Roasting Gradient

2010 January 18

In the process of roasting coffee, heat is transferred into the bean first drying it out and then causing several chemical reactions that produce the flavor components. The green coffee bean does not transfer heat very well, due to its structure. This means that it takes time to transfer heat all of the way to the core of the bean and have the necessary reactions take place. Depending on how the heat is applied, the outside of the bean can be much more completely roasted than the center. This creates a roasting gradient. Generally the faster one roasts coffee, the greater the gradient.

One can use a steak as an example of roast gradient. If we have a very hot pan and the steak is thick, the outside will be seared and the inside will be quite raw. This is a very dramatic roast gradient. If we put a steak in a medium oven, it will take much longer for the steak to cook, but the heat transfer will be much more uniform and therefore the color of the steak from outside to inside will be relatively uniform. There is very little roast gradient.

The diagram above is from my beloved “Ukers All About Coffee” written in the 1930’s. It shows two different ways of roasting coffee. One with little roast gradient and one with considerable roast gradient. One can see both in cross-section and in the ground coffee the visual signs of the gradient.

Which is better? When we think about an ideal steak, most people prefer one with some gradient. They don’t want a steak that is nicely seared but raw inside and they are not impressed by a uniformly cooked steak, as it lacks complexity of flavor and texture. Similarly, I prefer coffees with some roast gradient. Excess gradient with an underdeveloped core is quite unpleasant, whereas a coffee that is has little gradient may have clean flavor but lacks complexity.

Roasters that are virtually 100% convection, but have a reasonable roast time (not large-scale commercial “fast roasters”) tend to have little roast gradient. If we use 100% conduction as with stovetop pan, one tends to char the exterior of the bean and under-develop the core of the bean. This is really unpleasant. The coffees I have enjoyed the most have come from drum roasters which are generally 70-80% convective heat and the remainder conductive. This combination, properly executed, seems to give the best balance of roast uniformity and enough of a gradient to provide complexity in the cup.

8 Responses
  1. January 18, 2010

    This is a very interesting post for a few reasons. I myself had this very conversation today about which roasting method I prefer. Having started out in roasting on a fluid bed (hot-air) roaster and progressing to a commercial size drum roaster, I have experience in 2 of the 3 methods mentioned. Outside of traditional coffee ceremonies from Ethiopia, Kenya, etc., I don’t find many things about 100% conduction roasting methods desirable. It is inefficient, inconsistent and doesn’t yield very good results. Hot-air roasting delivers a clean/crisp cup and consistency (with experience and knowledge), but lacks complexity and body. Drum roasting combines these methods and solves many problems that negatively affect 100% convective and 100% conductive methods. I think this is the sole reason why many commercial roasters automatically look to drum roasting as a no-brainer decision. It is interesting that you chose to use steak as a comparable item to the roasting process. Aside from the fact that the maillard reaction affects both beef and coffee during the ‘cooking’ process, they don’t have much in common. What causes us to make the connection between these two vastly different food products? I only ask this because I too automatically used steak to create a metaphor for the roasting process the other day when educating a fellow barista on how to use my air roaster. Anyhow, I really enjoyed this post.

    • January 19, 2010

      Thanks for your comments. I think the steak metaphor works well because it is a common food and the cooking color is so dramatic and observable. Look forward to your future comments as a roaster.

  2. Jason Casale permalink
    January 19, 2010

    I agree with the this gradient information as carl staub founder of agtron tended to agree with the outside being darker than the inside. This the difference between agtroning whole bean then grinding the bean and measuring with agtron outer and inner the difference should be about 10 agtron points or less between outside roast level of the bean and the inside roast level of the bean.

    • January 19, 2010

      We used to use the Agtron system at The Coffee Plantation to check for conformance to spec when we were roasting. At that time we did not do the differential measurements to assess the gradient. I’d like to find a used Agtron to play with. I could then make my own observations regarding gradient and cup quality. Jason, thanks for your comments. I appreciate the insights of a seasoned roaster.

  3. Jason Casale permalink
    January 19, 2010

    I agree that drum roast to me provides the best combination of conductive and convective energy and drum roast seems to provide a more complex bean with resonating structure and body. I also have tried air roasting with small air roasters even home made ones I prefer drum roast as well for the above reasons. The first poster seems to agree with what I always preferred as well thru my own professional and amateur coffee roasting experiences.

  4. January 21, 2010

    Like in many quests if the answer is not in one clear side or another you probably will find the answer in the middle. In coffee roasting this argument of air versus drum have dragged for years and years (Sivetz versus the drums). Can an exceptional roaster be designed as a hybrid? Were all the energies can be applied to the green but also getting enough convection air at the end of the roast to make it clean of smoke contamination?

    • January 24, 2010

      Ron, thanks for your well reasoned comment. My current path is to try my best to create the roaster able to roast almost fully convection or almost fully conductive. This would allow for experimentation as to the timing of each mode of heat transfer within a roasting profile. Furthermore, it would allow one to adjust the relative ratio of heat transfer modes. I think this may be what you mean by a hybrid.

Trackbacks and Pingbacks

  1. Roasting, philosophy or science « This Life's for the Beans

Comments are closed.