Announcement

Collapse
No announcement yet.

Roast Profiles - Fact or Fiction

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Roast Profiles - Fact or Fiction

    Hey CSers,

    As my roasting journey is progressing and I am learning more I'm also starting to form some of my own opinions and ideas. I've been roasting, cupping and drinking my own roast for a while now. I've had some awesome success! I've also had roasts tasting like potatoes...
    ?.

    I generally pick up roasters have one of 2 strategies.

    A Roast profile per bean taking into account origin, processing, MASL, density, brew method and the like. Basically every bean has its own profile unlocking it's secrets. KL has this in the form of its MASL profiles.

    A single, well known, tried and tested profile. Only changing the roast level. Basically focussed on changing as little as possible, presumably focussing on cupping in order to fine tune roast leve and developmentl. Most guys I speak to roasting on Coretto or similar follow this strategy.

    I am keen on learning from other's experience.

    How do you approach roast profiles?
    Do roast profiles have a significant impact beyond roast level, DTR and total time?

  • #2
    Still figuring it out really. 91 roasts under my belt. My goal when I got the roaster (April?) was to be able to emulate my favourite dark roasted beans bought from Gunshop Cafe in Brisbane. I've used the same bean Kenyan AA Lena for this and I can now produce the same result batch after batch. Same charge temp, same bean, same 800 grams, fan speed etc. Formulaic but it is predictable. Drop at first crack plus 2 minutes.

    So I have the formula/profile for that sorted and know that my daily addiction can be satisfied with my thrice morning fix.

    About half my roasts have been pursuing that goal singnthe same bean and the other half have been with 15 or so assorted beans, just playing around and seeing what happens.

    Now I'm playing with light roasts and seeing if I can dial in formula for that. I've started with the bean bays Yemen Mokha Ismail which is outstanding but alas not available at the moment.

    This is a looooong way from being able to consistently get a pre determined result from a wide variety of beans.

    I guess to answer your question more directly, I tried some profiles with the first 50 or so roasts but didn't have the result I was after so I decided I just had to figure it out myself.

    Comment


    • yochiya
      yochiya commented
      Editing a comment
      Thanks Tom, this is really the exact journey I'm on now as well. Just wish there was a way to get there a touch faster.

  • #3
    yochiya are you doing any light roasts and if so have you got a favourite bean?

    Comment


    • yochiya
      yochiya commented
      Editing a comment
      I predominantly prefer lighter roasts so I've been working really hard to find good options. It's hard. I find the higher quality the bean is the better they work for light roasts. That is, OK quality beans roasted a touch darker make good Espresso but the same beans roasted lightly is horrible. Really good quality beans roasted too dark lose so much of their flavours, I suspect you might get a very similar results with a lower quality bean at that roast depth.

      My favourite light roasts have been the Sidamo Ardi Naturals and Guji Haro Lebetu Naturals. If you are able to get the green cupping score you can make a slightly more informed decision. My best results with light roasts were with beans having scores of 86+.

  • #4
    tompoland I reread your post and noted you mention dark roast with dropping 2 mins past FC. What is your total roast time and DTR like?

    2 mins past FC on a 10 min roast is 20% DTR. Still never sure when someone talks roast level what their reference is. To me, this would be a medium light.

    For filter coffee, I've dropped 45s - 1:10s after FC with a DTR of around 13%.

    Comment


    • #5
      I would have to check my roast history but the roasts are longer than 10 minutes. I use a Tonino color analyser to determine colour so definitely dark. I guess bean size, % moisture, charge temperature, roast temp, fan use would all come into play. But I'll take a gander and get back to you on that.

      Comment


      • #6
        You roast with a KL do you "yochiya"?

        Mal.

        Comment


        • yochiya
          yochiya commented
          Editing a comment
          Yes that is correct.

          In the KL context. A common theme is see is some roasters stick to MASL or more tailored profiles based on the beans. I also however see roasters choose a single profile, like D-roast or Raost and just tweak the roast depth.

      • #7
        This might not help anyone else, and I'd be keen to hear what others think. Here are my thoughts after reaching out to some fellow roasters some of who are commercial roasters.

        A green bean has a flavour unique to it. You cannot change nor fabricate it in another bean that does not have it.

        The darker you go, the more roast flavours are present and the less natural bean flavours remain.

        A roast profile can enable you to modulate or tweak the flavour. This can be achieved by extensions or shortening of the major roast phases.

        Rob Hoos has a shmick graph in his book showing how changing the duration of phases impact flavour. In a nutshell, you can affect body, clarity, acidity, sweetness and complexity through tweaking these.

        In a commercial setting, some roasters have a significant roast signature. Some of these roasters are better suited to having a profile fine tuned for them in terms of charge temp, batch size, temp and flow curves. I would imagine a roaster learns their equipment really well, finds what works best for the machine and sticks with it.

        You are able to get a good roast if you follow the basics of roasting (let's say 85% perfect roast). You are however able to fine tune a profile for a specific bean to enhance certain aspects (never the core flavour) by extension or reduction of time in certain roast phases. By doing this you could potentially achieve a better roast than a 'standard' profile provided your equipment allows it and does not have a significant roast signature.

        Understanding the impact of the phases is something I'm really keen on digging into more.

        Comment


        • tompoland
          tompoland commented
          Editing a comment
          There is no doubt that the darker I roast, the less of the original flavour remains. Apparently something like 85% of cafe coffees are dark roasted served with milk. A friend suggested to me that is a bit like ordering a steak well done, and then smothering with sauce: not a lot of the original flavor is retained.

          As mentioned above, my first goal for roasting was to be able to consistently roast a dark bean for my espressos and lungos. I can do that now and that's my go to drink every day.

          Now I want to be able to do that with light roasts and see how much of the original flavor I can enjoy.

          Gotta love this rabbit hole!

      • #8
        Rob Hoos has a shmick graph in his book showing how changing the duration of phases impact flavour. In a nutshell, you can affect body, clarity, acidity, sweetness and complexity through tweaking these.
        Have done my best over the years to follow the same principle but my inspiration for starting down this path was after reading though several articles by Willem Boot. Sounds like another Dutchman with similar ideas...

        Mal.

        Comment


        • yochiya
          yochiya commented
          Editing a comment
          Nice share, thanks.

        • 338
          338 commented
          Editing a comment
          Thanks for the link Mal

      • #9
        So much for my memory.

        Looking at my roast history, I'm dropping my dark roasts at 2nd crack, not at 1st crack +2 minutes.

        The last dark roast (as measured by Tonino analyser was a "68" on the Agtron color scale) was 800 grams of Kenya AA Lena (larger bean), charge temp 240, yellowing at 6:15, first crack 10:42, end of first crack 12:38 and dropped at second crack 14:16. Therefore development time as a percentage of total roast was 25%.

        Does that sound about right yochiya ?

        Comment


        • Andy
          Andy commented
          Editing a comment
          Surely you'd know if it cracked twice? Don't need an analyser for that, just ears.

        • tompoland
          tompoland commented
          Editing a comment
          Yes Andy, it's my memory that's shot, not my ears.

          Someone with your experience doesn't need training wheels such as a color analyser and maybe if someone had told me six months ago that second crack = dark roast then I could have saved on the Tonino purchase.

          But as it it, I'm a novice so I'll be keeping the training wheels on a bit longer yet I suspect.

        • Lyrebird
          Lyrebird commented
          Editing a comment
          Andy

          Not necessarily.

          I don't track the cracks as I don't think they mean anything. I could not tell you whether or when any cracks occurred in my last 100 roasts.

      • #10
        Sounds correct, development time is the time from FC until roast end as a % of total roast time.

        14:16-10:42 = 3:34 (development time)
        3:34 / 14:16 = 25% (DTR %)

        Interesting how misleading DTR can be when you don't have full context. I could do a much lighter roast with a similar DTR at a shorter total roast time.

        ​​​​​​​Thanks for digging up the numbers and posting.

        Comment


        • Lyrebird
          Lyrebird commented
          Editing a comment
          Yes, because we don't actually know what DTR means, or if it really has any meaning.

          The same thing can be said about every other roast parameter.

          Simply put, at the current state of play, we do not know what we are doing.

          Some will not admit their ignorance but as always:

          "Complacency is the enemy of progress" - David Stutman
          Last edited by Lyrebird; 7 November 2021, 10:40 PM.

        • tompoland
          tompoland commented
          Editing a comment
          Yep, and "the enemy of growing is knowing" :-)

      • #11
        Lyrebird "don't track the cracks" .. sound like a Detroit rapper.

        I guess that begs the question, what do you use for feedback about how the roast is developing? Temperate maybe? I heard a podcast where the roaster (commercial) said that he can tell a lot by how the beans smell. The fried onions smell around the yellow point is quite distinctive.
        Last edited by tompoland; 7 November 2021, 10:55 PM. Reason: typo

        Comment


        • Lyrebird
          Lyrebird commented
          Editing a comment
          I developed an algorithm that attempts to predict development from the integral of a function of temperature*.

          It doesn't work, by which I mean it is not universal: the algorithm has to be tweaked for each bean variety. I am currently trying to work out what factors govern the differences. I think green bean sucrose level is significant but I have not had the time to do the work required to chase this.


          * Based on Arrhenius's reaction rate equation, adapted to model caramelisation in a manner analogous to Pasteurisation Units, so I called it Caramelisation Reaction Evaluation Model by Arrhenius, or CREMA. Sorry.
          Last edited by Lyrebird; 8 November 2021, 12:17 AM.

      • #12
        First crack can be a useful marker, but not if you can’t hear it. There’s a number of commercial roasters that due to their design you can’t hear cracks (eg. Loring) so you can’t rely on that audible milestone to guide you. To your point yochiya regarding ‘signatures’ I’d say that this is only partly true in that a very good cupper/taster might be able to pick up hints of the type of roaster used or some of the features it has, but that is very unlikely to extend to the brand or make. Ann Cooper and Rob Hoos ran a course and a little competition a few years ago that a friend of mine attended. The competition at the end of the course was to taste the same coffee roasted on three different roasters and see if they could pick up the differences. He did the best out of anyone, but said it was as much luck as anything as it was virtually impossible to pick up any discernible difference. In my experience one of the things that can make the single biggest difference is airflow. I roast at home on a Behmor and at work on an Ambex 5kg drum roaster, neither of which have adjustable fan speed. I find this a limiting factor when trying to make smaller adjustments at just about any stage of the roast and it’s the one thing I’d want to have in any new roaster. At work we only roast Peruvian coffee from one small part of Peru. So we only have a single profile and don’t deviate from it (not intentionally anyway!) before first crack. Different results are achieved mostly through different drop temperatures but also occasionally by moving through and past first crack at a different pace.

        Comment


        • yochiya
          yochiya commented
          Editing a comment
          Thanks for sharing this, very insightful.

          U would be very interested to better understand how and when you use airflow adjustments and to what aim?

        • LeroyC
          LeroyC commented
          Editing a comment
          I don't really use airflow adjustments because of the limitations mentioned above. When we're roasting on the 5kg Ambex we are often forced to close the vent at the start of the roast as even on full gas we can't keep up with the profile. This is common in winter and cooler weather (it's a pretty cold climate here and the roaster is in a small workshop type building with no insulation), but in summer and/or warmer weather we may not need to use it at all. The quality of the gas we get is factor in this as well as it's LNG sourced locally in here in New Zealand. The butane content is typically higher in winter, but varies throughout the year. We could look to purchase pure propane as the roaster is able to run on propane I believe, but it's a LOT more expensive and supply can be an issue as well as certification believe it or not. That was a long way of saying that it's just another lever to pull to keep us 'on profile'. Because of the variability in the way it's used and the lack of variable fan speed we're not able to have a set process for use of the vent. Does that all make sense? Hopefully we'll be moving to a 15kg roaster of some sort next year and it will definitely have variable fan speed. The main way we'll look to use that at the start if we do is to just slightly or gradually increase the fan speed as we move through the roast. It could be as simple as moving up one speed at about 150-170degC BT, but it could be more than that. The main reason to carry out at least this one change in fan speed is to reduce smoke levels in the drum and reduce the chances of the associated flavours that come with even lighter roasted coffees swimming in smoke for a period of time.

      • #13
        Thanks for sharing LeroyC that makes sense. I'm always intrigued by how different roasting is in a commercial environment than a home environment. It might seem very obvious to a professional roaster but as a home roaster I find these types of responses very insightful.

        Comment


        • #14
          From hundreds of 350g-850g Corretto roasts I can safely say that profiles are a 'thing'.
          I've just been making tiny tweaks in my ramps to 1C to try and gain a little more acidity while still maintaining sweetness and body. My usual 'flat' ramp' is to increase my heat gun by 50° at 75°/100°/125°/150° measured then dropping back from 175°, which is lovely but misses a little acidity as an espresso. Recently I tried 60/50/50/40 — same overall rise but much more acidic and less body. Then I tried 50/60/50/40 — a much more balanced result with better acidity. Look at a recorded profile and you could hardly tell the difference in shape — in the cup is s different story! Let your taste be your guide…

          Comment


          • flynnaus
            flynnaus commented
            Editing a comment
            Interesting. Thanks for that. Why 50°? Is that an arbitrary temp increase or arrived at by trial and error? Do you adjust this for a larger batch?

          • Dimal
            Dimal commented
            Editing a comment
            Hiya Matt...

            I've actually reverted back to fewer steps in the profile towards 1st-Crack, having arrived at set-points that achieve the optimum rater of rise, that then declines at the Maillard extension phase then increased once more from there to 1st-Crack. Has definitely improved the sweetness of all my roast batches, all the bean intrinsic flavours come through and seem to last right through to the last brew out of the bag and a very extended finish to boot. Mouthfeel has also improved and remained consistent. Lots of trial and error to arrive at this point, as you say...

        • #15
          flynnaus from lots of trial & error ? … I found that a gentle, warming start then a ramping temperature increase gave the best results. I speed up or slow down a roast to deal with varietals but adjusting the starting temperature after the turning-point - but normally this is only 10-20° on the gun.

          For example, with an 850g batch for most beans I normally use a starting gun temp of 410° from turning point, and then 50° increments at 75/100/125/150 measured by the probe. I used to go hard then until 1st crack and drop a lot at 195° or so – but now I drop to 430° on the gun at 175 then 390° at 185 which gives a pretty good roll into 1st crack, then small 30° drops every 5° from then on to keep a nice 3-4°/minute until just before 2C. Here's a profile of a recent Ardi batch if you're interested…

          Click image for larger version

Name:	20211020-ardilatepick2145dr.jpg
Views:	281
Size:	96.7 KB
ID:	892146

          And here's a smaller 350g batch – same temperatures and process, but fan speed 2 on the heat gun rather than three gives and almost identical flavour profile, although the time can be a little faster.


          Click image for larger version

Name:	20210623-kona350g.jpg
Views:	228
Size:	96.8 KB
ID:	892147



          Dimal I've actually been playing with a simpler profile – dropping the 'Maillard extension' dip. My brother uses an identical roaster I built him with a simpler profile, and had brews that had more acidity as a ristretto, which I'd been hunting. But I'd be interested in knowing what 'fewer step' way-points you're using now?

          Comment


          • Dimal
            Dimal commented
            Editing a comment
            Hi again Matt...

            Sorry for the delay in responding, been feeling a bit out of sorts.
            Anyway, to keep a long story short, I have identified three starting point temperature settings that gets the batch to the end of the Maillard Extension phase, for a Fast, Medium and Extended profile that works well with my Big Loaf BM. These HG temperature settings range from 420C, 380C and 350C respectively.
            Upon arriving about 4.0C before the end of the Maillard Ext. phase, the temperature settings are moved to 580C, 540C and 510C respectively.
            At the approach to the start of 1st-Crack at an indicated bean mass temperature of approx. 190C, the temperature setting is reduced to 430C and then incrementally downwards 30C at a time in an attempt to get a smooth transition through 1st-Crack, ultimately settling at a rate of rise of between 4-6.0C until the end of the roast, wherever that may be depending on the results I'm looking for.

            Unfortunately though, of late I have been finding it very difficult during the transition phase because of very erratic temperature reporting from the HeatSnob at this critical stage of the roast profile. Makes it almost impossible for me to know what the bean mass is actually doing. Giving some thought now to applying the 'fix' identified by member "Lyrebird" to try and get over this hurdle.

            Anyway mate, that's the long and the short of it. Still a bit of fiddling with temperature settings but much less that I was doing previously. Lots of fun...
        Working...
        X