This is really perplexing. However, the root cause is that roast gases are being created at a higher rate than being cleared.
I wonder at what batch size the problem would go away. A combination of adequate fan speeds and smaller batch size would be my next guess to try.
I now feel like getting one of these air quality meters and testing the air quality near where I roast outdoors just for my own curiosity
Do you think so, Gaby? At this point, there are two AC Infinity S6 Pro fans running at 407 CFM each. They are able to move the air pretty efficiently and one of them is right above the exhaust of the roaster. When I was checking the ventilation, I put some chaff and dust on a piece of paper and put it underneath the duct hood. It sucked everything up rather quickly.
I donāt know if you already address this, but there is a third fan vent from induction heater, see photo. Could this be the issue since your air meter is right there?
I am clutching at straws Abouna. You got us all stumped 
Yes, I am certainly aware of it, but when my setup was in the garage, this is where I had the monitor.
Do you vent the induction fan separately?
Have you tried removing the filter altogether? Maybe a straight shot out would work better. But maybe youāve already tried that.
No, I do not vent the induction vent, but my set up is right next to my garage door. I just use a floor drying fan above the chaff vent, instead of pointing the bean vent under the unit I point it towards the garage door, and I point a cooling fan on the on/off switch side to blow under the unit out towards the garage door. I do not have air quality meter to compare to your readings. Maybe collect a reading at different points on the unit to narrow down location, which you probably have already done. Your set up is to notch, so the issue is baffling!
I seeā¦ sounds like a good setup. Certainly, I regret moving from my old setup in the garage to this new setup.
OLD
NEW
It looks nicer and had the promise of better ventilation with a vent hood going outdoors, but so far, none of that promise has materialized and it has become a much worse experience than when it was in the garage. My mistake, but I suppose one has to take risks to learn what works and what doesnāt.
Maybe the room is just too small? Then again, maybe it isnāt. Personally, I wouldnāt go lower than 2000 cubic feet, as far as room volume is concerned. But that might or might not be the issue. There could also be some air coming from different directions that moves the smoke before itās sucked in. With those measurements, you should easily be able to see the āairā, which is never a good thing. At those levels everything should be enveloped in smoke, or at least a haze of some sort. Maybe your air quality meter is at fault. When troubleshooting, always check everything, no matter how stupid it sounds. The last thing Iād try, although this one isnāt really going to help, but itās worth a try, is to get the roaster under the hole and leave just the fan and the shortest route possible hose to vent hole. If itās close by, Iād also try putting the hose out the door, just to see the difference. Iād also measure the rate at which the air is coming out on the outside, just to see if something is choking it. This you can do without operating the roaster. Maybe borrow an air flow meter from someone if you donāt want to buy it. While youāre at it, measure the air flow everywhere. First fan, filter box, second fan, If you see a drop somewhere, thatās where you intervene. Thatās about all that comes to mind. Iām just trying to help, but without actually being there itās a bit difficult. I hope youāll figure it out. Best of luck.
There definitely seems to be something going on with your setup. I run S6 and S4 fans ducted outside and also track my raw PM2.5 values during a roast. This is from a 500g light-medium roast on Sunday.
My air quality monitor only tracks in 1 minute increments so its not a responsive as yours, but you can see I am getting nowhere close to the particulate levels you are seeing.
One thing to check, are you sure your sensor is reading accurately? Iām not sure how you can really verify this without a second sensor.
Thank you both. I donāt have a way of verifying the air quality monitor other than to say that it reports GOOD air quality when Iām not roasting and the numbers are all fine. When it reports HAZARDOUS or goes berserk, it usually coincides with dry end or first crack, so it seems to be responding to some change in the air related to those roasting events. For this reason, I think it is working accurately, but something is going on with the roaster or the ventilation.
Today, I will borrow a thermal camera (like a Flir) to see if I can identity a leak. I also ordered a cheap air flow meter to measure output air flow and help identify whatās going on with the fresh air in the room, although to be honest, Iāve never used one and donāt know the first thing about it. But I can learn. Finally, I ordered new P100 cartridges for my respirator to ensure Iām getting the best protection in that environment.
Today, I fired up the roaster again and roasted a smaller 600g batch at a lower preheat temperature (265Ā° C) and higher fan speeds (F4-6). Sadly, there was no change. The air quality deteriorated at dry end and shortly before first crack. Dropping the beans sent the monitor readings berserk, but I didnāt see much smoke around me.
During the roast, I tried using a thermal imaging camera, but didnāt know what to look for. Everything that should be hot looked hot in the image. Nothing unique stood out to me.
Iām close to giving up on the Aillio Bullet, but will consider hiring a ventilation specialist to come and take a look first. When I ordered the Bullet, I believed that as long as one took reasonable precautions, it would be safe to use in homes, small cafes, microroasteries, etc. I think Iāve taken excessive precautions and this has not been my experience unfortunately.
I canāt help but think that everyone who uses one of these indoors (or even in a garage) needs to get an accurate air quality monitor and see whatās going on.
Abouna, do you mind sharing a link to the air quality meter you are using? I am curious to see what it measures.
My pleasure! It is this one:
https://www.amazon.com/dp/B08DNJ8L55
The specific variant I have is the LKC-1000S+ 2nd.
Thanks for sharing man.
I appreciate your effort for keeping a safe work environment.
I hope that you will find a solution and keep roasting.
Thank you Abouna. Great looking unit. I pray you figure the cause and find a solution soon.
Smoke naturally travels upwards. Just a suggestion based on my trials and errors, but the straighter and more vertical you can make the ducts, the more effective it will be at extracting the smoke. Forcing the air through a duct that went up then down and around a couple of bends pushed more smoke back into the room for me. I had to really fuss with the angles of the ducts, adding extra fan power didnāt help. Even a small downward angle made things worse.
I know you probably donāt want to throw more money at this buuuuut, go read the 1 star reviews on amazon for that air quality sensor. There are quite a few reports of major variation between units and odd behavior, including reporting very bad air quality when they shouldnāt be.
You could try getting an air quality sensor that uses the laser scatter detection type sensors (I couldnāt determine what type of sensor your unit uses) and cross compare the results. I have several sensors from Air Gradient (the DIY ones) that I have been quite pleased with (Indoor Air Quality Monitor). My sensors all agree with each other when they are in the same area and I have also tested them publicly available air quality sensors and generally trust the results.
Thanks, @rearbarbers0h.ZaI6, I appreciate the perspective. Itās tempting for me to blame my existing monitor, but one thing I canāt deny is that it has been remarkably consistent in showing high PM2.5 and PM10 readings at specific times over maybe 8-10 roasts now. The problem always starts around dry end and gets really bad immediately before first crack through the end of the roast. Moreover, I took my air quality monitor inside the house and kept it next to me for a good hour; it showed perfectly normal readings for indoors without any aberrations that would make me think there is an accuracy issue.
But who knows? Maybe a second monitor would be a good idea to compare. I donāt necessarily want to acquire more possessions, so perhaps Iāll ask my neighbors and see if I can borrow a device. What would be nice is a A/C powered device that is always on and connected to WiFi with a companion app/cloud service that continuously monitors and records air quality. The AirQuality One you sent seems to fit the bill quite nicely.
It is definitely grasping a straws, but my theory would be something like this: the current monitor is measuring something, but its sensitivity is wrong/broken. So once pm2.5 starts to rise (even my levels are clearly rising during the roast) the monitor goes from zero to 360+pm2.5 instead of the correct value somewhere around 30-50.
I live in the PNW and when we have wildfire smoke, the pm2.5 values will get into the high hundreds and you can SEE the smoke in the air. I would think you would be able to visible see the smoke if it was at that level in your room.
