Well, I also enjoy my espresso made fresh on the day of the roast.

Of course it depends on the origin of the beans. I like a very light roasted African bean.

There is interference from degassing, but then it is sort of like drinking a Beaujolais Nouveau.

The flavor does evolve a bit as time goes on, but the oxidization of the beans really masks that and makes the coffee unpalatable by the second week. It helps a lot to flush the coffee storage jar with nitrogen and keep a few dessicant packages in the jar, to slow down oxidation and remove excess moisture.

In a drum roaster, if you roast a smaller batch at the same profile as a larger batch, the drum temperature will be lower (you are transferring less heat to fewer beans in the same time, so you don't need as high a source temperature).

For cupped coffee, the lower the drum or source temperature the better, since one gets the aromas and origin flavors undisturbed by anything else. For regular brewing and espresso, one can overdo the low temperature roasting, since one likes some roasty flavors as well.

The difference between high and low source temperature roasts (or MET, maximum environmental temperature) is simple in concept. Low Met roasts are more even, bean inside to bean outside, so the flavor range on the taste wheel, from enzymatics to distillates is more focused, whereas with a higher MET roast, the flavors are more scattershot.



In reality, the differences aren't as schematic. Bad low Met roasts can be totally flat, bad high met roasts can have all sorts of indefinable skunky and sharp flavors

I dislike extremes in my espresso, so I like low MET roasts. However, there are roasters (mainly large scale commercial oners) who will use a heat blast to add some dark roast flavors to a few of their medium roasted espresso brands.

Jim thank you for that interesting perspective. It does show itself in what I've been doing lately with my roaster, pre-heating the drum with grill at a lower grill temp and I'm seeing faster roasts but done at much lower temps (from 700F down to 510F-520F ranges) still with a relatively large batch. The roasts progress very nicely and the espresso has an all new and improved cup character.

With too high a MET and too large a batch seems to suck the oils out much more which will mean an earlier death. More fractures too past 2nd. I you like it when it's best the first day, you will be amazed at how good it is when it's best at 3-5 days.

As usual, great insight Jim! You guys now made it clear what to look for in a smaller batch/lower ET roast: even more/clearer origin flavor/aroma and more required rest. Thanks!

The extra flavor/aroma is definitely a good motive for me to go ahead experimenting with 200g (or less!) in the next few weeks. Time for questions:

1. We talk about high/low ETs, how are they defined?
   
   1. As absolute numbers? e.g. 550F?
   2. Relative to bean mass? e.g. +50F?
   3. A combo of A&B depending on roasting phase (e.g. drying ET 350F=OK AND pre-C1 ET +50F=OK, etc)?
2. Is "with the same profile, the lower the ET, EDIT:the better the roast the clearer the origin" a good, solid, guideline to go by?
3. EDIT: Is MET just the maximum an ET gets or something else?
4. HB members: what are your ETs? Results?

This is getting interesting

MET is defined as the hottest temperature inside the drum or roast chamber, i.e., the highest temperature to which the beans are exposed. I don't use a Hottop, so I'll let one of the HT mavens tell you how to measure it.

I'll take a stab at how I look at MET and ET. First to say that considerations are very roaster specific. When I designed my hybrid homebuilt my goal was to get the maximum evenness of heat exposer to each bean or maximum energy transfer efficiency. On one end of a spectrum would be 1 bean suspended in the middle of a chamber with even heat distribution. On the other side would be something like a fry pan with a couple inches of beans and no stirring. In my first example my ET and MET would be the same at whatever moment in the roast and the overall roast MET would be the highest temp. the ET was during the roast. In the second fry pan example the MET during the roast would be the surface temp. of the fry pan and the ET would be almost useless to even consider. Obviously in the pan roast(without stirring) the bottom layer would quickly char and the top layer would take forever.
In a drum (where the heat source is outside the drum) The MET at any one moment is the hottest surface drum temp. on a non-perforated drum and the hottest air temp. entering the drum on a perforated one. The effect of the MET will have to do with drum speed (rpm), size of drum vs batch size, efficiency of any vanes in the drum, etc. So a small non-perforated drum with a slow rpm, lots of beans and no vanes will be at one end and a perforated drum with a faster rpm, small amount of beans and good vanes will be on the other. With say a popper the MET will be the temp. of the air as it first comes in contact with the beans and the ET would be considered by the efficiency of energy transfer which will be effected by the amount of air flow and the churning abilities and the depth of the bean mass.
So MET and ET are important to consider but must be considered and adjusted within your roasters design.
Here is the BRR (best reaction ratio) according to Carl Staub
"The best cup characteristic are produced when the ratio of the degradation of trigonelline to the derivation of Nicotinic Acid remains linear. The control model of this reaction ratio is a time/temperature/energy relationship. The environment temperature (ET) establishes the pyrolysis region for the desired chemical reactions while the energy value (BTU) and system transfer efficiency (STE) determines the rate of reaction propagation and linearity of Nicotinic Acid derivation to degradation of trigonelline. Because green bean density varies dramatically, under any given ET / BTU / STE format, the reaction distribution will vary. it takes longer to obtain comparable uniformity for a higher density bean. Monitoring the bean temperature offers a good method of approximating the reaction distribution during this phase of the roasting. The ideal environmental temperature, ET, for best reaction ratio, BRR, is from -401-424 degrees F, with 405 degrees F as the default value. The BTU required is determined by the systems transfer efficiency, or ability to impart the energy to the charge mass."
farm

Great post, Ed! Thanks!

If I'm to apply this to the HT, the MET would be the highest temp of the drum just before it touches the beans OR the highest temp of the air closest to the beans; which ever is higher. Right?

On the Staub quote, I must admit it's beyond what I can comprehend at this stage --maybe because of all the scientific-sounding names. But I understand it's conclusion to be: in the most efficient heat-transfer roasters, MET should be between 401-424. If this understanding is correct, I could tinker with HT and see if it's possible to sustain that low of a MET while maintaining a good profile.

hmmm, now all this might explain why well-controlled air roasters are said to be able to produce the maximum origin flavors/aromas

The Staub quote needs a good deal of decoding. He is discussing details of a proprietary roast control system without giving away any trade secrets. This makes the passage fairly incomprehensible.

Ratios of easily detected chemicals which breakdown at different speeds during the roast have been proposed as more exact ways of determining the degree of roast than color or temperature. The area is so heavily trade secreted, it's hard to know what's going on. But industrial roasters do have provision for various chemical "noses" or even real time mass spectrometers for roast control. Needless to say, this is very expensive.

So what is Staub talking about? 405F to 425F cannot be the ET at the end of any roast that is going city or full city, since the bean temperature exceeds this. The big clue here is that Trigonelline breaks down completely in the middle of the first crack. So I'm guessing he's saying that this environmental temperature is best for the ramp to the first crack and that his system stays in this range for this phase of the roast.

You'll need to go hotter to finish the roast.

Alright, I dug into my roasting notes and found the following roasts. These seemed to have very similar profiles (in terms of BT) with the main difference being ET (and eventually MET). #2 had WAY better flavor/aroma, although done at a higher ET and required almost no rest; just like I described on the 1st post. I'm not sure what to make of this. I'm still interested in doing ~200g (or less) but looking at these makes you wonder

#1 fancy Brazil, 225g, stopped at ~440F, not much extra drum pre-heat (click for a larger version):


#2 same fancy Brazil, 300g, stopped at ~440F, so much extra drum pre-heat (click for a larger version):