Learning from (other people's) mistakes

I had a visit to the roastery from a fellow coffee roaster this week. I met him at the Roasters Guild of Europe retreat last year in Poland and we connected through coffee and cycling and have become friends. We do some coffee swaps and talk geeky stuff about RoR, refrigeration effects, avoiding crashing etc etc. ...and bikes (he has a Colnago C60.. I used to have a C50…but I digress). to the point.

He’s installing another roaster in his business that operates in a similar fashion ‘heat-wise’ to the Petroncini. I was in the middle of a full days roasting and  had just charged the roaster with my max 6Kg batch of a coffee I’ve roasted a lot. I told him (let’s call him Simon) to pop his head around the back of the roaster and look into the spy-hole to see the burner flame in the burner tube. It’s a tight squeeze and there are some hot surfaces but I didn’t need to tell him that. What I did need to tell him was to be careful of the safety switch. As he bent to look into the spy-hole the power to the roaster went down. Then it came back for a second when he straightened up, then down again as he moved position. I told him to ease away from the roaster and I went through a few checks before engaging the power and the burner.

I’ve adopted the 1’ soak after charge as a default based on some tests I did last year after reading Scott Rao’s blog post. It allows for consistency in pre-heat and charge and my tests demonstrated no negative effects on the cupping table (the opposite in fact in most cases). Instead of the one minute soak Simons FU and burner re-ignition delay turned the soak into 2’30”. I had two choices: Dump the greens out and start from scratch or continue with the roast and hope. I chose the latter.


You can see the two curves with the dark blue the BT for ‘Simons’ roast, and the faded version that it sits under the standard BT curve. Bright red is ‘Simons’ ET with the power glitch and delayed ignition. You can see the RoR curves for both BT (bright blue)  and ET (orange) also. Whilst I continued roasting I told Simon to stick the two batches on the cupping table and we’d see if his batch was useable. I cupped it blind and one was softer, more fruit, cleaner ie. better. I looked at him, he looked back with a wry grin. I’ve roasted almost 200 batches of this coffee with a very consistent profile since it’s always tasted ‘good’ and why change a good thing ? It took an unintentional action to change the profile quite significantly and the result showed for the better. I remember watching a Joe Marrocco video a while back in which he talked about the roasters standard approaches to coffees. He said most of the time we adopt a certain way and it’s not until something goes wrong we really learn, sometimes with surprising results.

Overall the roast took 11'30" instead of 11'00" with perhaps the most significant difference being the lower max RoR. ie a 'softer' rate of change of the BT.

To make sure it wasn’t just ‘roaster’ talk convincing me that the 2’30” soak was better I had some friends help me with a cupping today. They’re not coffee pro’s (although one is a wine pro) and I asked them to give me their thoughts. They’d never cupped before, so we went through the basics (breaking the crust and the two-spoon cleaning technique) and I simply asked which they preferred and why. Simons roast won out. Softer, creamier, rounder.