Hey guys, I need some help with something that has been bugging me lately. We run an MCR 6kg with typical batch size of 10lbs. Since October-beg we have been roasting hard due to the season. But lately I noticed something rare... in short, I am seeing some abnormalities in the ET reading curve and the ET ROR curve that worsens as I roast more. - see images below. Some facts:

• The three images attached are a sequence of the same coffee, same batch, 3 consecutive days of roasting.
• I last cleaned the exhaust stack on Friday, Dec 12th (so fairly clean) and the exhaust fan on Sat Nov 28th.
• This issue does not happen on the first roast of the day. It is always smooth (I assume because the cooling fan is off)
• The spike on the ET curve happens as soon as I make adjustments to the exhaust fan (i.e. increasing from 15 to 18) and worsens when the cooling fan shuts off from the previous batch
• This issue started happening in October (after roasting about 1,000 lbs straight) and what I did then (from MCR advice) was to fully clean the exhaust stack and cap which I did and the issue basically went away
• I guess my issue now is that I am confused why if I just cleaned the exhaust stack last week, why is this issue reappearing so quick?
• Am I missing something? Is it related to the chaff collector?

Any help on this issue would be great

Update on 12.19.25

• Did a full evaluation of the condition of the exhaust vent cap using a cell phone and telescopic pole. Scary findings.
• Basically the top row of "window slots" of the vent cap was completely clogged despite believing that our cleaning process was actually doing the job
• One thing to clarify... our vent cap is 21ft above the roaster due to the design of the building and makes it very complicated to reach the cap from the outside.
• Our cleaning process didn't really account for the geometry of the vent cap vs the diameter of the exhaust pipe. Therefore in summary we were "cleaning" the vent cap but not really as our cleaning tool was not reaching the top row of slots on the cap.
• Following actions were taken:
• Inspection: Smartphone attached to telescopic pole to visualize interior and identify blocked slots.
• Cleaning: Drill-mounted spinner brush applied at low speed, increasing as needed; frequent visual checks to reach all slots; vacuum and wipe down debris.
• Verification: Post-cleaning roasts showed normalized ET and ROR behavior - perfect roasts
• Conclusion: We truly didn't follow MCR instructions to deep dive into the condition of the exhaust vent cap. We didn't understand the full geometry of the cap and the cleaning procedure was not doing the job.
• All roasts are back to normal after the deep cleaning of the cap and slots.