Espresso Extraction Ratios

(slightly revised post)

Speaking of Andy Schecter I must say that I am now a fully paid up member of extraction ratios when it comes to defining espresso drinks. I suppose I had always agreed with it but had been unable to put anything to the test because I didn’t have a good enough set of scales. Until now.

It is really great to have a set of lab balances that are accurate to 0.1g up to 1500g. I picked them up second hand so they may be a little old and beaten looking, but they still work very well. This means that I can now not only weigh shots, but also weigh portafilters as I go just prior to insert which means I can see how much I am dosing and how accurate my techniques are.

For those of you unfamiliar with all this the premise is relatively simple. Weight the amount of coffee you use and then weigh the amount of liquid you end up with afterwards. This, to me, seems superior to volume as it isn’t really influenced by the freshness of the coffee. A very fresh coffee tends to brew with a fluffier crema, as does a naked shot, but if we let the foam collapse we might find that 2oz through the naked is actually a different volume of coffee to 2oz from a spouted pf. With these two volumes it is possible to come up with an extraction ratio which is a very easy thing to communicate and use as a standard when comparing shots with someone else. (Divide the weight of the dry coffee by the weight of the liquid and multiply by 100 to make a percentage – I had this back to front first time I posted!).

At Home Barista Andy published this table

This is the new, simplified table:

After a session of weighing today I discovered that most of the espresso I enjoy is around the 60% mark, edging away from espresso slowly into ristretto (going by the table). What was also interesting was to analyse my habits with a naked portafilter versus spouted. Be it habit or a trick of the eye I woud typically pull a slightly shorter shot (in liquid terms) giving it a higher extraction ratio. Am I alone or are we all a little guilty of this? Are we attributing the mouthfeel from the naked when it is mostly down to it being a stronger shot? I don’t know.

I am not saying that this table is law, nor should it be. But I do think that with the right equipment it is easy to compare shots for shots (certainly with those we only meet online) if we are talking about brewing the same thing – regardless of machine or some degree of roast freshness. Perhaps the numbers in the table are a little too rigid or convenient, but that doesn’t dampen the idea for me one bit.

Still – you need a decent set of scales for this. Anything cheap is likely to be pretty inaccurate, certainly around the higher numbers. God bless you ebay…. (If anyone knows where I can get a manual for a random KTron scale then please let me know!)
[tags]espresso, brewing ratios, coffee science, barista, extraction ratio, coffee ratios, ristretto, lungo, home barista, andy schecter[/tags]

13 Comments

  1. i’m sure i’m guilty of shorter naked shots — but only because there’s limited space in a shot glass. the enhanced volume demands that you stop the pull sooner.

    pulling into a capp cup, however, is a different matter. one might be well advised to watch for the classic signs of a complete extraction instead of a volume level.

    on second thought, that makes me wonder — does a spouted pf change the crema’s color or visible consistency at all, versus a naked? this could also affect when you’re cutting the pour. i’ve often imagined lighter crema coming from a spouted pf, but am not sure if this is real. this would seem to veer toward surfactant science, and the whole small-bubbles-like-a-guinness effect.

    what say ye, james? we know spouts can alter the taste characteristics of a shot, and reduce crema volume. but do they have the potential for chemically altering the science behind the size and color of crema bubbles — thus affecting when you might stop the pour?

  2. I find all this fascinating. But still have real problems with the fundamentals of having a short or long espresso. How do you make it shorter or longer? What are you changing to cause this? I guess what I am getting at, is that I thought the type of espresso produced is determined by the blend of beans. Is it possible to create a ristretto, espresso, or lungo from the same blend using the same amount of coffee for each? Confused slightly by this chart!

  3. Alistair – Well I am not sold on lungos, but I think that is beside the point.

    I think it could be easily argued that you could brew a ristretto or an espresso from 18g of coffee, depending on the weight of the final beverage. What isn’t mentioned here is the factor of time – which is obviously crucial. I guess this just serves to replace the idea of espresso as a volume, which isn’t that useful a descriptor though it is very easy to measure hence its use.

    The chart is a bit of a headache. I just work out my percentage. 100% is very much ristretto for me, and I couldn’t really cope with anything past 50%. These are working with shot times of around 25-28 seconds depending on the coffee etc…

    I need to do some more work matching taste and brew ratio, but I do think it is a valuable system.

    Ben – Thinking very much off the top of my head I would guess that pours look lighter on a spouted because the stream of coffee is thinner, which would give the impression of it being paler. I think that makes sense.

  4. I have always worked on the fact that a Ristretto is about 1 ½ ounces of drink made from a double shot of coffee. Is this wrong? When you set up your machine for a particular blend what are you using as the constant, time or volume? You have to choose one or there are too many variables. I guess the problem I am having is that there doesn’t seem to be a fixed definition of what an espresso should be, which makes all this more confusing (or is this table trying to solve to problem of the definition?).

  5. James, as you know, I posted a simplified chart on HB that hopefully makes the concept a little simpler. But please note, in order to be in harmony with established SCAA literature, I dropped the name \”extraction ratio\” in favor of \”brewing ratio.\” This is more realistic, because the simple ratio doesn\’t really control the degree of extraction, which is of course affected by time, grind, pressure etc.

    Also, to be compatible with SCAA protocol, I flipped the ratio upside down. So instead of what you said (Divide the weight of the liquid by the weight of the dry coffee and multiply by 100 to make a percentage), it\’s now the inverse (divide the dry coffee by the liquid beverage weight and multiply by 100).

    You\’re absolutely right, the chosen numbers ARE too neat and convenient, but it was a place to start. Why not make it look like coffee brewing ratios were the result of \”Intelligent Design?\” :-)

    I agree with what you said, that much (although not all) of the perceived difference between naked and spouted shots is due to people misjudging the shot \”volume\” and pulling shorter naked shots.

    Rock on, James.

  6. “I have always worked on the fact that a Ristretto is about 1 ½ ounces of drink made from a double shot of coffee. Is this wrong?”

    It’s not wrong, Alistair, it’s just hard for others to duplicate what you’ve done because it’s so imprecise. Measuring espresso volumetrically is highly variable (crema varies), so it’s far better to weigh it in grams. Ditto with “a double shot of coffee.” Is that 14g or 19g?

    “I guess the problem I am having is that there doesn’t seem to be a fixed definition of what an espresso should be, which makes all this more confusing (or is this table trying to solve to problem of the definition?).”

    I think what an espresso should be is whatever pleases you and your customers. But yes, the chart is trying to find a way for us to develop a common descriptive language. In the long run, that makes everyone more productive.

  7. Thanks for the comments Andy. I can’t believe I’d written up the way to get the ratio wrong in the post (now corrected) – I had been doing it the proper way during my measuring session.

    So – the real question is how does one practically get this working in a bar scenario – should we be pulling shots onto scales or will it remain the tool of the blender/tester/enthusiast?

  8. Jim, for me it’s the same as experimenting with different tamping techniques, extraction speeds, whatever. It’s something to experiment with after hours or first thing in the morning when I have time to weigh shots, play around etc. Then once you’ve got a feel for the ratios, you don’t need to be weighing each one.

    Alastair, for me, a ristretto is a slower extraction (ie. I actually make the grind a bit finer), and I would probably cut it off a little earlier (time-wise i mean), so it’s less liquid, sweeter, and more concentrated than espresso. And I would say that your description of 1 1/2 ounces from a double shot probably fits the bill. But again, different machines, etc mean different crema and therefore different volumes, so the weight ratios help define the beverages more accurately.

    I would have thought that from a naked pf, you get better, finer crema. And as we know from jim’s excellent articles, the finer the bubbles, the lighter the colour. So with a naked it should look lighter, which would explain peeps cutting shots shorter. Just my two cents…

  9. I still don’t really get this. If I calculate with the images in this post. It’s 75% so a ristretto. But if this is 18 grams of coffee it would have to be a double. Since the double should give a double ristretto and I where to give these ristretto out in two cups, the cups would both have 1 drop of espresso in them, let alone when it’s up to 100%

  10. Yes it is a double ristretto, around the 1 ounce mark by the looks, so if you split it into two singles you would have around 1/2oz in each cup

  11. Thank you so much for this post!  I have understood the importance of weighing espresso for consistency but was never very clear on articulation of the concept.  So helpful!

    And just because the details matter, “weight” in the second sentence of the third paragraph should be “weigh,” no?

Submit a comment