Repeatability of a roast profile is the ability to reproduce a similar roasting result when the same setup is repeated on the same system. This isn't a built-in property of the roaster but the result of controlling many input factors at once, not just the numbers on the control panel.

Repeating a setup isn't the same as repeating a result

Entering the same heating power, drum speed, and roasting time doesn't guarantee two batches will follow the same heat curve. The starting temperature of the roasting chamber, ambient temperature, and residual heat left over from the previous batch can all make the bean's thermal response different even with identical input settings.

Inputs that need to be locked or recorded

To assess repeatability fairly, you need to fix or record charge weight, raw-material moisture, bean size and density, the machine's starting temperature, and the rest period between batches. If any of these variables changes without being recorded, a difference between two batches could be mistaken for unstable machine operation.

Placing multiple curves on the same frame of reference

The clearest way to assess repeatability is to overlay several temperature-time curves from batches of the same raw material on the same axes. Doing so lets you observe the average curve and how much each batch scatters around it. A batch far outside the usual scatter range is a signal worth checking, while small, consistent variation between batches may just be acceptable natural variation.

Post-roast output determines the meaning

Two similar heat curves still need to be checked against post-roast results: mass-loss percentage, bean color, and sensory evaluation if available. If the curves are similar but the outputs differ significantly, the cause may lie in the raw material or in an uncontrolled variable in the process, rather than in the machine's ability to repeat.

Designing a test to measure repeatability

A systematic way to check this is to roast a series of consecutive batches with the same raw material, the same charge weight, and the same configuration, while fully recording the input variables mentioned above. The number of batches needs to be large enough to distinguish random fluctuation from a systematic drift, such as the roasting chamber gradually heating up across consecutive batches in a day.

Natural variation versus process deviation

Perfect repeatability shouldn't be expected. Small differences between batches can come from the natural variation of agricultural raw material, which is never perfectly uniform bean to bean. What deserves more attention is systematic deviation — consistently drifting in one direction across several consecutive batches — since this usually signals an uncontrolled process factor, such as residual heat in the roasting chamber.

From repeatability data to an operating standard

Once an acceptable range of variation between batches has been identified, this data can become the basis for an operating standard: a suitable starting-temperature range, the necessary rest time between batches, or a charge-weight limit to keep the heat curve within the desired range. This is how tracking repeatability shifts from a one-off test into part of a regular quality-control process.

Practical conclusion

Repeatability isn't a fixed property of a roaster but the result of controlling raw material, equipment, and operating process together. Assessing repeatability requires looking at both the heat curve and post-roast results, with enough batches to distinguish natural variation from systematic deviation before concluding on the stability of a roast profile.

References

• Hernández, J.A. et al. (2007). Analysis of the heat and mass transfer during coffee batch roasting. Journal of Food Engineering, 78, 1141–1148.
• Batch-Scale Simulation of Heat and Mass Transfer of Coffee Roasting in Spouted Bed Roasters (2025). Beverages, 11, 162.
• Münchow, M. et al. (2020). Roasting Conditions and Coffee Flavor: A Multi-Study Empirical Investigation. Beverages, 6, 29.