The pressure applied during tamping has minimal influence on espresso extraction. We wanted to provide scientific evidence for this finding from daily practice. For this purpose, with the support of a food technologist, we set about the work and extracted, measured, and varied the tamping pressure of 150 espressos. In this article, we present the results to you.
What is Tamping?
Tamping is the pressing of ground coffee powder in the portafilter. A tamper is used for "tamping," which consists of a handle and a metal base that is pressed into the portafilter like a piston. The metal base corresponds to the inner diameter of the basket. If the distance between the basket wall and the tamper diameter is too large, some of the coffee powder will not be pressed sufficiently.
By pressing the ground coffee, uniform water distribution is made possible. The brewing water should encounter the same resistance from the ground coffee surface at all contact points. If the ground coffee is not compressed enough in places or if air pockets form below the pressing point, there is a risk of channeling. This can be prevented through careful preparation of the ground coffee, vertical compacting, and even tamping.
Instructions: Prepare espresso step by step (articles and videos).
Three classic tampers with a 58mm diameter.
Research Question: What Influence Does Tamping Pressure Have?
Two sub-aspects were examined. First, we wanted to find out whether a change in tamping pressure affects the flow time. Second, we examined whether the tamping pressure has an effect on the extraction yield and thus on the concentration of dissolved coffee particles in the cup.
To obtain the most precise possible result, work steps and methodology were strictly defined.
Materials
Rancilio Specialty, gastronomy espresso machine with precise flowmeter. However, each espresso extraction was re-measured with the Acaia Luna.
Etzinger etzMax, espresso grinder with integrated scale. However, all grinds were reweighed and repeated if there was any deviation (see methodology).
Puqpress, automatic tamping machines for gastronomy use. Tamping pressure adjustable.
VST Refractometer, for measuring TDS in percent, i.e., the concentration of dissolved particles in the beverage in relation to water.
VST Syringe filter attachment to filter solids from coffee solution and measure only the dissolved particles with the refractometer.
VST Syringe
Etzmax espresso grinder and Puqpress
Methods
The espresso grinder is set to an output of 18.5 g. Grinding results between 18.4 g and 18.6 g are used.
The Rancilio Specialty espresso machine is programmed for an output water volume of 45 ml and a brewing temperature of 93 °C.
The corresponding amount of coffee is ground directly into the portafilter. The coffee is then leveled with the fingers and tamped with the Puqpress. The portafilter is inserted into the machine and extraction is started.
The amount in the espresso cup is measured on a scale. After extraction, the portafilter is removed and knocked out.
A flush shot is pulled on the espresso machine.
After extraction, the espressos are set aside to cool. After cooling, they are stirred and the TDS value is measured with the refractometer.
The espresso is drawn into the refractometer via syringe and syringe filter attachment.
This process is repeated 50 times each for tamping pressures of 10 kg, 15 kg, and 20 kg.
The data is evaluated for flow time and TDS values using a one-way ANOVA with a predetermined error probability of 5%.
Results: Flow Time and Tamping Pressure
In the variance analysis of flow time, no significant difference could be found. In other words, whether pressed with 10 kg, 15 kg, or 20 kg, the results in our measurements did not vary significantly due to pressure.
Effect of tamping pressure on extraction time.
| 10 kg Pressure | 15 kg Pressure | 20 kg Pressure |
| 26 | 20 | 27 |
| 21 | 23 | 25 |
| 23 | 26 | 29 |
| 24 | 27 | 28 |
| 23 | 25 | 24 |
| 24 | 26 | 27 |
| 25 | 26 | 23 |
| 24 | 29 | 25 |
| 26 | 20 | 26 |
| 31 | 24 | 27 |
| 25 | 23 | 25 |
| 24 | 23 | 21 |
| 25 | 22 | 26 |
| 27 | 23 | 21 |
| 24 | 30 | 24 |
| 24 | 24 | 21 |
| 25 | 22 | 24 |
| 27 | 21 | 25 |
| 23 | 21 | 28 |
| 31 | 24 | 22 |
| 25 | 24 | 22 |
| 24 | 25 | 24 |
| 22 | 30 | 23 |
| 27 | 20 | 25 |
| 22 | 21 | 22 |
| 25 | 23 | 25 |
| 25 | 21 | 21 |
| 24 | 21 | 22 |
| 24 | 27 | 21 |
| 23 | 22 | 24 |
| 27 | 23 | 22 |
| 21 | 22 | 30 |
| 24 | 24 | 21 |
| 25 | 21 | 25 |
| 24 | 25 | 29 |
| 25 | 24 | 26 |
| 23 | 24 | 23 |
| 23 | 23 | 24 |
| 23 | 23 | 24 |
| 22 | 23 | 23 |
| 21 | 22 | 24 |
| 26 | 24 | 26 |
| 23 | 25 | 24 |
| 25 | 22 | 20 |
| 24 | 23 | 27 |
| 20 | 23 | 22 |
| 24 | 26 | 25 |
| 24 | 22 | 26 |
| 22 | 21 | 22 |
| 23 | 24 | 20 |
| Groups | Count | Sum | Average | Variance |
| Column 1 | 50 | 1212 | 24.24 | 4.51265306 |
| Column 2 | 50 | 1177 | 23.54 | 5.60040816 |
| Column 3 | 50 | 1210 | 24.2 | 6.08163265 |
| Source of Variation | SS | df | MS | F | P-value | F crit |
| Between Groups | 15.4533333 | 2 | 7.72666667 | 1.43133301 | 0.242301 | 3.05762065 |
| Within Groups | 793.54 | 147 | 5.39823129 | |||
| Total | 808.993333 | 149 |
| Fk(0.05) | 3.05762065 |
The test statistic F is smaller than the critical F-value, so there is no significant result. The actual P-value exceeds the set error probability of 5%. Thus, no significant difference can be found between the groups.
Results: TDS Values and Tamping Pressure
In the variance analysis of the TDS values, no significant difference could be found either. The concentration of dissolved coffee particles was not substantially influenced by different tamping pressure.
Effect of tamping pressure on concentration of dissolved particles.
| 10 kg Pressing | 15 kg Pressing | 20 kg Pressing |
| 8.53 | 7.77 | 8.15 |
| 7.99 | 7.94 | 8.15 |
| 8.03 | 8.02 | 8.16 |
| 8.13 | 8.03 | 8.27 |
| 8.01 | 8.08 | 8.2 |
| 8.01 | 8.06 | 8.08 |
| 8.1 | 8.07 | 8.08 |
| 8.05 | 8.26 | 8.01 |
| 8.2 | 8.12 | 7.96 |
| 8.24 | 7.96 | 8.03 |
| 8.17 | 7.98 | 7.99 |
| 8.15 | 7.99 | 8.03 |
| 7.91 | 7.85 | 8.11 |
| 8.09 | 8.02 | 8.02 |
| 8.09 | 8.88 | 8.02 |
| 8.09 | 8.12 | 8.01 |
| 8.11 | 8.11 | 7.93 |
| 8.15 | 7.99 | 8.04 |
| 7.96 | 7.99 | 8.04 |
| 8.3 | 8.05 | 7.99 |
| 8.25 | 8.15 | 7.99 |
| 7.99 | 8.2 | 8 |
| 8.07 | 8.28 | 7.99 |
| 7.99 | 7.81 | 8 |
| 7.81 | 7.84 | 7.97 |
| 7.8 | 7.92 | 8.01 |
| 7.93 | 7.9 | 8.01 |
| 7.94 | 7.87 | 7.66 |
| 7.86 | 7.97 | 7.8 |
| 7.93 | 7.88 | 7.89 |
| 8.06 | 8.02 | 7.99 |
| 8.08 | 8.02 | 8.02 |
| 7.99 | 7.95 | 7.92 |
| 8.04 | 7.7 | 7.99 |
| 8.05 | 7.89 | * |
| 8.13 | 8.01 | * |
| 8.2 | 8.11 | * |
| 8.19 | 7.95 | 7.91 |
| 7.98 | 8.02 | 7.96 |
| 7.94 | 8.07 | 7.7 |
| 7.82 | 7.94 | 8.01 |
| 8.08 | 8.04 | 8.08 |
| 8.08 | 8.05 | 7.92 |
| 8.07 | 7.96 | 7.83 |
| 8.06 | 7.97 | 8.03 |
| 7.88 | 7.78 | 7.97 |
| 7.88 | 8.01 | 7.97 |
| 7.93 | 7.81 | 7.91 |
| 7.97 | 7.73 | 7.65 |
| 7.96 | 7.85 | 7.69 |
TDS values. *these fields lack measured values
| Groups | Count | Sum | Average | Variance |
| Column 1 | 50 | 402.27 | 8.0454 | 0.01824167 |
| Column 2 | 50 | 399.99 | 7.9998 | 0.03163465 |
| Column 3 | 47 | 375.14 | 7.98170213 | 0.01635356 |
| Source of Variation | SS | df | MS | F | P-value | F crit |
| Between Groups | 0.10546284 | 2 | 0.05273142 | 2.37573217 | 0.0965805 | 3.058928 |
| Within Groups | 3.19620383 | 144 | 0.02219586 | |||
| Total | 3.30166667 | 146 |
| Fk 0.05 | 3.058928 |
| Fk 0.1 | 2.33979948 |
The test statistic F is smaller than the critical F-value, so there is no significant result. The actual P-value exceeds the set error probability of 5%. Thus, no significant difference can be found between the groups.
Raw Data and Measurements from the Tamping Test
The raw data and further information can be found in the Excel sheet. If you can extract any other interesting findings, we're curious to see the results.
Download Results - Tamping Project
When analyzing the data, a linear correlation was found between the following datasets between the extraction time and the TDS values obtained, which was determined using the Bravais-Pearson correlation coefficient:
- Tamping pressure 10 kg: Extraction time and TDS: weak positive correlation → r=0.5068
- Tamping pressure 15 kg: Extraction time and TDS: strong positive correlation → r= 0.6854
- Tamping pressure 20 kg: Extraction time and TDS: weak positive correlation → r= 0.5423
where: |r|= 1 → perfect linear correlation
|r|= 0 → no linear correlation
Pressing Pressure, Taste, and Additional Information
Within the scope of the present test, the taste of the espresso was not examined. Therefore, no statement can be made about this. However, in the sense of the findings from the test, we interpret that varying pressing pressure in the range we examined has no or only marginal influence on taste.
We also cannot make a statement based on our test as to how pressure lower than 10 kilograms affects extraction. Based on studies and from partners and our own experience, we recommend tamping with at least 6 kilograms of pressure.
Discussion
When investigating whether tamping pressure affects the extraction time of espressos, no influence could be demonstrated. However, this does not mean that no influence exists, but only that in the test model used, no influence could be demonstrated. Confounding variables include, for example, inaccuracies in preparation and the insufficient precision of the test machines used.
The equipment used consists of machines designed for professional use in gastronomy and not precise laboratory equipment.
However, in practice, there are far greater confounding factors in the preparation of an espresso than in the experimental procedure used.
Thus, in practice, there is no reason to assume a major influence of tamping pressure on extraction, since there are a multitude of inconsistent parameters affecting the extraction process. In this case, tamping pressure is only one parameter among many.
---
The tamping test was conducted by Michel Aeschbacher in collaboration with a food technologist who wishes to remain unnamed. Our thanks are certainly due to you!
We are open to further research partnerships and are happy to use our materials and expertise to develop insights around the topic of coffee.
