Coffee grounds samples for laser particle sizing analysis, May 2013

This post follows on directly from the previous post (and other recent posts) about grind, grinders, grind size, and the laser particle sizing tests I’m having carried out on a set of grind samples.

A lovely customer, @mathewsmith1, to whom I’m extremely grateful, is making this possible, and I admit I’m very excited to see the results, which I think will be fascinating!

Ideally, I might have liked to a compare a washed to a full natural process coffee, but a suitable natural was not available at quite the right time, so the comparison/s is between two completely different washed, filter profile coffees, which will be equally illuminating, I’m sure.

The grounds being tested are all in the fine-filter through to coarse FP range. 

Below are my notes of the details of the samples submitted, the testing of which is currently in progress.

As ever, this is just very casual research I’m doing to satisfy my own curiosities and obsessions with everything coffee, and to try to learn more, some of which I’m putting ‘out there’ in case it happens to be of interest. As such, much of the specifics are over my head, and my interpretation of the subject is not necessarily accurate, relevant, or anywhere near as properly or scientifically conducted or presented as it might be by those actually skilled in these matters, for which I apologise. 


Two coffees (coffees roasted by James’ Gourmet Coffee Roasters):

1) Rwanda Koakaka Coop Filter profile (Washed process) 

Roaster’s notes:

“Rwanda Koakaka 100% Red Bourbon

Late harvest, super fruity balance

Farm 1,316 Cooperative members

Town/City:  Huye

Region:  Karaba district, Southern Province  

Process: Fully Washed and dried on screens

Farm Owner: Karaba Coffee Growers Cooperative (Koakaka)

Varietals: Red Bourbon

Altitude:  1,700 – 2,000 Meters above sea level”

2) Colombia Finca Santa Barbara Lot 238 Filter profile (Special, unique washed processing for this Lot – I have additional details from the roaster and farmer.)

Roaster’s notes:

“Farm: Finca Santa Barbara

“La Joyeria”= The Jewelry

Owner Pedro Miguel Echavarria

Varietal: 100% Colombia (Caturra and the Timor Hybrid. It is highly productive and resistant to coffee rust. First released in 1982 Expect to see this ALOT MORE AS CLIMATES CHANGE)

Area :Antioquia


MASL: 1400-2000 with an average of 1630 for this lot”

Three grinders:

My home Mahlkonig Tanzania (T) (main focus of test). An essentially brand new Tanzania with new burrs (having only seen light domestic use for two months plus some old beans through for seasoning – maybe 5-8kg max total). Still at original factory calibration.

My home Rancilio Rocky (R). My (largely retired) domestic filter and FP grinder for last 5 years with original burrs and only light use – perhaps 40-80kg total. True zero on my grinder is roughly +5 dynamic/running and -1 static.

Santos (S). The cafe’s FP grinder. Set at a coarse FP setting. Ancient model. Fitted new burrs about 8 months ago. These have maybe seen maybe 60kg +/-20 since then.

Two roast ages:

Fresh Rwanda Koakaka and Colombia Santa Barbara Lot 238 for majority of samples (Roast date 1st May 2013 for both coffees. Samples ground on the 5th and 6th May). Sealed one way valve bags. Both bags opened (and subsequently resealed) at the same time, about 24-48 hours before grinding samples, and treated to exactly the same conditions at all times.

One sample of much older Koakaka (Roast date: 28th March 2013 with the bag also having been open for about 4-5 weeks already, although resealed during this period whenever not in use) (admittedly, and necessarily, a different roast batch, so the precise filter profile may have been adjusted accordingly by roaster between the two batches, but the same identical bean, at a similar or even identical filter profile, nonetheless).

Two batch sizes:

All samples are at 15 grams (into grinder) (samples are therefore equivalent to the amount of coffee that might be used for an actual, small batch, brew size).

Except, in addition, one larger sample of 30 grams (I, T7) (fresh Koakaka).

The aim:

To accurately determine and compare the precise grind sizes in microns, the % amounts of these particle sizes (including the average, and median sizes), and therefore the shape of the distribution curves, produced by each grinder with each sample submitted. The different grind settings being tested are from fine filter, through medium, to coarse on the Tanzania, around fine to medium filter on the Rocky, and coarse only on the Santos.

To test all submitted variances of the two coffees (freshness, batch size, etc) at one same setting on one grinder (T7), as well as at the other various settings across the 3 grinders.

Detail (excuse the rather repeated waffling):

To compare two different coffees of exactly the same freshness at the same grind setting (T7).

To compare the effects on the distribution curve with just one coffee (fresh Koakaka) over a variety of grind settings, on three different grinders (settings T4-9, R24, 28 and 33, and S coarse) (some reports often indicate progressively more fines at finer filter grind settings as opposed to coarser settings for grinders similar to the Tanzania).

To establish this size and spread of particles at the different settings, in order to better understand the current calibration of the grinders, to compare grinders’ distribution curves both individually, and also relative to each other, and to better understand the grinders’ settings, as well as the brewing and taste results.

To compare the effects on the same identical bean (Koakaka), at the same grind setting on the Tanzania (T7), when ‘stale’ compared to when fresh.

To establish the effects on the distribution curve, of a larger single dosed weight of beans (100% more) (equivalent to a larger batch size) going through the Tanzania, of the same identical coffee (fresh Koakaka), at the same freshness, and at the same setting (T7), as also tested for the other samples.

To see the distribution curve produced by the cafe’s Santos, also with the same coffee (fresh Koakaka) at the same freshness, at its coarse FP setting.


Many reports indicate that the Tanzania settings generally approximately correspond to the average size in microns at each setting, for example T6 might equal 600 microns. Although, as there will be a variance between the calibration of each different grinder, this might not always be true, and so these tests will confirm/clarify this in relation to my grinder specifically, at its current calibration.

Samples for Tanzania and Santos fed into already running burrs. Rocky samples fed into static burrs, then switched on. This accurately replicates my current normal usage for each grinder.

For all samples, a minimum of 4 grams of a bean precisely identical to the sample, were purged through on each new setting on each grinder before taking the main sample for that bean and setting, in order to try to ensure the integrity of the sample, and achieve an accurate, uncontaminated sample (applying this to every single sample – different coffees, roast dates, grind settings, dose sizes, etc).

Grind samples all taken on 05/05/13. Except for the Santos – this was taken one day later on 06/05/13.

Samples then due to be analysed at the lab at some point during the following month or so (all analysed on the 29th May in the end).

The same zip-lock plastic bags used for the containment of all samples to try avoid the samples being compromised in any way, and to standardise the samples.

Samples ground directly into these bags, then sealed.

It’s my understanding that the lab will separate each sample into several portions, and thereby take several readings from each sample – to create a more accurate overall picture of each sample. This should maximise the overall accuracy of the readings, and the validity of the analysis as a whole, by counteracting and minimising any effects of static, mixing, settling, or any other factors that might result in any declassification of the grind particles within each sample as a whole.

The samples (13 in total):

A)  T4

B)  T5

C)  T6

D)  T7

E)  T8

F)  T9   (all fresh Koakaka)

G)  T7   (old stale Koakaka)

H)  T7   (fresh Finca Barbara)

I)  T7   (larger 30 gram weight of fresh Koakaka)

J)  R24   (fresh Koakaka)

K)  R28   (fresh Koakaka)

L)  R33   (fresh Koakaka)

M)  S   (coarse FP setting – fresh Koakaka)

24th may 2013 evening pours

A gratuitous, almost completely irrelevant, and quite unexceptional picture of brewed, lovely, Santa Barbara.

Aside from this interesting diversion of grind size analysis, brewing at home recently with the Tanzania has rather turned everything on its head. With this grinder now as a sound benchmark to work from, and whilst readjusting and reassessing all the parameters for each filter brew method to get to grips with the new grinder, the infinite range of different techniques possible with each brew method has become more evident, and has been brought into sharper focus, than ever…

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: