealed the hopper to the firebox. Due to a material mismatch in the scrap (the hopper is stainless), the two were initially brazed together. A weld was attempted along the side, for structural support as you can see it didn't come out that great welding stainless with an AC stick.

There are 2 slots towards the front to insulate heat from creeping up the hopper. It turns out they also allow air to attack the pellets in that area, helping them feed. I'm not sure whether to keep them or fill them with stove cement - for now, they are filled in.

Note, that slanted hopper section is too long, resulting in ugly shutdowns where the last of the pellets smolder. I plan on shortening it for the next test.

Attachments:

[rectifier]

The firebox now has a perlite-insulated, 1' riser inside it.

Attachments:

[rectifier]

Here is the test setup for the rocket burns, this is just about all of the flue (about 1' out of photo).

It's still quite leaky, there is no seal between the firebox and the rocket chamber. There is a large gap that undoubtedly draws massive amounts of excess air - as such I expect higher performance from the final model, just need to get my hands on more 4" tubing

Attachments:

[rectifier]

Here is what happens when you open the inspection port! No more photos of the inside of the chamber guys!

Attachments:

[rectifier]

Here is the stove installed and keeping the shop's coffee room warm(er) than usual - still freezing as it's just a part of the large shop, heat rises out of the coffee room and gets away. However, a noticeable temperature rise did occur in the coffee room.

Exhaust was straight out that window, looked like an old diesel truck smoking during warmup (30min) and completely transparent during rocket operation. The stove was operated for 3 hours this afternoon, and consumed 5.6lbs of fuel at an estimated 9610btu/lb (pine) for a power rating of ~18kBTU/hr.

Stove temperature was 600F on top, 450F on the sides, 280F on the black flue section. Quite efficient for such a small stove.

During the 3-hour burn an interesting self-moderating effect was noticed in the small rocket with a negligible amount of exhaust stack. While the shell is cold, it causes the exhaust gases to cool and drop faster away from the rocket chamber, drawing a higher internal draft. As it comes up to temperature, that differential effect weakens, causing the draft to weaken and the power level to drop. This causes the shell to equilibrate at ~600F with power changing to maintain that temperature.

As such if I decrease the throttle at all it will go out after it is up to temp. Fortunately the only thing I have to do to restart it is pull it back out to full throttle and thump the hopper to drop more pellets and they will catch on the coals.

Note this stove still has a ton of air leakage as that joint is still loose (probably looser than before). I expect higher power than 18kBTU once it is sealed up.

Attachments:

[rectifier]

I went to make some changes and found the cause of recent jamming under low fire and possibly the root of the gas creep issue.

Jamming was caused by a buildup of creosote/tar/carbonized pellets on the forward face of the hopper. These were scraped out and then the face was oxy-washed to burn away sticky tars. Fresh start.

The air that caused this was likely rising up the hopper to the leaky braze that connects the two pieces of the hopper. The holes were small, but on inspection showed creosote inside, proving hot gas was leaking out of them. The slots for air wash likely helped oppose that draft drawn upwards, but they may have been a solution to a problem that shouldn't have existed in the first place. The whole hopper is sealed now, no air wash, we will see what happens.

The angled part of the hopper was shortened significantly to minimize the amount of pellets left on shutdown to smolder. It is tack welded on the sides and front for structural support, and then sealed with stove cement. The leaky braze was also sealed with stove cement (I do not have the equipment to weld stainless properly, nor do I have a piece of 4" round steel pipe. This will have to do)

Forward of the hopper a slightly smaller than 1/2" hole was drilled and a 1/2" fine nut welded on over it. This hole will serve to pour in the pilot diesel and ignite it. The bolt will come down and seal over the hole, and the fine thread should seal any leakage that gets past it (should be at negative pressure anyways during operation)


With these changes I feel this project has now officially taken the turn away from reverse engineering to actual engineering - this is now my burner, inspired by the photos of the Riley design. I am going to stop trying to emulate their burner, as that wasn't working perfectly to begin with - I think I can solve the few remaining problems and produce a better stove.

Attachments:

[rectifier]

Operated the tuned-up stove for 6 hours today with no interference other than checking fuel level. Consumed about 7lbs fuel and kept my tea kettle hot all day. 11kBTU average power.

An interesting note on load regulation to interpret as you will - the stove was warmed up with no kettle, coming to equilibrium at 600F on the top plate after about an hour. Firebox was running at 580F by the bolt hole, which seems to be a good indicator of the power input. The (large and cold) kettle was placed on top, cooling the top plate to 350F over a period of 10 minutes. During this time the firebox temperature steadily increased to 650F, and exhaust volume increased - however the exhaust flue stayed constant at 250F on the barrel output. This shows that an increase in heat extraction from the 'barrel' will be balanced by an increase in draft and thus a compensatory increase in power - a negative feedback loop that will tend to try to stabilize the 'barrel' temperature.

In any case, 11kbtu is not enough power - the draft was insufficient to burn charcoalled pellets, which piled up and stalled the feed at hour 6. Another issue was the lack of draft did not blow the ash out of the combustion area, contributing to the pileup. This stove simply needs a taller heat riser - otherwise, the new hopper worked great with a minimum of smoke-back and no combustion up into the hopper (confirmed after the stall by poking the charcoal heap - fresh pellets directly above)

I apologize for all these posts if I am spamming the board a little. I am doing a lot of testing as work is slow at the shop and also the shop is cold! Writing about the day's testing helps me think about what occurred and hopefully other people are interested to read about the progress.