Coffee Data Science
Cold Start Espresso
Changing water temperature during the shot
In 2022, I bought a Decent Espresso machine in part because I was more curious to look at temperature effects on the coffee during the shot. I ended up doing many other experiments, and I finally got around to doing some temperature experiments. However, the temperature was slow to change, and many of my one-off experiments were inconclusive. Then I saw a video on cold bloom in pourover, and I thought I would try again.
After doing thermal pre-infusion, I have been able to drop the temperature of my brew water, and I have played around between 60C and 90C, but usually, 60C brews have to run longer and taste like they are missing something in body.
Cold blooming means using cold water for the first pour and letting the coffee bloom. My modification is better called a cold start because I use the same bloom ramp profile, but I use colder water at the beginning of the shot. Cold in my case is 70C so really, it is just colder water than the brew water at 90C.
History of Cold Bloom
Cold blooming is not new, but it has been gaining some attention. I will cite a few sources that are easy to find, but I suspect the roots of this technique go back even further.
Back in 2017, Lucio Del Piccolo made a cold bloom with a moka pot. He mixed water into the coffee grounds before the start of the moka, so the water had contacted the grounds for a bit before the shot. I tried this technique a few years ago, but I didn’t experiment further and don’t remember why.
In 2018, there was a discussion on Home Barista about using cold blooms for darker roasters, which cited Lucio’s work.
Back in June 2023, Vincent at Tales Coffee made a cold bloom video. In the video, he talks about realizing that hot water on citric fruit brings out a bitterness, and so back in 2016, he tried cold blooms for darker roasts. I get the feeling his thought process was also similar to others who like darker roasts.
Hiver van Geenhoven at Chromatic Coffee advocated for cold bloom for robusta, but he threw out the output of the cold bloom as it usually contained the harsher flavors. Samo Smrke has also been investigating cold bloom to better understand what chemicals are being extracted using this type of brewing.
I would be very curious to know how far back cold blooms go in history.
For espresso, I haven’t seen much about it because most machines don’t allow you to change the temperature during the shot, but that is changing as more machines have become computer driven.
Equipment/Technique
Espresso Machine: Decent Espresso Machine, Thermal Pre-infusion
Coffee: Home Roasted Coffee, medium (First Crack + 1 Minute)
Shot Preparation: Staccato Tamped
Pre-infusion: Long, ~25 seconds, 30 second ramp bloom, 0.5 ml/s flow during infusion
Filter Basket: 20 Wafo Spirit
Other Equipment: Acaia Pyxis Scale, DiFluid R2 TDS Meter
Metrics of Performance
I used two sets of metrics for evaluating the differences between techniques: Final Score and Coffee Extraction.
Final score is the average of a scorecard of 7 metrics (Sharp, Rich, Syrup, Sweet, Sour, Bitter, and Aftertaste). These scores were subjective, of course, but they were calibrated to my tastes and helped me improve my shots. There is some variation in the scores. My aim was to be consistent for each metric, but some times the granularity was difficult.
Total Dissolved Solids (TDS) is measured using a refractometer, and this number combined with the output weight of the shot and the input weight of the coffee is used to determine the percentage of coffee extracted into the cup, called Extraction Yield (EY).
Intensity Radius (IR) is defined as the radius from the origin on a control chart for TDS vs EY, so IR = sqrt( TDS² + EY²). This metric helps normalize shot performance across output yield or brew ratio.
Cold Start Espresso
After lots of testing, I found a profile that gives 22% EY in the 1:1 shot ratio. It is called a ramp bloom, and I use a 0.5 ml/s flow rate. This is a typical flow, and the shot is much longer than normal espresso shots, but the flavor is amazing and worth it. I will note that the espresso process has been driven by commercial interests and time efficiency which is not necessarily the best espresso.
So I tried two forms of cold start:
- Cold start until after pre-infusion (After PI)
- Quick cold start that lasts only 2 seconds at the start of the shot (Quick Start)
After PI cold start is a single longer step at 70C:
Quick Start is a very short step at 70C which means thermal pre-infusion happens at 70C, but the water temperature quickly goes up. The mixed water going into the basket reaches 90C at 64 seconds instead of 130 seconds for the After PI approach.
Data
The results were very surprising and pushed me to quickly test a larger data set. This data is over 5 roasts and a few shots per roast. In terms of taste, the difference was huge. It was more pronounced for a longer start (After PI).
Looking at the individual score metrics, the cold start outperforms the baseline greatly.
I split out the data by Quick Start (8 shots) and After PI (9 shots), and I compared the difference to the baseline because they had difference baseline comparisons. They both agree on bitterness, but the longer cold start seems to be better.
When looking at TDS, EY, and IR, all of these quantitative metrics were lower with a cold start. It might be possible to re-optimize this profile by extending bloom time to bring EY back up.
If I plot taste vs EY, this shows a trend I have not seen before where taste is better for lower EY. Typically, taste correlates or loosely correlates to EY.
From a statistical standpoint, the quantitative metrics (TDS, EY, and IR) are statistically lower. For taste metrics, everything but Syrup has a statistically significant difference. This is particularly surprising because syrup generally is tied to a higher TDS reading. It might be possible that a higher number of pairs could see the Syrup have a statistically significant result.
These results suggest one should try using a cold start. This investigation also led me to want to try a few things:
- Improve the cold start profile to get a higher EY like the baseline.
- Try this profile on a lever machine (challenging)
- Try this profile on a Gaggia Classic (fun fact: I already did, and it is possible!).
I wonder how much soaking the puck at the beginning would allow such a cold start to work.
If you like, follow me on Twitter, YouTube, and Instagram where I post videos of espresso shots on different machines and espresso related stuff. You can also find me on LinkedIn. You can also follow me on Medium and Subscribe.
Further readings of mine:
My Second Book: Advanced Espresso
My First Book: Engineering Better Espresso
