Development Time Ratio


The Golden Ratio

I occasionally see discussions online of Development Time Ratio (DTR) and a recent discussion on the Barista Hustle Facebook Group motivated me to write about this at length.

I wrote The Coffee Roaster’s Companion in mid-2014. At the time, most specialty roasters were not yet using Cropster or other software, underdevelopment or and baked roasts were the norm (baked still is, but more on that another time), and roasters were routinely discussing development time but no one had ever, to my knowledge, referred to development time ratio. Likewise, I was not aware of anyone proposing that the ROR should always decrease during a roast (it should).  After twenty years of roasting beans on a variety of machines, I had many such ideas to share. I couldn’t discuss them in too much detail, though, lest the book become confusing to readers.  As it is, I receive a fair amount of mail from confused readers seeking clarification.

At the time of writing CRC, almost all curves I’d seen from other roasters and new clients looked something like the following curve, which I’ll call “The Crash and Flick

 big crash and flick, 23%

Given that backdrop, it was important to propose some basic guidelines that anyone could follow easily, with any roasting machine.  I didn’t intend the book to dive into roasting theory to the depth that I would go into with a client; my goal was to offer a foundation on which roasters could build ideas and discussions about roasting.  Prior to publication, there was precious little open discussion of roasting among professionals.

I designed the guidelines with the typical small specialty roaster in mind: someone using a classic drum roaster, charging with >60% of capacity, and dropping the coffee between the end of first crack and the beginning of second crack.  I think that described the vast majority of professional roasters at the time.

I chose 20%—25% because:

– They’re round numbers that work well with most medium and light/medium roasts that dominated specialty coffee at the time.

– Of the 20,000+ roasts I’d tasted over the years for which I had access to bean-temperature data (several from a few of the best roasting companies in North America), probably 18 out of the best 20 happened to have a DTR of 20%—25%.  That’s not “proof” of anything, but it’s a correlation that gave me the confidence to make the broad recommendation of 20%—25%. Please note that 20 out of 20,000 represented the top 0.1%; these were not just “good” roasts, they were extraordinary. My goal has always been to study and practice the methods that produce exceptional coffee, not just to roast good coffee.

These days I still prefer a 20%—25% DTR for most roast levels between the end of first crack and the beginning of second crack. Of course, I’ve had excellent roasts with DTRs below 20%, but none have cracked (no pun intended) that top 20.

Many roasters have opined that they don’t often like their roasts at DTRs above 20%. Some have posited such sweeping generalizations as “higher DTRs create less body”, which is rubbish. In my experience, roast degree and development, not DTR, are what determine body.  If a roast’s DTR is 20% but not well developed, it likely will lack body.

DTR is not a guarantee of development; it’s simply one of many indicators of how a roast progressed. It’s a handy rule of thumb, and one that can be broken successfully, but like most rules, one probably shouldn’t break it before mastering it.

For example, if your roast curve looks like the Crash & Flick curve shown above, and you blame the roast’s baked (flat, cardboard, straw-like), underdeveloped, or roasty flavors on the DTR, you’re missing the point that DTR is just one piece of a complex puzzle.

The Exception To The Rule

There’s one major exception worth noting:  if the ratio of your burner capacity to batch size is high enough, the target DTR range should probably decrease. For example, it would be reasonable to use a DTR of 15% when roasting 10kg of coffee in a Probat UG22.  However, it would be more difficult to roast a high-quality 20kg batch in the same machine with a 15% DTR. This relationship will partially explain for some readers why they can easily develop coffee using a much lower DTR in a sample roaster than they can in a production roaster.

This Is How We Do It

For those who protest that it’s not possible to achieve a good roast with a  20%—25% DTR in a reasonable roast time, here are a few examples.  In future blog posts I’ll explore how to achieve such roasts in virtually any machine, and with almost any coffee. Until then, feast your eyes on these lovely curves 🙂

This is a 22% DTR from Ben Toovey  of Genovese Coffee in Melbourne.  I’ve been working with Ben through email for a short time and he’s rapidly achieved great control over his ROR curves. This curve was roasted on a customised 5kg STA Impianti, an indirectly heated roaster. 


Ben 22%

Below is a lovely 21% DTR roast from client Jackie Suiter of Kona Coffee Purveyors in Honolulu. Jackie roasts on a Giesen W15.  I’ve shared Ben and Jackie’s curves to illustrate similar, excellent curves with quite different roasting machines.

smooth 21% DTR

NB: Please do not try to mimic these temperatures and times in your own roaster; it is very unlikely the numbers shown in these curves are appropriate for your machine, probe, and batch size.

The Bottom Line

Roasting is quite complex. I hope roasters carefully test new methods over the course of hundreds, if not thousands, of batches, before drawing conclusions, or critiquing those methods online. If we embrace that approach as an industry, we will move forward more steadily.  I did not intend the recommendations in my book to be the final words on roasting; I meant them to be the first words of an open, much-needed, professional-level discussion about roasting. I’ve started this new blog in hopes of furthering that conversation. Welcome and thanks for reading.

77 thoughts on “Development Time Ratio

  1. Yes yes yes! Great blog and I am very happy, that you are start bloging. I have every book from you and I am really happy that I found this blog. Keep up the good work and I cant wait another postst.

    Also can I have a question? Could you write something about dark roasting? Many of my clients like the taste of coffee which starts about 2.nd crack. Its very hard to be master in darker roasts. To produce really sweet and chocolaty coffee and not to burn it.

    • Hi Peter,
      Thanks for your comment. I’ll consider what to write about dark roasting…I honestly don’t have many clients who ask for help roasting past the beginning of second crack, but I have a little experience with it. Roasting carefully gets pretty challenging in second crack.

      • Although I hate dark roasts and simply reject such client requests, I feel (I say “feel” because I have almost no experience in dark roast) that mastering dark roast (i.e. yes, hitting that sweet spot) is crucial for specialty coffee roasters because it’d improve their light/medium roasting abilities. I do not know if that makes sense, though 🙂

    • Hi Erhard,
      There will be some coffee-related links at times, but these links were mostly just for fun.
      Please note that underlined words allow you to see their definitions when you scroll over them.

    • Hi Dan
      While I’m not privy to all the relevant info I’d need from the competition (machine, probe, location, etc), I know that most successful competitors (wisely) use rather small batches, often 50% or less of the machine’s stated capacity. If you read the section of my post “The Exception To The Rule”, you’ll see that a high gas power: batch size ratio allows you to finish roasts faster with less concern about development.

      • Thanks. Is there any flavor benefit to roasting smaller batches? (Other than not wasting a $150/lbs ninety plus). Shortened maliard, more delicacy?

        • Hi Dan,
          It’s easier to develop smaller batches (in a given machine). That’s the main benefit. There may be a small benefit as well, on some mmachines, to being able to use less flame and therefore keep the drum surface cooler all roast.

          I would advise against adjusting roast curves based on ideas like “shortened Maillard”… I understand the intention but it’s not so simple and it’s not a useful approach in my experience.

  2. hey scott
    we just won 2. place on the swiss championchip in brew coffee this year. (damm, we could made it)
    hence, we were the only swiss roastery in the top ranks.
    thank you for your consulings in our place. black&blaze enhanced a lot in our understanding of puzzling all this factors;
    times, ror decrease, top-ror in start phase, dtr, loadings, and last but not least: momentum of heat-transfere in relation to batch size and entrance temperatures.
    its very truh what you were writing: the batch size in relation to the drum capacity and the total roast time matters a lot about dtr.
    my personal feeling:
    espresso or drip coffee discern in its dtr; the sweetnes or acidity aspects sensory matters different by the preparation of the beverage.

  3. A quick word to say welcome in the world of blogging 😉 and how I think it’s more than wonderful for us to have you sharing your knowledge and opinions. Thanks a lot! Merci beaucoup de Montréal.

  4. Hi Scott, thanks for the post – for keeping the discussion going! Glad to see it happening 🙂

    If you have time: Thoughts on managing evaporative cooling during first crack? From memory your recommendation is to never increase the heat input only decrease; for high grown washed coffees if we follow this our RoR will crash, if we maintain a higher energy input earlier in the roast we’ll get facing. 66% batches in a Diedrich IR-12 (which appear to have relatively low thermal mass, high burner output).

    Cheers Scott. Thanks again for sharing so openly! Glad you’ve started a blog.


    • Hi Adam,
      With apologies, I’m still trying to work out the boundaries of the advice I’ll offer on the blog. At the moment, questions like yours cross a little too far over the line of consulting for me. I will offer that in my opinion, 66% is too much for a small Diedrich for the quality and control I want over a roast; I never exceed 50%. And of course, if you’re getting facing, your drum is much too hot.

      • Seems I have a related problem with my Turkish AKS1 roaster. With 1 kg load I must apply too much heat to
        get a 10°C Delta BT in the second phase ( from end of drying to C1 ) Mostly I can achieve only Delta BT of 8°C
        and it takes too much time.The effect on the endphase is a too high Delta BT 7-8°C to reach a desired colour and end temperature in a recommended time. Besides there is also danger of facing.
        So I should reduce the load and increase air flow from the beginning?!

        • Hi Erhard,
          I’m sorry, but not only is this too complicated to try to figure out here, but to answer your concern, I’d have to know much more information. But if you’re roasting 1kg in a 1kg roaster, your first move should almost certainly be to decrease your batch size.

          • Dear Scott,
            Of course you should know more, but it is clear that I can get more flexibility when reducing load first.
            Thank you anyway! Your advices helped me a lot.

  5. Great to see you writing again Scott, I hope all is well with you and keep the information flowing! Big thanks and hope to see you in Oz soon.

    • Hi Nico!
      Thanks for saying hello. I should tell you that the delicious, small batch of Colombian you gave me 1.5 years ago that had the very low DTR was the inspiration for my connecting high power:batch size ratio to lower DTRs.
      I hope to see you this January in Byron

      • Ah yes I think it was the Caturra lot by Daniel Sanchez. A memorable coffee! Would love to catch up, see you in January, email me dates so we can make time to surf and slurp. Peace. N

  6. thanks Scott for starting this blog,
    At the start of DTR, those beans will fall under your 20-25%. The crack profile along with the crack structure change will then determine the DTR and effects from, every other bean from then on. Obviously some will be much less. So then we get into something like the average DTR, ADTR, and outlier considerations.

    • Hi Ed,
      Thanks for commenting. If I understand you correctly, then yes, I totally agree. That’s part of why I stay on the safe side with my recommendation of 20%–25%.

      I hope roasters recognize that bean-temperature curves and RORs are just proxies for roast progression and also involve a lot of averaging, both of noise in the data, and also of what’s happening in different beans in the same batch.

  7. hi scott
    first of all is an honor and a pleasure and from COLOMBIA we send many successes, it was very exciting aver known your book and even more exciting to read this blog, DTR, I really agree with you encourage discucion to enrich our knowledge about the production … high quality coffee.
    But that way we can give farmers understand clearly and simply (which today are marketed directly their roasted coffees) during sign remove intermediaries from the commercialization

  8. Dear Scott, I have a quick question regarding the “length of the first crack”. Lets suppose that, for two roasts, you have exactly the same total roast time (say, 12 mins) and development ratio (25%) but they differ in the time first crack actually lasts. Say, in case of batch X, crack starts at 9:00 and ends at 10:30, but in case of batch Y, crack lasts until 11:00. What kind of differences should we expect between these two cases?

    • Hi Burak,
      Thanks for your comment. I don’t know if we can generalize about that… length of FC can be related to overall roast time (slow roasts will generally have longer cracking phases), uniformity of bean size, batch size, ROR at FC, etc. There’s too much going on.

      I would say that for a given coffee and batch size, a slower crack is good and bad. Good, because it allows you to control the ROR a bit more, bad because it implies a less uniform roast. BUt that’s just speculation. Sorry to not be more helpful.

      • Actually, you’ve been very very helpful. The comment about the relationship between the length of FC and the “uniformity of the roast” has been very inspiring for me. I will think more on that. Thanx!
        If you allow me, I would like to ask an even more bizarre question: Why is crack sometimes very silent or almost inaudible. I had this with a Brazilian natural yellow bourbon (I know you hate naturals, sorry 🙂 Is it because of something I do wrong or is it because of the structure of the bean?

        • Hi Burak,
          I have some ideas on that regarding moisture and density and ROR but I don’t like to speculate too much publicly. But I do not want you to think that a soft crack is “wrong”.
          PS Had you asked about the crack of a washed coffee, I may have been more inclined to answer 😮

  9. This is interesting as I just started roasting on a 30 kg drum at 45% capacity and noticed that at 20% DTR I was getting a lot more development in the bean than at 23% DTR on my previous roaster, previous drum was at 60% capacity.

    Looking forward to more insights from you! Thanks

  10. Nice post over here. Keep doing it Scott !

    By the way is that possible to have more informations about coffee density influence in roasting please ? It will be a pleasure to read and learn more about it !

    • Hi Guichard,
      I don’t have much to say about density other than what I’ve written in my book. But I’m sure it will come up a few times as a topic in this blog. Thanks

  11. Great post, Scott!

    I have been trying to figure out how better to roast with my humble set up of a modified popcorn popper and intuition as I cannot afford a drum roaster or electrical doodads to add onto my current set up. I’m sure that as I read more from you and pay attention to my analog thermometer more, I can figure something out to increase quality until I can have nicer equipment.
    I look forward to what you have to say about fluid bed roasting!

  12. Hi Scott,

    I would like to say that your book, The Coffee Roaster’s Companion has been invaluable to us as we are in the startup phase as a small commercial roaster. It’s great that you have started this blog and furthering open discussions on roasting.

    I have a couple of questions if you able able to answer them in this blog:

    1. In the book, you mention that air/fluid bed roasts should usually be in the 7-11 minute range. What are your thoughts at present?…is this still the right range to be aiming for? Any particular reason to be aiming for one end of the range rather than the other? (apart from batch size / burner size considerations; which end of the 20-25% DTR range one is aiming).

    2. I like the roast curves you posted in this article blog…something I’m aiming for but not quite there yet. One issue I’m grappling with is how long one should wait before making temperature adjustments when the bean temperature deviates from the desired curve (I’m using an air roaster, so can make quick adjustments in temp; Artisan roasting software and 5 second intervals for temp recording).

    Looking forward to future posts.

    Many thanks.

    • Hi Graeme,
      Thanks for the kind words.
      1. the time-range recommendations in the book are just ballpark targets. I don’t have a strong opinion about exact roast times in your machine, and can’t really recommend a particular range without knowing much more about your setup
      2. I think you should move your interval to perhaps 15 seconds. As for how/when to respond to deviations in roast curve, that’s another one of those questions that’s impossible to answer without a lot more information. (Which goes beyond the scope of my role here, sorry)

  13. Hey scott
    Steg here!
    thanks for the info great to see you getting a online presence
    looking to reading the posts to come
    everything is going swimmingly over the pond here
    are you back in aus anytime soon?

  14. Hello Scott
    Your blog is great.
    I would like to know if you do any consultation via mail?
    and also if you do What is the fee structure.

  15. Thanks for blog Mr Rao, I apreciate it.
    We see the fast roast at the roasting competion like 8 to 10 minutes , Do you thing enough time for development,even usıing half batch with high energy..maybe they do in 6 minutes in the future,, not sure how effect to on proteins and acids high energy in less time

  16. Fantastic article!!

    I have so many thoughts on this subject. After pouring over years worth of probe data from my own roasts, seems that some of my own theories parallel these (and some from your book) to a large degree.

    I’m going to read this a couple of times, and let your thoughts sink in.


  17. Dear Scott,
    Could you give some ideas about DTR in relation to the origin of the coffee?
    I mean – slow start, fast finish versus fast start, slow finish.
    As an exampel: how would you roast an Indonesian coffee and how an Ethiopian?
    Does your book deal with these questions ?

    • Dear Erhard,
      I don’t think of it that way. Please remember, DTR is just a guide or general range to aim for, assuming a roast is not extremely light or extremely dark.
      I think the best analogy is extraction percentage… if I were to recommend a 19-21% extraction range with a given grinder, your extraction target may vary very slightly for different coffees or brew methods, but it would be unreasonable for me to try to pinpoint what’s best for someone else, given the tremendous number of variables at play.

      • Dear Scott,
        I understand, it is very much about details, but on the other side there a guidelines for different types of coffees and I am sure experts like you could describe the organoleptic effects when changing a roast in different phases in time or ( and ) in temperature,air flow etc.. Hobbyists like me are always on search for simple rules, such as: African coffees could be better with a fast start and a slow finish. Of course within borders. It is not about exact advice but overall directions. And yours answers as all in coffee could be very subjectiv and challenging. The same is not good for all!
        So I wish that you once would write a book about your favorite coffees and how they were roasted. I am sure you will find a very “coffeiny ” title for this edition. Anyway thank you for having me answered ! I am very proud of it.

        • Hi Erhard,
          I think most roasters assume such knowledge is out there, but it isn’t. I know many roasters who seem to think they have mastery over flavor…. and they may have some success with some coffees when attempting to manipulate flavors, but given that I’ve never experienced or heard of a roasting company that is fantastically consistent in its roasting, it stands to reason that no one out there has the kind of knowledge or control that would be needed to offer the types of rules you’re seeking.

          Whenever you hear people offer fast, general rules like “Africans require a fast start and slow finish”, they’re usually speculating and basing their recommendation on a limited amount of experience with a limited range of variables, and they’re usually incorrect. A company with a particular machine, coffee, and batch size may find a certain pattern works well for them, but that is very far from being able to translate their success to a general rule.

          Even my simple recommendation to shoot for a 20%–25% DTR (a very broad range) has stirred up an immense amount of controversy, an even greater amount of misunderstanding, and countless people distilling 80 pages down to one sound bite: “have a DTR of 20%–25%.” In this case, almost all of my critics are in the 2% of the industry that roasts very light (they may not see it that way because they tend to focus most of their attention on their light-roast peers, but when one steps back and looks at the coffee industry as a whole, the 15%-DTR third-wavers roasting close to Nordic preferences probably represent about 1% of coffee roasted worldwide.) But if they had to offer a DTR range works well for 90% of roasters, they’d probably have to choose 20%–25%, too. Maybe they’d even have to recommend a higher top range, given the large amount of coffees out there even darker than Starubucks.

          My point is that everyone wants a set of general rules, and I’ve already done my best to offer those. I will have a few more such ideas to offer in the next year or two, but it will take time to confirm their validity, because I’ll have to test them on a variety of equipment before feeling comfortable making recommendations.

          Thanks for your comment.

          • Dear Scott,

            your detailed comment makes me happy and it proves that all roasters, hobbyists and professsionals
            are still dependent on their own experiences and must not seize to struggle for consistency.
            Nevertheless I am looking forward to your offerings in the future.
            Best wishes

  18. Hi Scott

    I was the one initiating the debate on BaristaHustle – asking for people experiments with the DTR 20-25%. Thanks for elaborating 🙂

    Especially, I like that you emphasise that DTR “is just one piece of a complex puzzle”. And “DTR is not a guarantee of development; it’s simply one of many indicators of how a roast progressed.”

    My reason for asking was that I too often heard roasters focus solely on DTR to determine when to stop the roast. Expecting a coffee to taste good as long as ROR is nicely falling and DTR is between 20 and 25%. I have tasted coffees roasted like that – most didn’t taste good. So I was getting tired of this narrow focus, so thanks for clearing up 🙂

    When you say “Of the 20,000+ roasts I’d tasted over the years (…) probably 18 out of the best 20 happened to have a DTR of 20%—25%.” I have been thinking: it doesn’t work the other way around: that as long as DTR is between 20-25%, the coffee will taste good. I understand it as “icing on the cake”. As you write here: “is just one piece of a complex puzzle”.

    Also very interesting that you explain the difference depending on batch sizes (sample roaster size vs. micro roaster 15-30 kg, I presume) – and the %-fill up of the capacity of the roaster. I myself have used batchsize to optimize the taste – apart from development time, onset of FC and level of preheat. And doing the exploring for each new bean.

    Again thanks for elaborating. And thank you for venturing into these things in the first place 🙂

    Roasting is so wonderfully complex – and a never ending journy of exploration 😀


  19. Hi Scott,

    Quick question-please could you give me an example of a burner/batch size ratio? What metrics to you use?

    Just trying to work out what the X to 100g would be for me.



    • Hi Henry,

      I’ll do my best. I’m sure some scientist reading this one day will tell me I’ve made some errors, but hopefully I’m approximately correct. Most calcuations have been rounded slightly.

      An old Probat L12 has about 100,000 BTUs/hr (106,000 KJ/hr). So if you were roasting 15lbs (6.82 kg)—a good choice for that machine—you would have a ratio of:

      6700 btu/hr per lb
      15,500 kj/hr per kg

      Per-hour-per-lb is a little awkward, so I’m just going to call it btu/lb and kj/lb here.

      according to probat:

      Their sample roasters offer 3,600 “btu per drum”… I’m going to assume they mean btu/hr… in that case, your ratio when roasting 100g is 16,000 btu/lb or 38,000 kj/kg. That puts your sample roaster at approximately 2.5x the power/weight ratio of the production machine roasting 15lbs!

      I’m pleased you asked this question, as I’d never bothered to do such a calculation, and just assumed in my head the typical sample roaster’s ratio was about 3x the typical production machine’s ratio.

  20. Two questions:
    1. How do I calculate the Development Time Ratio with my roast curve?
    2. Is there a specific temperature/time that you are looking at to measure the 20-25% DTR?


    • Hi Sam,
      Development time ratio is simply the % of total roast time that occurred after first crack began.

      For example, if first crack began at 9:00 and total roast time was 12:00, DTR would be 25%.

      No, there is no specific time or temperature involved in this.

      I know you haven’t had a chance to read the book yet, but when you do, it will all make more sense.

      • So I do all my roasting on a Quest M3. When I hit 1st Crack I typically increase air speed to prolong/draw out 1st Crack and drop the temp because of the exothermic reaction. Is there anything else I should be doing to hit a longer development time without going into 2nd Crack? I’m always worried about going too far. Thanks!

        • Sam,
          Questions like this are a little difficult to answer, because actions like “increase air speed to prolong 1st crack” do not work the same on all machines. On some machines, opening the air late in a roast will accelerate the roast. Beyond that, of course, you should be lowering your gas if your goal is to slow the roast.

  21. Hey Scott,
    Cheers for smashing so many coffee myths for me along the way. On the subject of Diedrich’s i’d like to ask a question. on Diedrich’s with with the stepped airflow adjustments of ”Closed” “half” and “full” would you advocate playing airflow gymnastics(offsetting airflow adjustments with burner changes) and say start with it “closed”, at mallard move to “half” and at FC “full” or would you pick one setting say “half” and stick with that throughout the roast?

    And I’d love to talk to you more about Loring roasters if you’d like as I’ve heard you havemixed opinions on them however we normally are hitting most of your “comandments” on our kestrel35 and really enjoying the results



    • Hi Anthony,

      The answer depends on the year your Diedrich was made… Stephen Diedrich told me two years ago that they had increased the fan size in their smaller machine quite a few times over the years ( I think he said 9 times). I find that with the newer machines 100% air is more than necessary, and 50% is plenty. But on some of the older machines, you’ll need to use 100% at times. I’ve never been a fan of charging a roast with very little air… I don’t think it’s a terrible practice, but it seems more of a workaround to compensate for a small burner (again, the older machines had much weaker burners) and to allow for simultaneous roasting/cooling, rather than because it’s the best roasting method.

      If you’re happy with your Loring, there’s not much to say. I think the machine has many good points, but would prefer not to offer a full review on the blog.

  22. Scott, What’s your impression of the Jopers? I’ve noticed they are quite powerful, and when looking at their BTU/hr in their fact sheets it is quite high for their rated batch size. I am currently roasting on a W6 Geisen, and am quite happy with it after separating the cooling fan and re-probing the machine. I’ve found 4kg batches develop well around the 10-minute mark, my DTR aren’t quite up in that 20-25% range. Everytime I’ve pushed the DTR further (with great ROR curves and same target roast degree) I lose a bit of the clarity in the coffee. Is it just my preference for cleanliness over the caramelized tones? or am I missing something. Do you have a modified approach for nordic roast degrees, that results in better-developed coffees? **side note**I have always loved the coffees of the roasteries that you’ve work with, and thanks for the blog.

    • Ben,
      I have a lot to say about the Jopers, but this isn’t really the place to do it. the short version:
      -they’re single drum, though Jay Endres has told two of my cients they have a double drum, in order to get sales. The guys in Portugal confirmed the smaller machines don’t have double drums (the 25kg may, but I’ll believe it when i see it now)
      -I do not agree with the clarity vs. caramelized opinion in relation to DTR. there is something else going on there.
      -nordic approach: that’s a little too far over the “consulting line” for me to write here. Sorry.

      Glad you like some of my clients’ coffees.

    • As I am sure you are well aware, Giesen are advertised as one of the few roasters that can roast full batch sizes with ‘control’. It reads as if your preference is 4kg batches to achieve development and clarity many of us reading Scotts blog desire. Does that vary significantly among origins? Or is that your sweet spot? Do you have any experiences you can share with a full 6kg batch?

      Yo Scott – love the books and blog! Keep rockin!

      • Jason,
        I put no stock in what the machine manufacturers say. Giesens are better at roasting a larger % of capacity than most other machines but I still would not put a full batch in if you want optimal results.

        Yes, the max “quality and control” batch size varies a bit with origin, density, bean size, etc.

        I have no positive experiences to share with a full 6kg batch (presumably you’re referring to a W6.)

        I’d like to share two insights with you regarding the use of the W6 in roasting competitions:

        -On the one hand, almost all of the top batches in the comps are done with quite small batch sizes. Not every competition batch will be great, but smaller batches is a sound strategy.

        -A few years ago, Cropster began publishing the top few roast curves from each comp… in the (I think) first year, most of the ROR curves looked like roller coasters, up and down, up and down, and it was impossible that they were accurate. The funny/frustrating thing about it for me was that on some online forums, several people noted the crazy RORs and used them as evidence that my proposal of a always-smoothly-declining ROR was obviously BS.

        Obviously, the winning RORs were no such thing: anyone with experience on several machines would recognize that the RORs looked as they did because Giesen had installed a (very fat and slow) bean probe much too high on the face plate to read such batches with any accuracy. The probes were likely immersed in beans at times and out of the beans at other times, and when the air temp in the drum was much different from the bean temp, the ROR would reflect the volatility of a probe reading mostly beans then mostly air, then mostly beans, and so on.

        The point is that context is critical when interpreting bean-probe data. If you look at a curve and want to interpret it with any confidence, you’d need to know, at the least, the probe type, size, and location, the machine, and batch size. And even that’s not really enough.

        And thanks for the kind words.

  23. Hi Scott,

    Thanks for much for this blog as well as all your work before hand, i’m gutted I won’t be able to meet you in Adelaide! As I gear up to make the shift from barista to roaster your works have been an invaluable source of information and inspiration for me. I’ve spent time working as a roasters assistant and bought my own green to experiment with but am far from understanding the finer intricacies of roasting yet. I’m struggling to know the differences between the makes of roaster and their dramatic price differences to boot!
    Anyway just an overall massive thanks!

    • Thanks for the kind words, Alex. I’m sorry I won’t see you in Adelaide, but am excited to finally see the city. Looks like a lovely place. I’ll be there 28/12–31/12; if we’ll overlap at all, I’ll stop in your coffee bar and we can chat roasting machines.

      • Unfortunately I’ll be out of town during those times. But I hope you enjoy Adelaide, it’s seen an incredible boom in the last few years in specialty coffee and i’m sure many places will live up to your expectations and benefit from your time there too!

  24. Thanks for your reply,

    Overall, did the single-wall drum negatively affect the way you could roast coffee or the results they produced?
    Surely you can pay a company to wrap the drum in a second skin and then you have a very well-equipped roaster?


    • You can roast great coffee in a single-walled drum. The double drum just gives you more protection against bean-surface damage, and makes the coffee a bit softer in flavor.
      And yes, it’s easy to build a second skin; many have.

  25. I would like to add here, that while this blog has an interactive voice, there is no replacing Scott’s “Coffee Roaster’s Companion” book — as it goes deeply into this. I have read it twice, and I intend on two more times by the point I get my first REAL roaster.

    Scott reminds me of what a great psych professor once told me: “Once you reach your gift, the next thing to do is share it with others.” Mr. Rao is all that.

  26. Hey all!

    Was there ever a follow-up to this post, where Scott talks about achieving this type of roast?

    Getting back into the roasting saddle after a year hiatus, working with a larger machine, and I’m having issues with getting my DTR over 19%.

    • Hi David,
      There has not been a follow-up post to this. My goal is to discuss ideas and lead roasters in the right direction, but I save the “how to” for my consulting business.


Leave a Reply

Your email address will not be published. Required fields are marked *