Article on controlled temperature incline/decline in espresso extraction-Rancilio XCelcius

#1: Post by **TomC** » September 12th, 2016, 7:21 pm

Today an article was posted on Reddit about what appears to be a sponsored (biased?) study looking at tight temperature controls applied to espresso extractions. It looks like a fascinating read, so I'll share it here for others to view and comment if they like.

Fascinating snippet: In the washed arabica samples, it was noted that the increasing temp profiles improved perceived aromatics and decreased astringency, which I found interesting.

https://www.dropbox.com/s/cgy5npx1n8qhr ... n.pdf?dl=0

September 12th, 2016, 8:40 pm

One way to interpret this study is that the ever-increasing temperature of a La Pavoni group head is really a desirable feature

!

September 13th, 2016, 2:43 pm

I had a chance to play with a three group one last year. According to the barista who was there to evaluate it, rising temperatures was Rancilio's advice at that time. He had set up the three groups to flat line, drop, and rise; and preferred the flat line and drop equally, and the rise least.

Would it be crass to suggest that finding cases where rising is best would sell the machine more readily, since it is not a profile found in conventional machines? On the other hand, this is a really sweet machine for any kind of tailoring temperature to coffee; so I doubt they need to stoop to complete BS.

September 13th, 2016, 3:13 pm

another_jim wrote:Would it be crass to suggest that finding cases where rising is best would sell the machine more readily, since it is not a profile found in conventional machines? On the other hand, this is a really sweet machine for any kind of tailoring temperature to coffee; so I doubt they need to stoop to complete BS.

I'm no expert but the study looked like they were trying to find benefits of the rising temperature profile. The were able to find benefits according to the language of the article. It's not clear to me how the jargon-speak they used applies to real world espresso. I agree that having temperature profiles to choose from seems like better than not having it. They implied that different coffees might be best using different temperature profiles.

#5: Post by **jonr** » September 13th, 2016, 5:45 pm

... the ever-increasing temperature of a La Pavoni group head is really a desirable feature

Is there data to show this for the temperature of the water delivered above the coffee during a shot? I pull the entire shot volume of too hot water into a much cooler group just before a shot - my guess is that the water temp drops as the water loses heat into the cooler group and that this more than offsets the slow, gradual rise of the group temperature.

September 13th, 2016, 5:58 pm

Thanks for sharing Tom! Interesting read.

Based on the temperature profile I've seen, HX machines also tend to have a temperature incline profile especially when pulling ristrettos - if inclining temperature is something of desire.

I had a detailed read on the paper yesterday. It was pretty all over the place to say the least - some part of the discussions don't quite agree with the data. The data itself doesn't show a consistent trend when using different beans. The taste test contradicts with the data completely. The only takeaway I could get away confidently are the methods that they're using, and that taste experience can't be quantified by hard data.

I was also expecting to see some temperature graphs for the three profiles - but they're not there (how do they know if they've executed the temp profile correctly and was not affected by flow/coffee?).

#7: Post by **jonr** » September 13th, 2016, 6:19 pm

The use of 88-93°C gradient of temperature contributes to increasing the quality of all espressos investigated.

Sounds like this conclusion is too broad. Maybe they determined something more like "... increased the quality with the parameters we tested".

I have no doubts that a flat temperature profile is often not optimal - and that the taste differences are often significant.

https://www.ranciliogroupna.com/single- ... experiment
https://www.coffeegeek.com/forums/espre ... nes/527217

#8: Post by **h3yn0w** » September 13th, 2016, 6:37 pm

samuellaw178 wrote:Thanks for sharing Tom! Interesting read.

Based on the temperature profile I've seen, HX machines also tend to have a temperature incline profile especially when pulling ristrettos - if inclining temperature is something of desire.

I had a detailed read on the paper yesterday. It was pretty all over the place to say the least - some part of the discussions don't quite agree with the data. The data itself doesn't show a consistent trend when using different beans. The taste test contradicts with the data completely. The only takeaway I could get away confidently are the methods that they're using, and that taste experience can't be quantified by hard data.

I was also expecting to see some temperature graphs for the three profiles - but they're not there (how do they know if they've executed the temp profile correctly and was not affected by flow/coffee?).

HX machines can also be coaxed to produce a relatively flat or declining profile as well depending on flush and wait times.