Batch cookstove with a small riser

[Deleted]

Hi everyone,

I was delayed because my testo had been lost in the postal services.. happily it came as a gift at midday today.

First results are very good BUT the trials have been made with a hot heater so I still need more testing with a cold heater and in various situations.

For now, combustion-wise the heater performed 6 times better than the french norm on mass heater.

About the methodology :
- wood at 20% humidity
- lightning from the bottom since there are some red coals left
- the biggest wood is on the top
- I don't touch anything except the bypass at the beginning; I don't touch the air inlet nor the wood
- the plancha has no airtightness joints

About the Testo measures :
- I follow Peter's methodology which is similar to the methodology used for norms testing
- the testo is calibrated at 21% O2
- measures begin around 19% O2 and stop at 16.8% O2. At the end of the testing there are still big red coals left.
- measures are made approx. 120 cm above the plancha level
- CO at 13% O2 is the standard measure for mass heaters. I calculate this CO level as if the mean O2 was 13%. It's a ponderated sum. In France, mass heaters must have a CO level at 13% O2 under 3000 ppm.

About the heater :
- the idea of the 5B riser plus the small bell was mine; the idea of the half octagon was Peter's. This is an addition to the batchrocket so all credits belongs to Peter. License CCBYSA4.0.
- the riser is NOT made with insulating bricks
- the EXACT sketchup plan is here : uzume-asso.org/assets/docs/experiences/cuisiniere_donkey/batch165_sidewinder_cuisiniere_121317.skp
- 5 m of chimney stack; insulated on the 2 last meters; diameter is 150 mm
- ISA of the cookstove + 120 cm of uninsulated chimney = 5.8 m2. I count the surfaces where the gases are going upward as half the ISA.

Results :
1st run :
- heater was moderately warm (100°C on the front lintel)
- the third fire of the day

Mean values :
- global efficiency = 85,1%
- combustion efficiency (calculated) = 97,8%
- TF = 153°C
- CO = 436 ppm
- COnd = 1720 ppm
- CO13% = 485 ppm
- O2 = 13,8%



2nd run :
- heater was warm (160°C on the front lintel)
- the fourth fire of the day
- bigger wood but not a full batch, maybe around 60% of the firebox

Mean values :
- global efficiency = 83,2%
- combustion efficiency (calculated) = 97,3%
- TF = 167°C
- CO = 528 ppm
- COnd = 1969 ppm
- CO13% = 640 ppm
- O2 = 14,4%

[matthewwalker]

Very nice Yasin!

[Deleted]

For the last few weeks I've been observing the heater. I understood some important things I think :

- the threshold for the floor channel is very very important. Without it the vertical part was suffering greatly in a matter of weeks, which means that there was not enough secundary air. Adding a threshold changed completely the behaviour of the core : the flames are greatly slowed down in the firebox but there are more flames in the "riser".
On the first part of the burn, the flames go up quite frankly which is bad in my opinion because it shortcuts the turbulences of the riser.

On the second part of the burn it's a completely different matter. There is a projection of two simultaneous double vortexes. The higher one is the same than in the first part of the burn. The lower one seems to stay on the floor of the riser. Lots of turbulences in there.

Pictures of the two parts of the burn :



The problem is that there is a lot of red coals staying in the firebox for quite a while under the level of the threshold in this configuration. Much more than without threshold.

I think it a big problem because it forces the user to keep the air inlet and the closing valve open for a longer period of time. The heat transfer efficiency of the heater is lowered because of that.

Maybe we could replace the threshold by a simple brick floor whose level would correspond to the top level of the threshold ?

[Deleted]

Does anyone here knows how to calculate combustion efficiency using the data gathered with the Testo (CO,COnd,O2,CO2,TF,Tair) ?

I have a method that gives me 99,996% of mean combustion efficiency.. maybe that's it ?

pyrophile, matthewwalker, peterberg ?

[peterberg]

Yasin, these results are very nice. The second diagram shows there's more excess air in a warmer heater. Logical, the air opening stays the same but the air velocity is higher so there's a larger volume of air entering the heater. So you need to restrict the air inlet running a second or third load. This is in line with my own heater, by the way.

I also have the same experience regarding the floor channel equipped with a threshold. First half of the burn is violent and producing loud rumbling, while the second half is quieter and very clean burning. A negative effect is the large amount of coals that stays in the firebox behind the threshold. It might be that raising the firebox floor say, 50mm would act the same way to get more air in the floor channel. As an aside, my floor channel made out of 2 mm mild steel is about halfway in its second full winter and seems to survive so far.

In my heater I can't see the riser from the top but I think it's the same picture as you describe. The floor channel is definitely different from the p-channel and the second picture is looking much more like the chaotic fire fountain as compared to the orderly fashion of the double vortex. The fountain is probably more resistant to overfuelling is my experience.

Maybe it's time to say the floor channel and riser combination isn't necessarily the same as the p-channel and riser. By changing the secondary air provision so drastically the parameters of the riser could be simplified, or so it seems.
So you are running a 6B square riser with half an octagon as high as the port at the bottom. No backsweep in the riser's back as well. This could lead to a new definition of a combustion core configuration.

Thanks for the work, there's still a lot more to learn I'm inclined to think.

[peterberg]

Dec 13, 2017 9:53:32 GMT -8 @yasintoda said:

Does anyone here knows how to calculate combustion efficiency using the data gathered with the Testo (CO,COnd,O2,CO2,TF,Tair) ?

I have a method that gives me 99,996% of mean combustion efficiency.. maybe that's it ?

I've got a spreadsheet from Lopez Labs 8 years ago. Fill in the average temperature, CO and O² plus the weight and moisture content of the fuel and all three efficiency parameters rolls out, combustion, heat extraction and overall.

[Orange]

thanks for the experiment Yasin! What are the temperatures of the cooktop and the bench?

[Deleted]

Thanks Peter, I added the combustion efficiency to my previous post.

I was thinking along the same lines. In your design the triangles in the firebox are approximately 7% of the CSA of the firebox. If we take them out, the height of the firebox must be 2,77 times the base value in order to keep the same CSA for the firebox (I believe that's important).

So in this heater I can take off the triangles and replace them by a flat brick floor of.. 50 mm. Right on, Peter !
The secundary air will still be underneath this flat brick floor.

A small correction : the half octagon is 3B in this configuration

Orange, the temperatures on the plancha are on the previous page. They can get higher than described is the fire is more intense. The top of the bench can get as hot as 70°C on the hottest point. For the next benches I'll try to have the walls of the bench at least half as thick as the top of the bench in order to correct this.

Regards,

[Deleted]

Hi everyone,

That little heater is making me happier and happier !

This morning I tried a firing with the heater much cooler than the previous tests to see how it would perform. In my configuration, it seems that I must close the bypass valve when the flue gaz are at 200°C. Before that temperature, I think that there is not enough draft to ensure proper turbulences.

1st run :
- 7:50 am
- heater was lukewarm (38°C on the front lintel)
- 1st fire of the day
- initial draft before the firing : 5-6 Pa; TF=48°C
- small wood for lightning + wood of uniform size (4-5 cm of diameter)

Mean values :
- global efficiency = 89,5%
- combustion efficiency (calculated) = 97,1%
- TF = 112°C
- CO = 947 ppm
- COnd = 3899 ppm
- CO13% = 974 ppm
- O2 = 13,2%



2nd run :
- 9:10 am
- heater was hot (126°C on the front lintel)
- 2nd fire of the day
- initial draft before the firing : 12 Pa; TF=108°C
- no wood for lightning, only wood of uniform size (5-6 cm of diameter)
- there was a big layer of red coals on all the firebox floor. The top level of coals was just under the top level of the threshold.



Conclusions :

- The best combustion results are achieved when the flames in the "riser" are symmetrical, the double vortex well formed, the front of the flames touching the opposite side of the riser and the flames going up to the plancha and making a 90° turn. Whenever the flames are going out of the first bell it's an overfueling situation.

- Even tough on the second run there was a big CO spike, I'd say that this core is not so overfueling prone. The CO spike happened because I did use very small wood (5-6 cm diameter) for a refueling.. normally I would use logs at least 10 cm in diameter. It's good to know the limits and one must remember that there is no insulating bricks in the riser.

- The biggest problem in my opinion is the amount of red coals at the end of the firing. I'll try that flat brick floor.

- Maybe the overfuelling would be better dealt with if the little bell on the top of the "riser" was a proper bell with a lower exit.

- A video of the best part of the burn :

[Orange]

great results for uninsulated stove! You may want to leave the door open overnight so the stove cools down completely.

[independentenergy]

the problem of the carbons could be attributed to the uninsulated core.
during cooking of foods that require long time the residual coal could be a problem, because to have a hot cooking plate it is necessary to have the flame coming out of the riser, the accumulation of coals could prevent a necessary refueling.
what seems to be in this case is an isolated core rather than a riser :-D.
in any case, great work, congratulations.
New horizons are being discovered

[Deleted]

Hi everyone,

I've made some more runs.

A last run yesterday showed the same behaviour as before : good overall but bad in the end and a lot of red coals.

I took off the triangles on the side of the firebox and the floor channel as well. Both being replaced by a flat brick layer of 60 mm. The heater refused to go under 14% O2 and when I finally succeeded to get it to a higher burning rate it lead to a big CO spike.. bad configuration obviously. I think the port was too narrow in that configuration.
I've also realised that a vertical part in the secondary air is REALLY important -- if there is no vertical part the coals and ashes can fall in front of it, leading to an obstruction to secondary air.


Another run this morning with the floor channel back in place. The height of the threshold has been reduce to 8 cm instead of 12 cm. The idea was to add some primary air to reduce these unburned coals. Results are better now but I need some more trials to see whether or not the results are consistents.

At 23 min I refueled the batch to see how it would behave under 10% O2. That's why there is these little spikes.



Regards,

[Deleted]

Hi everyone,

This afternoon I tried the same configuration as the previous run to check the consistency of the results.

In my previous post I forgot to mention two things that I also changed.

- The little bell above the riser. Now the exit of this bell is 11 cm lower. The CSA of this opening is a little bit more than 1,5 times the CSA of the system.
Visually it is very clear that the flames are forced to stay longer in the bell. I can see it from the black oven door which has a glass. In this run I think it avoided an CO spike thanks to this because in earlier experiments, when I saw the flames from the black oven door it meant a CO spike.

- I took off the small triangles on the sides of the firebox. So now the firebox height is higher than the recommended value.

I know it's bad testing but if the results are not consistents I'll go back to the previous configuration to test things one by one.

Mean values :
- global efficiency = 86%
- combustion efficiency (calculated) = 97,7%
- TF = 148°C
- CO = 570 ppm
- COnd = 2140 ppm
- CO13% = 600 ppm
- O2 = 13,4%



Until now it's really good !

Regards,

[Deleted]

Hi everyone,

Some more runs allowed me to understand some very important things I think.

I tried one more run with the previous configuration and then another one with a different setup : instead of the bell, I tried to mimic a riser with a 90° turn.

Two pictures to show what I mean :





Then I tried a different approach to characterize each core : instead of making three runs to see how it behave in average, I make one long run and I try to push the core to its limits. Basically I try to get an intense fire (under 10% O2 at the very least) and then I check the correlation between CO spikes and a percentage of O2.

For example, on my first setup (no real bell, threshold of 12 cm) the tipping point appeared to be at 9,0% O2 (9 examples, standard deviation of 0.35%). Meaning that whenever the fire is on a clear acceleration (O2 percentage going down) and goes under that fatal 9,0% O2.. chances are very important that there will be a CO spike. The same goes (at the inverse) when the fire is on a clear deceleration and go through this tipping point.

On the fourth setup (real bell, 8cm threshold, flat floor in the firebox) the tipping point appeared to be of 8.5% O2 (5 examples, standard deviation of 0.35%). So the fire can be more intense without negative CO spikes.

On the last setup (riser with a 90° turn, 8cm threshold, flat floor in the firebox) the tipping point appeared to be of 9.4% O2 (7 examples, standard deviation of 0.5%). The combustion analysis :



As of now the best setup is the fourth, so I'll go back to that and tweak from that point.
What's very nice is that in this approach we can go very easily to the tipping point and have a lot of examples to know where were going.

matthewwalker , peterberg that could be a way to characterize the performance of our cores.. a core with a very low tipping point would be best as it is the proof that the core is able to withstand an intense fire without smoking.

Peter, from our previous discussions I'm sure that you've noticed that tipping point in your measures.. what was the tipping point in your standard batch (in the redbell) ?

Regards,

[peterberg]

Relatively speaking, 7% oxygen and lower is the level where it all can go wrong in the best of repeatable runs that I've seen. So I tried to stay above 8% at all times using a straight batchrocket. But in your implementation with a short riser it might be above 9% as the safer bet. The height of the floor channel's threshold in my heater is 60 mm, by the way.