Preheating with sous-vide mechanism induces changes in roasting tendency and sensory profile

Preheating with sous-vide mechanism induces changes in roasting tendency and sensory profile

Text – Youngjun Cho
Study participant – Sungjoo Hyun, Austin Shin, Seula Lee, Hyungsoon Jang, Myunghoon Cha, Youngjun Cho
Translation – Jijung Jung


Roasters question if it is necessary to preheat green coffee beans before roasting. In this study, we have simulated the effect of preheating by applying sous-vide method. Total of 9 samples were tested at three different temperature: 50°C/70°C/90°C. At each temperature, green beans were preheated for 10min/20min/30min. Then, each sample was roasted using the reproduction function of Stronghold S7 to reach Agtron 62(high).
Additionally, beans that have not been preheated, that is, stored at room temperature, were also roasted as the standard to make a comparison. The samples were compared by performing the cupping method.
As a result, it has been confirmed that homogenizing the condition of green beans by preheating at an appropriate temperature helps to add a positive effect on expression of flavor and balance of taste.
However, excessive preheating rather breaks homogenization of green bean condition and affects flavor. In further studies, we would like to find the optimal roasting profiles and points that can lead to the utmost taste at the most favored preheating temperature found in this study.


Whether in roasting or extraction, temperature is always one of the biggest issues. Particularly, roasters have conflicting opinions on whether or not to preheat the raw beans before roasting. Thus, I wanted to find out if the process of preheating indeed affects the result of roasting, and if so, which temperature and duration provides the most effective result.


In this study, we used a sous-vide water circulator (sous-vide machine) to create a similar preheating environment of putting green beans on the hopper. The preheating temperature was set to 50/70/90℃, which corresponds to the turning point(TP) during roasting. Samples were roasted using Stronghold S7 to reproduce the roasting profile of the room temperature sample.
For each sample, water activity, water content, and density of green been were measured. Water activity is a measure of the activity of free water, which is pure water remaining in green beans. It also indicates the activity of microorganisms in green beans. Generally, normal range is 0.2 ~ 0.6Aw. When the water activity is too low or high, it may affect the shelf life of roasted coffee beans. In addition, the total amount of water, which is generally recognized as a measure of freshness, measures the total amount of free radicals and binding water (binding water: water that has been bound with protein, sugar, etc.). Lastly, we will also measured the density of green coffee bean to determine if physical changes had occurred as a result of preheating. Using these three indicators, we can determine if moisture profile of green coffee bean changes during preheating, and how this change affects the roasting profile and sensory difference.
The tools used in the study are:
  • Green Bean: Costa Rica Las Lajas Black Honey
  • Sous-vide Machine: Anova Precision cooker
  • Roaster: Stronghold S7 (*Used to reproduce the same roasting profile)
  • Measuring tools
    • Water activity: Aqualab Pawkit (error range: ±02 Aw)
    • Water content/density: SINAR Bean pro (error range: water ±3 STD% mc; density ± 1.0g / ℓ)
  • Sensory evaluation
    • 4 skilled panelists who have completed calibration attended the cupping
    • Mahlkonig W401 Grinder was used to medium grinding degree (Level 5 out of level 1 ~ 9)
<Figure 1>
Experimental Method
The experiments were carried out in the following order.
  1. Preheating green coffee bean
    • Packing: subdivided into 1kg each in heat-resistant zipper (can be heated up to 100℃)
    • Temperature classification: RT / 50℃ / 70℃ / 90℃
    • Preheating time for each temperature set: 10 minutes / 20 minutes / 30 minutes
    • After preheating, samples were rested for 30 minutes before roasting
  2. Measuring
    • Immediately after sous-vide: water activity, water content
  3. Roasting
    • Roasted at the same profile of RT sample using Stronghold S7
  4. Sensory Evaluation
    • Calibrated with RT samples
    • Cupping was performed for 1 cup of each sample
 <Figure 2>


How are the physical properties of green coffee beans changed?

Table 1. Changes in physical properties of green coffee beans
The overall water content tended to increase when preheated for a longer time. Particularly, there was an interesting tendency in water activity. Preheating at a low temperature (50 ° C) did not significantly affect water activity. However, as the temperature and preheating time increased, water activity changed to a greater extent. In addition, there was an inverse relationship between water activity and total water content. As water activity increased, total water content was decreased. This indicates that preheating indeed affects green beans’ status. For instance, activation of free water during preheating led to evaporation of water in green beans.
On the other hand, the density of green beans was not significantly changed after preheating. However, it is interesting to note that among all temperature sets, preheating at 70°C led to the most stable density of green beans. In fact, such effect was also reflected to roasting and sensory performance.

2. What are the effects of preheating on roasting and sensory performance?


Roasting was performed using Stronghold S7. First, beans stored at room temperature (RT) were roasted and its roasting profile was saved to repeat with other samples.
During roasting, there was no significant change in the profile itself, except the turning point (TP) was slightly advanced than that of the RT sample. It was surprising to find out that the reproducibility was much better than what we had expected. The reproducibility of total time was about ± 1 second and a discharge temperature was ± 2 ℃. The Stronghold S7 has a mechanism to reproduce the roasting profile based on RoR (Rate of Rise) and discharge beans based on the end temperature. This study proved that Stronghold’s technology regarding these processes is quite advanced.
The sensory evaluation was carried out for the 10 kinds of samples in a cupping manner. Rather than focusing on detailed flavors, the evaluation was focused on how overall balance and preference changed in comparison to the RT sample. Panels have freely discussed about overall tastes and were asked for the preference of each sample. The results are as follows.

Table 2. Sensory evaluation result

As shown in the table above, samples preheated at 70°C were most preferred. When the preheating temperature was too low or high, a negative reaction was observed. It was interesting to find out that when preheating at a lower temperature of 50°C, longer preheating time resulted in a better taste. Whereas, when preheating at a higher temperature of 90°C, a shorter time resulted in a better taste. As a result, it can be assumed that when preheating, there is an appropriate range for both temperature and time.

Conclusion and Future Progress

From the above results, it can be concluded that a certain degree of proper preheating helps stable roasting and flavor development. Then, is it always better to preheat before roasting? In this study, we used sous-vide machines and could control other variables in closed condition. However, in most of the cases, preheating is processed in opened hopper. This might endanger green beans as a double-edged sword because there certainly is a risk of over-evaporating moisture out of the beans.
The conclusion of this test is summarized as follows.

  1. Homogenization of green beans by preheating at an appropriate temperature helps to develop positive flavor and balance of taste.
  2. Excessive preheating rather breaks the homogenization of green beans and affects flavor.

In the next study, we will further develop this research by focusing on the preheating temperature of 70 ℃. We would like to find out the optimum roasting point at this most preferred and stable temperature.
In this study, beans were roasted to about 62 (high, Agtron standard). In the next study, I would like to tune the roasting profile more precisely and find out which point results in the best development and expression flavor.

Youngjun Cho

저는 prism coffee works의 창업자로 한국에서 전문 바리스타 대상의 이벤트 및 교육을 진행하고 있습니다. 일본, 중국 등의 아시아 커피 시장은 물론, 콜롬비아 등의 산지를 방문하며 커피에 대한 다양한 지식과 인사이트를 쌓았습니다. 이러한 지식을 바탕으로 커피에 관련된 다양한 정보를 이해하기 쉽게 전하고자 노력합니다.