How to Calculate Weight Loss in Coffee Roasting

Measuring weight loss in coffee roasting is a great tool to understand your roast level. Here’s how:

Coffee loses bound moisture in the bean during the roast process, which is the primary reason that a batch of 100 grams comes out weighing 85 grams when it’s all nice and brown!

While coffee expands in volume due to the expansion of the green bean, it diminishes in weight, mostly because of loss of water weight. Recording this weight loss as a percentage can be a really effective way to get consistent results, or to match the roast you did for a particular coffee when you roast it again, maybe weeks later.

While it’s easy when you are working with certain green coffee batch weights, it’s a little harder to calculate when you roast 120 grams or 454 grams, about 1 lb.

I roast 100 grams when using my sample roaster, so I don’t need a formula. When I weigh 85 grams of coffee post-roast, thats 15% weight loss.

Our Short Video on Calculating Weight Loss in Coffee Roasting

A video we made to encourage using a scale for roast batches and calculating weight loss in coffee roasting!

Here’s a formula for calculating weight loss in coffee roasting

How do I calculate weight loss in coffee roasting?

First, weigh your batch before and after (it’s important not to miss a single bean when doing so!).

Then, to covert weight loss to a percentage, use this formula:

(Green Weight – Roasted Weight) / Green Weight * 100 = Weight Loss %.

Another way to phrase the roasted coffee weight loss formula is:

  1. Subtract the post-roast Roasted Weight from the pre-roast Green Coffee Weight.
  2. Divide that number by the Green Coffee Weight
  3. Multiply that by 100
  4. Equals Weight Loss as a %

So if I start with 95 grams, and end up with 85 grams, then 95-85 is 10, divided by 95 is .1052, and times 100 = 10.52% weight loss. Got it?

Roast Weight Loss and Degree-Of-Roast Names

At Sweet Maria’s, here are the approximate values for weight loss ascribed to the roast level names we use. (Note: These are average between air roasters and drum roasters so they might not fit 100% with your machine)

1st CrackCity-CityCity+Full CityFull City+FrenchBurnt
412 f418 f425 f432 f438 f442 f448 f455+ f
– 10.3%– 11.5%-12.7%-13.3%-14.5%-15.1%-15.6%-16.6+%
Weight Loss during home roasting depending on degree of roast.

Problems and Pitfalls with Measuring Weight Loss in Coffee Roasting

I mention this in the short video … but one issue is that green coffee types do not always share the original moisture content. Generally yes, they are in a range hovering around 10.5% or so. But outliers can be coffees from Ethiopia, Yemen, and some oddball lots like India Monsooned coffee.

Why does this matter?

  1. It makes it harder to compare roast levels between a coffee that starts with 10.5% moisture content versus one that has 9.0%. You can roast to the same weight loss percentage overall but will have 2 different roast results.
  2. If you roast a coffee that only has 9% original moisture, and your target is 15% weight loss from green to roasted, your roast will end up a lot darker than you expected.
  3. You probably don’t have a moisture meter for green coffee laying around your house, so you can’t check green coffee moisture content easily. Ours at Sweet Maria’s costs around $1800, and while there are some available for $250 or so, they aren’t super accurate or consistent in my experience.

But these are small drawbacks really: The fact is, you can compare one great roast batch of Yemen to another , and enjoy a more repeatable result … even if comparing a Yemen to a Sumatra might be thrown off by moisture readings.

For more on roasting and roast levels see:

23 Responses

  1. I have been keeping track of weight loss for years, and it’s a good way to see the results post roasting. I even record the weight loss on the label of each batch. But you only find out the weight loss when the roast is done. There’s where the watt meter comes in.I switch mine to kilowatt hours, and record the total number at various stages of roasting. When I have roasted a particular coffee, say Mocha Hajjah, at starting weight 6.5 oz, in a SR800 with extension tube, maintaining the temperature in the range 447 to 453F fan/ heat between 5/4 and 5/3, it takes about 14 — 15 minutes to get to 350 watt hours (=0.350 kwh), which when I stop. This yield a weight loss of 17.5%. Without the extension tube, I get similar results in 12 minutes or so, with 330 watt hours, weight loss 17.0%. The temperature reading was kept in the range 447 plus or minus 2.

    I’m wondering, temperature readings on the SR800 without the extension tube may not be equivalent to temperatures when the extension tube is in place. The roaster is (according to the UC Davis Engineering coffee faculty) a fluidized bed reactor. The coffee is more fluid in the extension tube, but the temperature, measured at the bottom of the tube, may not accurately reflect the temperature of the beans as they move in the tube. I’m experimenting.

    The watt hours measure fan and heater activities, confounding variables. Thus when gauging the progress of a roast by watt hours, one needs to be consistent with the fan/heat settings.

    1. That’s really interesting. I need to try your method of using the kilowatt hours … as you say you need to do the same heat / fan changes or it throws off the measurement but seems like an interesting approach toward consistent roasts.

      Agreed on the issues measuring bean temperature vs environmental temperature in any air roaster – difficult to measure bean temp in a rapid air environment for sure

  2. Can we assume that as green coffee ages it loses moisture? Would that happen within weeks, months, or years of acquiring coffees? And I’m guessing this is particularly the case with heirloom coffees. Although green coffees wouldn’t age out all that quickly as candidates for roasting, I’d be interested in knowing how roasters decide when green coffee is no longer worthy or keeping for roasting.

    1. A very good question and could take that in many directions. In short coffee loses weight with age particularity if it’s not in a high barrier container or bag and the atmosphere is switching out and if that’s happening it’s also carrying out positive flavor and aroma compounds ( or precursors to those compounds ) so there’s quality loss. Plus moisture is the medium to help transfer heat through the bean so roasting gets harder. Definitely there’s a point where the coffee is beyond hope. But interestingly, the roasters girls, this year is focused on Roasting old coffee, and getting the best result, a super great roasting challenge!

  3. Ever since the roast color card became available, I’ve had it posted in my roasting room. Super helpful. Thanks Tom.

    FWIW: I’ve been using a minor variation for measuring weight loss: (final weight / original weight) – 1

    1. I have been trying to find a printer to produce a higher quality version of the card. The color squares would be much easier to compare, printed using spot colors, like paint chips are at hardware stores. Not a lot of luck but still working on it. I also want to come up with weight loss charts for specific roasters, because they do vary a lot. Behmor weight loss correlated to degree of roast / color (like Full City) is a lot different than a Popper or a Freshroast.

  4. I think being aware of bean weight loss is a very interesting idea. I just tried it and got a 14.5% value which I believe corresponded quite well to the roasted bean color.

    At the risk of sounding nit-picky, I’ll just raise a question about dealing with the chaff weight. The removed chaff probably doesn’t weigh all that much, but it’s not zero. That would mean that the removed moisture percentage calculated using the before and after bean weight is somewhat overstated since along with the moisture, chaff is also removed during the roast. In addition, you get different amounts of chaff from different beans which further adds to the potential variation.

    This is probably not very significant, but I thought would at least mention it.

    1. Actually its a really good point, and not nit picky. I’m going to see if I can actually check this tomorrow with an air roast of 100 grams and see what the chaff weighs.

      BTW a interesting side note here. I found out years ago that a trick of the trade in commercial scale roasting for ground coffee (Folgers, MJB etc) is to collect chaff, impregnate it with water, pellitize it, and then grind it back into coffee before packaging. Chaff is technically coffee, so your label can still read “100% arabica” or whatever, but what they mean is “beans and chaff”. ingenious really. Nothing gets wasted … but chaff has a very astringent grainy taste. And this still happens in large scale roasting today, with that they call RG coffee – roasted ground (vs whole bean)

  5. I’m roasting in one of the stainless steel popcorn poppers purchased from Sweet Maria’s. I’ve often wondered if the venting helps or hurts the roasting process. Should I try to dampen the venting, allow the steam to escape only through the vents, or would the roasting process be better if I left the lid open from mid roast? I stopped using the thermometer also and feel like I just guessing based on how the beans look just past the first pop. I am afraid I will burn the beans! I’d love some feedback

    1. It’s actually a question I have never had, and never seen discussed. Not sure why! But what I imagine is that as the steam escapes the coffee (in particular first crack, when the internal bean temperature turns moisture to steam) that there’s no way to contain or hold that moisture in the roast chamber / drum / popper / pan etc. It’s basically gone as soon as it leaves the bean due to the temperatures of the roast environment. So I guess we can’t pressure cook coffee (in a hot pot?) but it makes me really curious about what would happen if you tried to keep the roast environment humid!

      Anyway, my sense is even if you tried to prevent the popper from venting, keeping a lidn tightly shut, it would lose the moisture anyway.

  6. i was intrigued as to what pressure you’d need to reach say 415F (215C) in a pressure cooker to try cooking coffee that way so i looked it up by trial and error here and it turned out to be 20 bar (20 atmospheres) or around 300psi. Sadly we won’t be doing this anytime soon with anything handy as conventional home pressure cookers typically do at most 1 bar (15psi) above normal atmospheric pressure.

    But back to roast weight loss: roasting in a 5LB Ron Kyle drum in a BBQ for many years now, mostly recording weights, typically for my setup droping roast just before 2nd, 15 to 17% weight loss depending on bean. Almost never less than 15%.

    i do (1 – (roasted / green) ) * 100 but i do it just roasted/green giving numbers like 0.8427 which i record as 84.3% net which translates easily to get 15.7% loss

    1. Ok interesting. That is a lot of pressure indeed! Beyond 9 bar espresso machine pressure. I imagine 20 bars is a bit dangerous in fact – more like something in a commercial boiler

  7. For years, I thought I was the only one using moisture loss as a tool for roasting. (I live in a Cave, my garage). I’m nearing a 1000 roast on my Hottop. I make spreadsheets for monitoring my roasting. I shoot for Vienna Roast, (Missing on your bean chart) I roast 300gr and log in the weight loss. I don’t venture outside my garage and don’t know what other roaster’s are doing. Glad to see I’m on the same page. This little snippet of information is crucial to me. I’ve been trying to ween off French Roast and oil and have landed a notch lower to Vienna. That is for me, going from 16% and 17% to 15%. I also “stretch” out the roast between first crack and second crack. I try to avoid “divots” and chaff still on the beans when roasted. “To each His Own”…… From the one minute mark to 400 +- degrees, I have a slope of 40 degree rise. It’s the “Divot” thing. Between 1st and 2nd crack with the “stretch” time, my fan is on 10 and Heat set to 40%. These settings aren’t a constant as every roast has it’s challenge. The bean chart : bring back the term “Spanish” Bring back “Vienna” and C++ and C+++ for the nit picky and me. And “Quaker” (300 x 1000)/454 = how many pounds I have roasted in my Hottop “Life is Short, Drink Coffee”


    1. Fantastic – looks like you are getting great results with your system, Mr. V ! I think what we call Full City + truly is Vienna, but its all semantics. I think the frame of reference for Vienna level was lost on the “third wave coffee” crowd, plus they were all in a competition to roast lighter and lighter. It was outside the frame of focus. But the names for darker roast levels, from Vienna to French to Italian, and perhaps darkest of all “Spanish” did seem subjective all along. After all if you go to Italy and are served a light roast, isn’t it still an Italian roast? ha ha .

  8. Mostly out of curiosity, I made a rough attempt at measuring the chaff weight loss during roasting. I roast using a Behmor which results in post-roast chaff in the chaff collector, the bottom of the Behmor, and that which remains inside the drum both free and still attached (slightly) to the beans. I made no attempt at collecting the chaff from the inside of the drum or that which escapes the drum after I take it out of the Behmor since I blow air on the drum from a small leaf blower to rapidly cool the beans and vigorously shake it to get rid of the remaining chaff. For what it’s worth, the chaff I collected from the collector and the bottom of the Behmor totaled 0.9 grams from the 8 ounces (227 grams) of green Monkey Espresso Blend that I roasted. If that were all the chaff, it would amount to about 0.4% of the initial green bean weight. I’m guessing that the chaff that I missed might have raised the total weight loss to around 0.6% to 0.7%.

    My rule of thumb would be to figure on a 0.5% weight loss due to chaff. So, if you measure and calculate a 15% total loss, it’s closer to 14.5% moisture and 0.5% chaff loss. Of course, different beans can have different results and a more complete chaff collection may provide a different result as well.

    I should also point out that the very well-done chart on the website comparing roast level to moisture loss percentage did not include any consideration for chaff weight loss – nor should it. The chart works just fine without even thinking about the pretty insignificant chaff weight loss.

    1. Wow this is great! Thanks for doing this test. It’s very informative and honestly something I really didn’t think of much when considering weight loss until it came up in discussion here. I do think different coffees are going to have more chaff, in particular those very “yellow” dry process coffees. So maybe I will try to compare some roasts between wet process and dry process to build on your effort here. Again thanks Dick

  9. As others have noted, moisture loss tells you how you did, not what to to, as you can only guess at the moisture loss until the batch cools and you can put it on the scale.

    I roast by color and smell first, IBTS temperature second and then compute moisture loss to see if there is an outlier. I’m usually shooting for Full City or perhaps a tad lighter.

    I wish a bag of retail green coffee from SM was about 520 – 525 grams, which would give us a roasted pound when finished.

  10. Like your color and moisture chart. Just wondering what is happening when color is right as far as I can tell but weight loss is off. Usually
    to high. Example full city color but 16% weight loss.

    1. This can happen for sure… even in the opposite direction: a coffee at Full City with only 10% weight loss. I can get that with Ethiopia coffees that arrive with low mositure levels, sometimes 8.5% even. Normal range for arrival is 10.5%-11%. It’s possible some Indonesia wet hulled coffee arrive at 12.5% or so, meaning more moisture loss during roast. Also depends a lot on the type of roaster. For our card we averaged between drum roasters like Behmor and Bullet, and Air Roasters. With out Probat roaster we can see 18% weight loss for full city. So the roaster type has a big impact. Maybe we should produce a card for air roasters and another for drum type ?

  11. Do you measure the moisture content of each lot? Would that be a number you would be confident sharing on a Bean’s product page?

    I use the color chart you created, but I also roast a lot of decaf. So, I tend to use the percentages more to get a sense of when I ended up. If I knew, to some degree if a been had a high/low moisture content to begin with, it would help me dial that in even more.

    1. We measure the moisture content in a lot of coffees, but not all of them. We definitely check it if we suspect it could be higher low, but that’s mostly on offer samples, not on the arrived coffees because they are normally in a pretty tight range of 10 to 11%, such a small range wouldn’t have a huge bearing on weight loss measurements, but I can understand if you really want to pin down the weight loss, the moisture level is a factor. I can look into what it would take for us to add those numbers to reviews.

  12. 16 burnt? The color I get is more a full city. This value with percentages are too extreme. Hoos coffee they mark 15-17% is still in medium range why are you guys saying it’s burnt lol

    1. It does depend on the roaster. A probat drum roaster will be City+ at 16% weight loss.

Leave a Reply

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.