Not a member yet?
Click here to register.

Forgotten your password?
Request a new one here.
You must login to post a message.

10/26/2021 1:55 AM
Welcome, Ukwoody, liamsmith, timbarnes, SydHagen and GISpoIRiS !

10/25/2021 3:16 PM
MendoEB, Welcome to HRO! Welcome

10/21/2021 2:41 AM
muckymuck, mckinleyharris and williamjonex welcome cup

10/19/2021 6:22 AM
Welcome tarunk!

10/17/2021 12:40 PM
Ploni and nader fouad, Welcome!

In Memory Of Ginny

Latest Donations
JackH - 25.00
snwcmpr - 10.00
Anonymous - 2.00
Anonymous - 5.00
Anonymous - 5.00
Users Online
Guests Online: 27

Members Online: 0

Total Members: 7,377
Newest Member: Ukwoody

View Thread

Who is here? 1 guest(s)
 Print Thread
Convection, Conduction, Radiation - Heat Transfer Ratios!
Hi all - first up, thanks for the welcome! Great first impression!

I have some rather esoteric questions to ask the brains trust (ie. you) about heat transfer ratios in different roaster types. I admit the theoretical nature of these questions. I am looking to use this info to build a better roaster, but I recognise that a theoretical perfection won't necessarily correlate to cup quality. There's my disclaimer!

My background: a mate and I roast with a Corretto (heat gun, bread maker) and have been doing so for nearly 18 months now. We've roasted a lot of coffee in that time (nearly 200kg/450# - we have thirsty friends!) and have experimented constantly with different ways of utilising the heat gun's output. Our system works best with 500-800g (1.1-1.75#). Any less and the mass of beans is too small to maintain its heat evenly; any more and the HG needs to be turned up too high to reach FC by 10-12mins. Still I sometimes wonder if it's the best it can be. Koffee Kosmo has generously shared his Turbo Oven roaster ideas here and elsewhere and that method has me intrigued. As is true for many of us, the idea of a 2kg/4-5# roaster is very appealing for future ventures...

So I started looking into various designs for roasters of this size... Unwisely, I decided to have a look at some of the first principles of roasting. What heat transfer methods do they use? What kind/rate of mixing? How much airflow? How do these (and other factors) affect the structure of the bean post-roast, moisture loss percentage, rate of staling, etc.

Pretty soon I found this article by Terry Davis (Ambex Roasters) where he asserts that modern drum roasters are 80% convective in terms of heat transfer. I later found this thread on this very forum where someone was reporting back having spoken to Karl Schmidt from Probat. Karl had given him a similar story for their (drum) roasters; 80% convective, 16-18% conductive, 2% radiated. That thread also gives some information on fluid bed roasters (96% convection).

There are several other articles/threads I've come across - you probably all know what it's like looking for this kind of information; especially on forums, the really useful info is just so disparate!

I want to build some kind of roaster that will roast close to 1kg/2#. In this thread, I just want to throw some ideas around about how knowledge of heat transfer methods and other thermodynamics will help me (and others) to build better roasters.
Here are some starters:
- Drum: 80% convection, 16-18% conduction, 2% radiation. Low airflow but excellent physical agitation. Large heat mass (all that metal!) so can afford a small temperature differential (am I right there?).
- Fluid bed: close to 100% convection. Very high airflow with good (though not even?) physical agitation. Large temperature differential, mitigated by high air flow.
- Corretto (heat gun/bread maker): I'd guess ~90% convection. High airflow, though the air doesn't go "through" the beans... Concentrated stream of very hot air can be a problem. Very inefficient system (ie. no lid) such that heat input is required to be very high.
- Turbo oven (think KK's design): I'd guess that the TO roaster would have a higher radiated heat transfer component, but still >85% convective. Lower air flow, though it does pass through the beans. Seems to be quite efficient (eg. 1400W can handle ~700g/1.5# roasts) with available heat meaning that heat input can be more gentle.

My perfect roaster, based on the above, would have:
- good airflow
- a significant component of heating that is not convective
- efficient use of available BTUs
- excellent physical agitation
>> all of which are meant to encourage gentle heating.

What think you?
Now that's a first post to chew on! Welcome, Stuart!

I'm a Corretto man as well. I also want a 1 kilo roaster, which means you and I are not so distant brothers!

I haven't the time at the moment to dig deep, but I would guess that conduction - from steel to bean (not bean-to-bean) - would be the least desirable of them all. Tipping and scorching and the like. Radiation would be second worst, and convection being the best/ most gentle, and therefore, my preference.

I started working on a roaster, but haven't had the time/ability to finish her...

More discussion later,
Roaster: CoffeeAir II 2# DIY air roaster
Grinder: Vintage Grindmaster 500
Brewers: Vintage Cory DCU DCL, Aeropress, Press, Osaka Titanium pourover
i would think that a solid drum would with low air flow/venting would maximize conductive heat transfer, but i thought that that was not desirable as per the points seedlings posted above.

i'm pretty new to this whole scene, but the roaster i'm building is essentially a drum roster with added convective heat, assuming that the more homogeneous the temp inside the roaster is the better (so long as i can control it :)
"If it wasn't for venetian blinds it'd be curtains for us all"
Koffee Kosmo
Hi Stuart & welcome to Home Roasters

I have also been thinking about a larger roast with the Turbo Oven
The original Turbo Roaster design was made for me and the DIY market by utilizing off the shelf products not dissimilar to the Corretto

My next project will be to build my own turbo oven by utilizing a round 2000w oven element and fan that is currently used in most household ovens in the capacity of the fan forced system

I have a friend that is an electrician that will help but will be made sometime in the future as I am tied up with building the kit roasters

Hopefully by utilizing the current lower set up and the new turbo it will roast 1.5 kg as the extra heat and extra fan force will be used more efficiently & effectively

I have some other designs that incorporate dual heat sources but I will need a good workshop to make one

I even have a Twin Turbo Design drum roaster

Koffee Kosmo attached the following image:

I home roast and I like it. Designer of the KKTO
Roaster Build information
Blog -

Bezzera Strega, Mazzer Robur Grinder, Pullman Tamper Convex,
(KKTO) Turbo Oven Home Roaster.
Here's a GREAT (7.7Mb) animated read:


On P27, "Specific heat requirements for Arabica coffee at a specific roast
color and with a green coffee moisture of 11.5% is approximately
470kJ per kilogram of green coffee (theoretically)."

Great stuff.
Edited by seedlings on 09/03/2009 8:22 PM
Roaster: CoffeeAir II 2# DIY air roaster
Grinder: Vintage Grindmaster 500
Brewers: Vintage Cory DCU DCL, Aeropress, Press, Osaka Titanium pourover
Welcome, Stuart! You pose some fundamental but complicated questions. But they're the ones we like the best. We don't know any answers, but that doesn't ever stop us from having opinions!

Thanks for the links. I enjoyed reading the Ambex page.

You mention possible dangers of high roasting temps. Have you seen the paper from SCAA and Agtron on Basic Chemical Reactions by Carl Staub? He makes a good case for keeping temps no higher than 520F. I think he's been quoted by other experts like Kenneth Davids.

Most of the roaster types you mentioned are specific enough to have approximate numbers assigned (guessed at?) for the balance between heating mechanisms. But it seems kind of risky to put percentages on something as variable as a "drum" roaster. There are too many wildly different drum and roast chamber configurations to generalize on one set of percents for air/conduction/IR. I know in my home-built drum a high percentage of the roasting comes from IR. I used a 73% permeable screen for the drum to enable direct air and IR from the electric heaters. I run air temps near the drum no higher than 510 and use only 1-5 CFM for air venting.

Fluid bed roasters are also somewhat variable. I believe that they have the potential to roast with the smallest temperature differential. High speed air is very good at transferring energy. But it needs a high pressure blower to lift and mix significant percentage of the beans. Still, fast air can be much better than conduction which always suffers from a small point of contact. I suspect that a lot of what some people have subscribed as contact heating is actually long-wave IR from hot drum metal. It can be tricky to figure out what mechanisms are actually heating the beans.

I learned a lot from the book Coffee Technology by Michael Sivetz. He goes into a fair amount of the engineering behind the machines. I wish I could find more technical references like his.
11 years old... forever!
>home-built roasters and fair trade
I'll add a theoretical equation for some computating fun.

(Remember: in theory, theory and practice are the same; in practice, they're different.)

seedlings attached the following image:

Roaster: CoffeeAir II 2# DIY air roaster
Grinder: Vintage Grindmaster 500
Brewers: Vintage Cory DCU DCL, Aeropress, Press, Osaka Titanium pourover
About that NCA2007 Probat-Burns figure of 470 kJ, I'd sure like to hear how they arrived at that. When I figure it I get more like 609 kJ per kilo, 277 kJ per pound. And I'm just crusty enough to think I'm right!
11 years old... forever!
>home-built roasters and fair trade
I would suggest analyzing your roaster design in the shoes of a coffee bean with an engineering degree. What variables are available to the roaster? How do they affect the bean's experience? Take each variable to the extreme limit to gauge effect.

I wouldn't suppose either conduction or convection to be superior; a design that allows appropriate and even energy transfer is preferred
Having the ability nail post 1st Crack progress, neither stalling nor running is the highest concern, IMO.
-Marshall Hance
Asheville, NC
Hey all - fantastic to see such good input from people!

#CHAD (long lost brother! :P) - I think you're right about metal-to-bean conduction being the least ideal method of heating the beans. I guess large drum roaster (whatever style) have excellent physical agitation which helps to soften the conductive heat transfer (and avoid scorching). My thought is that this mitigating effect would be less effective the smaller the roaster. In a 1-2kg design, I suspect lowering the conduction component would be the best way to handle that. One way to do this would be to have a perforated drum. With a perf. drum I guess you risk too direct a heat (radiant) if your heat source is cranked up to cope with the large bean mass - so it's a fine line! To mitigate harsh radiant heating, you'd need to keep the heating gentle, meaning that you'd need good efficiency of over all heat transfer.

As for radiation being second worst... perhaps it depends. This paper abstract talks about how long-wave IR roasting results in fast, even heating through the bean. On the other hand, I agree (intuitively) that direct radiant heat would tend to be quite harsh.

And thanks for the links.

#itsallaroundyou - is this a small(ish) drum roaster with convection? If so, I'd love to know more... is there a thread I can look up?

#KK - there a whole lotta cool-factor to the twin-turbo design... I very much like the notion of a KKTO-style roaster capable of up to 2kg. For readily available sources of heat, I think turbo ovens have the right mix of power and air flow. We're lucky in Oz that we have more watts to play with from a power point (what is it, like 3000W?).

#dBndbit - info on max roast temps is very useful, thanks! I guess that's air temp within a drum? Agreed about drum roasters being highly variable. And that's just commercial ones! Deidrich's that use IR are clearly very different... I suspect perforating the drum gives a very different style of drum roaster, too.

Does anyone have figures for the air flow rate of heat guns and turbo ovens? I haven't read into air flow too much, but I do suspect that our Corretto dries out the beans too much. We lose 17% of the original weight of beans on average - that's for 16-18 minute roasts to ~SC. KK said his TO roaster gives 12-15% (from memory). What do other get, and is it simply proportional to air flow and roast length?

#endlesscycles - I'm yet to meet a coffee bean with an engineering degree, but here's hoping. I like your point, though! I'm in the 'heats ain't heats' camp of thinking, here. I suspect that it matters greatly what kind of heat is used - though you can't use one without using the others. Having said that, you're right about post-FC control being super important. IMO that's another topic - though related.

Great discussion. I'm keen for any other/subsequent thoughts!
John Despres
Stuart, way to kick off a great discussion with your first post! I'll follow carefully as I'm beginning to think about building another roaster myself. Bigger and better this time, though.

Respect the bean.
John Despres
Fresh Roast 8, Gene Cafe, JYTT 1k, Quest M3, Mazzer Mini, Technivorm, various size presses and many more brewers.


dBndbit wrote:
About that NCA2007 Probat-Burns figure of 470 kJ, I'd sure like to hear how they arrived at that. When I figure it I get more like 609 kJ per kilo, 277 kJ per pound. And I'm just crusty enough to think I'm right!

I believe you, Jim!

Now, my question is what to do with the ~100,000,000 J figure required to heat 1 kilo of beans ~400F. Do you divide that by the 15 minutes of roast time anticipated to get 8,300,000 J/minute? And then, how does that figure back into calculating heat requirements? Guessing is almost as easy.

I earned an "F" my first time through Physics at Mizzou.

John! If you haven't already, download that Probat presentation and save it. There's great info in the last 15 pages - questions you've asked recently.

Edited by seedlings on 09/04/2009 7:49 AM
Roaster: CoffeeAir II 2# DIY air roaster
Grinder: Vintage Grindmaster 500
Brewers: Vintage Cory DCU DCL, Aeropress, Press, Osaka Titanium pourover
John Despres
Great bedtime reading. That is definitely something for the notebook. Thanks, CHAD!
Respect the bean.
John Despres
Fresh Roast 8, Gene Cafe, JYTT 1k, Quest M3, Mazzer Mini, Technivorm, various size presses and many more brewers.
Chad, about joules, watts and BTUs...This is definitely something under the heading of "FWIW". But for the last month or two I've been pulling together all the numbers and equations I have on physics/engineering and coffee roasting. Mostly just trying to get my own head around all the details, trying to eliminate gaps. I include the background info on specific heat of coffee beans. Maybe it would be useful. Hopefully next week or so I'll have it wrapped up for "peer review" by all you peers.

Short answer is that you do have to include time as you think it applies to a roast profile. A watt is 1 joule per second. And a BTU is about 1055 joules. That makes a watt (1 j/s) about 3.41 BTU/hr. But there's more going on than just number conversion. For instance: calculations for a coffee roaster are like designing a car to drive over the Rocky Mountains. Do you need to have a car with the average power over the whole trip? Or does the car need the power defined by just the uphill part of the trip and the brakes defined by the downhill part?

And I do know why my joules don't equal Probat's joules. But it would take some more explaining that I don't think Stuart would like me dumping on his thread. I want to hear some more discussion on convection/IR/conduction.
Edited by dBndbit on 09/04/2009 9:14 AM
11 years old... forever!
>home-built roasters and fair trade
We need an experiment!

Roast an amount of coffee X in a popper (mostly convection).

Roast the same amount of coffee in a frying pan in the same amount of time and to the same roast level (mostly conduction).

Roast the same amount of coffee, same roast level, same amount of time in the oven, spread out, not touching each other, on a sheet of aluminum screen (mostly IR).

Compare and contrast.

I know what the popper coffee will taste like. I suspect what the frying pan will taste like, but am unsure of the oven.

Alternately, you could start with any of these methods, then try to mimic the time and roast level by other means.

Edited by seedlings on 09/04/2009 10:08 AM
Roaster: CoffeeAir II 2# DIY air roaster
Grinder: Vintage Grindmaster 500
Brewers: Vintage Cory DCU DCL, Aeropress, Press, Osaka Titanium pourover
Thanks for this thread. Include me in CHAD'S Brotherhood. I am looking for a roaster that will do about 1or maybe 2 kg. 1 would be great to start. Because I live in Cleveland Ohio would have to roast indoors and vent somehow. I will follow this thread with joyous anticipation. I don't think I can go with a 1/2# max batch that I have.:eye-popping:
May the Force be with your cup
Hi again,

Thanks for even more responses! Seems like there's a few of us (4 have said so on this thread so far) who'd love a 1-2kg capacity roaster! I really hope we come up with some helpful info on this thread.

#Jim - sounds like good, interesting info... but I am partial to keeping threads "clean" (makes info so much easier to find at a later date!)! Please feel free to post a link to that info from here, though!

#Chad - your experiment is a worthy one, but while it will provide interesting data on the different heating methods' contributions to flavour, it's not quite the experiment we need. I say that because according to those two sources (listed above, somewhere...), convection is still 80% of the total heat transfer in a drum roaster. [I guess we should limit that to commercial roaster - where air flow through the drum is clearly important.] As such, our real variable ranges are 80-100% for convection, 0-18% for conduction, and 0-15% for radiation (not sure on that last one - wild guess based on some IR home roasters?). What might be interesting is to be able to use a drum roaster a) with no IR and low airflow, b) with IR and low airflow, and c) with no IR but normal airflow. All we need is someone VERY skilled with a drum roaster that has very adjustable air flow and two heat sources! Easy. Any volunteers?

Here's another one. I do like my heat gun. It's powerful (2000W) and adjustable. What I don't like is that its air flow is quite strong and quite concentrated. What designs have people come up with that utilise a heat gun but do so with gentle (and hopefully efficient) heating? I have visions of a HG output being piped in under a bean mass in a bread maker... would this be feasible?

Keep the thoughts coming - I'm really enjoying this discussion!
stuart, with my roaster, i'm aiming for between 1-2lbs, hoping for closer to 2. the basis for my roaster is a convection oven, two banks of heating elements (independently controlled) and an exhaust fan. there is a thread going here: http://homeroaste...ad_id=1478

the biggest challenge so far has been getting my heat retention/insulation up to an acceptable level (for roasting), so this thread will be really helpful i think.
Edited by seedlings on 09/04/2009 1:52 PM
"If it wasn't for venetian blinds it'd be curtains for us all"
Stuart, here's a problem with the heat guns. I'd guess that the heat elements are designed to have a specific air flow. Lessen the airflow without reducing current to the element and the nichrome will probably fail. So, the only other option is to reduce the heat AND airflow simultaneously. That's fairly easy with a triac-timer circuit. Basically a 555 timer circuit, the output of which is a triac or Solid State Relay. It's like turning the power on and off rapidly, which has a similar effect to reducing the voltage to the fan and the element (whole heatgun), which has the side-effect of longer life of the fan and element.


Edited by seedlings on 09/04/2009 2:02 PM
Roaster: CoffeeAir II 2# DIY air roaster
Grinder: Vintage Grindmaster 500
Brewers: Vintage Cory DCU DCL, Aeropress, Press, Osaka Titanium pourover
Thanks guys,

#Chad - with the heat gun, I was thinking more of spreading the air flow out somehow, rather than reducing it outright. Moving the heat gun further away from the bread maker pan is one way to do this, but decreases efficiency even further (and it's already pretty low!). So I guess I was thinking of ways to pipe the HG's output into/onto the beans in such a way that it was spread out/not so concentrated.

Having said that, I'd still like to know how the total fan output of a HG compares to that of a Turbo oven...?

#Mike - good thread. Some useful info there - would love to see some photos of what you've got so far. Best of luck with it; I'll be sure to follow the thread.

Jump to Forum:

Similar Threads

Thread Forum Replies Last Post
Larry Cotton's 'wobble disc' heat gun build Heat Gun Roasting 12 09/24/2021 9:16 AM
Simple fluid-bed roaster attempt: wiring heat element Fluidbed Roaster 4 06/14/2021 7:27 AM
Heat, smoke and chaff - fluid vs. drum Roasting Coffee 7 04/22/2021 5:40 PM
Looking for heat gun alternative Fluidbed Roaster 7 04/20/2021 4:28 AM
Heat gun bread machine roaster--how can I change container? Heat Gun Roasting 8 03/05/2021 4:14 PM
Homeroasters Association Logo, and all Content, Images, and Icons © 2005-2016 Homeroasters Association - Logos are the property of their respective owners.
Powered by PHP-Fusion Copyright © 2021 PHP-Fusion Inc
Released as free software without warranties under GNU Affero GPL v3
Designed with by NetriX