Measuring the extraction of Clever Coffee Dripper and Aeropress brews (Mixed Phases Methods), using a refractometer.

Coffee preparation techniques besides espresso like pourover.
thebeanvagrant
Posts: 4
Joined: 3 years ago

#1: Post by thebeanvagrant »

Greetings :D

Wondering whether someone can advise...

I routinely use a refractometer (VST Lab III specifically) for my manual brew methods at home, to measure TDS and calculate extraction %.

I find it extremely useful for dialling in both pourover filter percolation brews (such as Kalita Wave, V60, and Chemex, etc), and immersion brews (French Press).

However, for some time I've been uncertain about the measurements/results for my hybrid brewers such as Clever Coffee Dripper and Aeropress, since reading Jonathan Gagné's post on Coffee Adastra where he calls these brew methods "Mixed Phases Methods":

https://www.google.com/amp/s/coffeeadas ... ields/amp/

The VST refractometer and coffee tools software appears to cater for measuring such methods, as it specifically lists them under 'Immersion'.

However, Jonathan's post describes that a special procedure and equation is necessary to get a proper measurement with these Mixed Phases Methods, and "This particular equation is not currently supported by the VST application".

However, I don't have the time or inclination to undertake the additional complicated procedure and manual maths he details (frequent refractometer measurements and brew notes, etc, already take up enough time and effort when making a coffee!).

So I guess my questions are:

Does using the immersion setting on the VST refractometer software for these brew methods give a reasonably accurate reading, assuming fairly normal recipes are used (and samples are syringe filtered, of course), albeit perhaps not as completely accurate as the procedure Gagné describes?

Or is any result I get from the VST on the immersion setting for CCD and Aeropress going to be so inaccurate that it is basically completely unreliable and misleading?

Thanks :)

Note as per guidelines:
I originally posted this question over on Reddit a week or so ago, but had no response, so am reposting the question here :)

millmountain
Posts: 203
Joined: 4 years ago

#2: Post by millmountain »

Hi, I don't have a VST, but if I distill your post down to the following:
thebeanvagrant wrote:I find it extremely useful for dialling in both pourover filter percolation brews (such as Kalita Wave, V60, and Chemex, etc), and immersion brews (French Press).

However, for some time I've been uncertain about the measurements/results for my hybrid brewers such as Clever Coffee Dripper and Aeropress, since reading Jonathan Gagné's post on Coffee Adastra where he calls these brew methods "Mixed Phases Methods":

....

Does using the immersion setting on the VST refractometer software for these brew methods give a reasonably accurate reading, assuming fairly normal recipes are used (and samples are syringe filtered, of course), albeit perhaps not as completely accurate as the procedure Gagné describes?

Or is any result I get from the VST on the immersion setting for CCD and Aeropress going to be so inaccurate that it is basically completely unreliable and misleading?
...then I am left with a question about what you actually want to do with the result. The title of the article you linked is "Measuring and Reporting Extraction Yield," and scanning through it seems that all of the formulae and measurements for mixed-phase methods is about getting the EY correct. I suppose you would only need to get the value absolutely correct if you are going to use their coffee control chart in a strict way.

So isn't a relative value sufficient for comparison in personal use, as long as you are measuring in a consistent way?

Advertisement
thebeanvagrant (original poster)
Posts: 4
Joined: 3 years ago

#3: Post by thebeanvagrant (original poster) »

Hi,

Thanks for your reply :)

Yes, you could make the case that as I'm using my refrac in a personal rather than professional or scientific setting, I don't need 100% accuracy or certainty, which is true.

And that as I use the refrac in a careful and consistent way, the readings and measurements obtained for these types of hybrid brew methods would still give me useful information by comparing each measurement relative to the others.

However, when using the VST to measure my standard pourover and immersion brew devices, I feel pretty certain that the results are very accurate, which gives me a great deal of confidence when using this information as part of my dialling in process.

And also, I'm confident that the levels of TDS and extraction can be compared between and across the various different brew devices measured.

Therefore, a fairly high level of accuracy is important to me when using the refractometer.

And ideally, I'd like the same level of confidence when measuring the hybrid 'mixed phases methods', so I can be reasonably certain of the results, and be able to compare the results from the hybrid methods relative to the standard percolation and immersion methods with a good level of certainty.

I would have assumed that I could, as the VST software lists CCD and Aeropress under the immersion setting, and it was only after reading the article I became uncertain.

So, although I'm aware of how well respected Gagné's work is, and don't doubt it's validity, I wondered whether others well versed in using refractometers and/or the VST specifically had any thoughts on the subject.

- Whether those users felt the complicated process and formula described in the article is truly necessary for the average user, or whether they have found reasonably high levels of accuracy that they trust, simply by measuring as normal using the immersion setting (with samples syringe filtered).
- And whether they feel the results can be compared fairly confidently against all the other standard percolation and immersion brew devices.

Basically I'm wondering whether people feel the results measured with the VST for these hybrid devices are fairly reliable and certain for general purposes or not - and whether Gagné's method perhaps just gives ultra-accuracy for those who require it?

Thanks again :)