Studying ground size distributions
Recently, a collaborator provided me with a flat-burr grinder with a continuous grind mode to more efficiently fulfill ground coffee orders. Once it was clean and properly adjusted, it was time to assess whether the grinder was suitable for my intended use. For me, this step is crucial, as the particle size in ground coffee greatly influences the extraction of soluble compounds when preparing the beverage. A more uniform grind reduces the presence of excessively fine or coarse particles, which improves extraction consistency and can minimize undesirable flavors such as bitterness or unbalanced acidity.
Using the somewhat slow but new and well-calibrated conical burr grinder I use in production as a reference, I proceeded to compare the grain size distributions obtained with both grinders. For simplicity, I will limit the scope of this article to each grinder's optimal setting of around 300 microns, suitable for espresso, for example. We will identify each grinder with a combination of letters and numbers to keep the trademarks anonymous. Thus, we will call my production grinder the KA58 and the grinder under study, at its most similar setting, the AS08.
Samples of the same medium-roast specialty coffee (City) were analyzed. The KA58 grind produced particles with a mean diameter of 313 ± 81 µm, while the AS08 grind generated particles with a mean diameter of 307 ± 89 µm. Both data sets fit the normal (Gaussian) distribution well, although the reference set showed a slightly thinner distribution (less dispersion), indicating a somewhat more uniform grind. The implications of these differences for coffee sensory quality and extraction efficiency are discussed.
Materials and methods
Coffee and grinding
Specialty Arabica coffees from Costa Rica, Caturra and Catuai, were used, with a City roast profile. Two commercial grinders calibrated to a suitable grind for espresso were used:
KA58 Mill (conical grinding wheels): Stainless steel conical grinding wheels, 40 mm diameter.
AS08 Mill (flat burrs): Stainless steel flat burrs, 58 mm diameter.
Each grinder ground 25 g of the same coffee described above. The first 10 g were discarded to avoid cross-contamination. Of the remainder, only a small sample extracted with a laboratory spatula was used.
Particle size analysis
The analysis used a software program developed by astrophysicist and coffee enthusiast Jonathan Gagné and the methodology recommended on his blog .
Each analysis was replicated five times per sample, using different areas of the photographs obtained from each profile. The mean diameter (D[4,3]) and standard deviation were recorded.
Results
| Windmill | Average diameter (µm) | Standard deviation (µm) |
|---|---|---|
| KA58 | 313 | 81 |
| AS08 | 307 | 89 |
Both distributions were approximately normal (see Fig. 1), with skewness coefficients close to zero (0.05 for KA58, 0.08 for AS08).
Figure 1: Histograms of the particle sizes analyzed for both millings.
The greater dispersion in AS08 suggests a somewhat less uniform grinding, which is visually reflected in the greater width of the distribution curve (Fig. 2).
Figure 2: Theoretical Gaussian curves of both coffee particle size distributions.
Conclusion
For a grinding setup with similar average particle sizes (313 µm vs. 307 µm), the difference in standard deviations (81 µm vs. 89 µm) indicates that the reference mill produces a slightly more uniform grind than the test mill. The difference is small and can only be detected by repeated measurements [...], since it falls within the range of uncertainty attributable to the Gaussian distribution fit.
Previous studies have linked grind dispersion with sensory differences perceived in the cup. No detectable differences are expected between the two grinds, although the theoretical result indicates that the KA58 produces a more controlled extraction and is therefore the preferred grinder for espresso-ground coffee orders.