Advise on 1kg espresso roast

I have been able to roast many successful 600g filter roasts using the bullet but I have just attempted some 1kg espresso roasts but finding I am having to charge the roaster fairly high and it still seems like it isn’t high enough to get a smooth declining ROR to fit the parameters i have chosen.
I am aiming for 5 min dry, 3.30 - 4 min to FC and 2-2.30min development time with a 9-10 degree rise before dropping.

I plan to be roasting for my small coffee shop so want to keep the weight as high as possible so it reduces the amount I need to roast weekly.

It also seems like my I-R sensor doesn’t work the same on the larger roasts ?

any advise on how to improve these, or any thing I am missing would appreciate it.

I don’t roast 1 kg batches but I can confirm that IBTS vs. Bean Temp has less separation as batch size increases. However, the reversal of BT & IBTS late in these roast profiles is possibly an indication of “junk” on the IR sensor and it needs a little cleaning. Beyond that the profiles look well formed, but you need to evaluate the taste- pretty graphs don’t always translate to a good cup.



My largest batch was 850g…However I echoed Bruce’s comment on a possible cleaning of the IR sensor is required and more often than a 500g or 700g batch.
The IBTS/BT cross each other happened a bit too early (which was in the early stage of the FC or even before the FC showed in your roast above)

Your profile looks very well controlled and I am sure a lot of us can actual learn from it on how to deal with large size of beans. Again, your final taste is the most important one.

With an expresso and a 2:30 development cycle in mind, I would consider keeping the P3/F5 toward the end of the roast rather than further down to P2/F6.
Too high (higher than p5 + longer than 20 sec can impact the aroma)
The goal for me on an expresso, is to keep RoR after FC @ 5~8C. If using roast degree, is to keep Agtron within 65~73. Hope that helps.

1 Like

I didn’t know you had to clean it more often with bigger batches, I mostly use the IR reading so it was super confusing when it stops reacting near FC. Will give it a clean every day on these bigger batches.

I still find the coffee is coming out to dark for me, maybe this will be Easyer to control when I have the IR reading but any thing above 210 seems to far.

I have cupped the beans and them seem fairly sweet with no noticeable defects still a little red fruits present, will put them in the hopper today and give them a test.

I would recommend looking at some of the roasting profiles by cash0216 and lav - they both regularly roast 1kg. I’ve been basing my 1kg roasts on cash0216’s profiles and I am pleased with the results. I am not hung up on how the ROR curve looks as long as it cups well.

A preheat of 300C / 572F is needed given the amount of bean mass. The BT and IBTS will track much closer to each other since there is more bean mass for the BT to accurately measure bean temp, compared to 800g or 500g. Crossing of the BT and IBTS does happen. My last 1kg roast was after a deep cleaning and you’ll see that they pretty much overlap during development time, but as Bruce pointed out if it crosses in a significant way it means you need to clean your IBTS. I’ve had them crossed before like yours but I wasn’t too worried, and yours wasn’t a significant cross to worry yet.

In RW you can search users by their username, and if you search for the two I mentioned - they are good references.

1 Like

for any one interested this roast turned out great, tested on espresso and filter.
Notes of chocolate, caramel and red fruits.

Very sweet and balanced on espresso with no roast flavor, any thing that goes over 209 tastes to roasty for me, not sure if this was a result of baking the beans but smoothing out the ROR changed the flavor massively and using the smell I pulled the coffee out when it smelt the best and it really shows in cupping.



I have done over 100kg of espresso roasts for my coffee shop now and this is my current 1k espresso roast - Washed Guatemalan

ended up extending the roast time but keeping same color just to calm down the acidity, very chocolatey and sweet.


Thanks for coming back to share how you evolved your profile. Between your May 27 post and the one yesterday, how would you compare the two that you posted on this thread in terms of the cupping results?

1 Like

They both cup great, I only really like light roasted coffee so I was very against taking the roasts longer at first, but in the cafe with milk based drinks the acidity at times became a problem so I found lengthening the roast and Dev time really chilled it out while managing to to maintain the same flavours and made it a lot easier to dial in.

So the latest roast is very sweet and balanced with less acidity but still a light roast.


It sounds like you found a way to get a light roast shine with less acidity. I may have to poach your last profile to try it out if someone I roast for wants a light roast. Most of my friends want med/med dark or dark :slight_smile:


Oh interesting, you went highest power to start [p9] then dropped down by 2 [P6] almost halfway through. I’ve struggled between this concept and steady power decline advise in a course I took w S. Rao. I’ll have to try your approach and sample it to his method to see how it compares. Thx for sharing!

Its in Scotts second book, he’s suggested bigger power changes on electric roasters as they are not as immediate as gas. I am constantly making small adjustments to this roast as i do over 30kg a week it allows me to really dial, lots of cupping and small tweeks.


Could you share your recipe for this? I would like to try this recipe, as I have been looking toward shifting from 1 pound to 2 pound roast potentially.

This is the recipe as of last night

That profile aligns quite closely with the CoffeeMind approach of constant power up to a first power drop a little before a 9 min FC, and and another power drop after FC to get to the desired degree of roast.


Anyone have a simple profile (ala CoffeeMind) that can produce a nice curve and RoR just into the first cracks of second crack? So far, I’m still scorching coffee on my Bullet. Seems like beans are getting stuck at the front of the drum from other beans falling on them, with the drum just sliding past the beans.

Just took a quick peak at your most recent roast on RW (a MX Chiapas 760g batch) me thinks the PH is high coupled with the P9 and F1 at charge time for over 5 mins - that’s a lot of heat in the drum and may explain the scorching. Just my opinion… My 500g roasts starts with 410F and a P6 or P7 at F2, but I bump it up by 2 Ps immediately after the turning point. I am also playing around with D speed given the lower temps to stir up the beans more to soak up the heat gently initially.

I haven’t had the time to study (and watch any of the videos) CoffeeMind’s methods so I have no opinion there. Like everyone else Morten, Hoos and Rao has their philosophies and in the end it is up to us to choose who they want as their “teacher” and extend from there. I haven’t subscribed to anyone except to leverage and amalgamate collective feedback/experiences on these forums and just experiment for myself :slight_smile: It’s a journey…


I just updated my “preventing scorching” post with this and two other beans roasted using the same profile. Two are burnt tasting, one is great. See here: Advice to prevent scorching - #49 by toddjohnson

Thanks for sharing this! My roast profile looks very similar to this and I Am also roasting espresso roast for a small cafe. Makes me feel a lot better !!

1 Like

Quite a late response… Anyway, have recently introduced a soak into my roast and have managed to mitigate any tipping on my beans. It was during a cupping session that someone pointed out some minor tipping on my beans ( had not noticed it myself). Based on that feedback, have dropped my starting P to 4. Did not want to go lower than that since I start my increase my F to 3 on charge. Right at turnaround, I push up the P to 9. That has helped in many ways. Further, am going for a hotter roast overall to decrease development time.

1 Like