Pencil Icon

Watch your head space.

I’ve realised that, when I talk about coffee, what I say isn’t backed based on a data that’s been proven or experienced myself. This frustration has lead me to actually experiment on few things, rather than pondering and hypnotising myself that my pocket science (very little of it) is correct.

One subject that I always wanted to, but had no commitment to experiment, was the importance (if there is any) of headspace between the shower screen of espresso machines and the coffee bed.

In my perfect little world, I always believed that it did, that having some head space produced higher percentage of consistent shots, without any data to back my opinion on. I started to have firm believe because of several reasons, firstly, I got so frustrated when, even though I have the same dose, grind setting, extraction time and temperature of the machine in each shot, sometimes the results fluctuated quite abit. I started to dissect and see what variables I could control and those I couldn’t, I didn’t even bother about the uncontrollable variables, but I’ve realised that I’ve never accounted head space as a variable.

Those of you who use the Nuova Simoneli Aurelia ,would know that each group head has a brass dispersion block. Which disperses the hot water into 8 different holes for even distribution of water from the group head to the coffee bed.


Now this brass block comes in 2 different thickness, 3mm and 5mm ones. I first thought 2mm difference would only make a minute difference but the results below proved me wrong.

To compare and note the difference between the 2 different sizes of the dispersion block, I wanted to gather 2 different sets of data. First being consistency, and second, rate of flow/rate of espresso extraction.

So firstly, consistency:

I wanted to see the consistency of espresso produced in grams, with different size of dispersion block, when other variables become constant. (Impossible but I’ve tried my best!)

Machine used: Aurelia T3

Temperature: 94C

Time: 28 seconds

Basket: 18g VST

Dose: 19g

Coffee used: Jirmiwachu, Ethiopia SOE


If we ignore the fact that, 3mm block produced big shots that wouldn’t normally cut the mark, it is apparent from the data that, 3mm block produces fairly consistent shot in terms of weight , you have to remember that I’m just looking for consistency in weight of espresso produced.

In the other hand, the 5mm block shows that the shots produced are not even, it fluctuates rapidly after shot 4, it produces some values which could be considered as an outlier, and its hard to produce a mean value.

I’m no scientist, and I think people cleverer than me could give me a better answer, but here’s an educated guess. If we think of the shower screen as a perfume, and surface of our clothes as the coffee bed, applying perfume from a distance will cover broader surface area of your clothes evenly. On the other hand, spraying your perfume in close range will highly saturate a specific area. If we want the water to be evenly distributed amongst the coffee bed, it makes sense to me to have a gap between the surface of the coffee bed. I mean, we are talking about 93~94C water under 9bars of pressure, 2mm difference in my opinion, gives the water some space to actually distribute the water evenly.


*Note that spent coffee bed from 5mm leaves clear mark of the screw, where 3mm does not.

Lastly, I wanted to record the rate which the espresso was being extracted, by plotting a time/espresso weight graph. The reason why I wanted to obtain this data was to observe the pattern in which the flow of water through coffee bed was affected by the head space.

The result is quite interesting:

Now this graph was plotted using average values from 5 different set of shots, you can see from the result that, 5mm block produces espresso faster than the 3mm block; and that flow is maintained until 17 seconds, where the 3mm block actually overtakes 5mm block and produces more espresso at the end.

Now why does this happen? why does the flow rate gain momentum from 17 seconds on the 3mm block? Here’s my guess, the reason why 3mm block took longer to produce any espresso initially, is because the water starts to fill the basket and slowly, and saturates most of the coffee bed before extraction occurs.

Where as the 5mm block (this is just my opinion), the coffee bed is mostly pushed up against the shower screen, and when the extraction occurs, the water isn’t able to fully saturate the the coffee bed evenly and finds a path which allows extraction to occur without full saturation.

With my shallow knowledge, I wanted to know why there is an increase in flow from 17 seconds with 3mm block, and I came up with this.

Darcy’s law states that, flow rate of fluid increases as pressure increases, until it reaches a certain pressure (correct me if I’m wrong), so it would mean that there are less resistance created by the coffee bed than the 5mm block, allowing the water to flow through the coffee bed at an increasing rate, which means full saturation at the start of the extraction loosens up the coffee bed, and allows the water to flow through  evenly, which also benefits even extraction throughout the coffee bed?

I really enjoyed the shots coming out form the 3mm block, not saying the 5mm block didn’t produce any good shots, but 3mm block hands down delivered the consistency, and I think we all need that in our lives.   

The pocket science rant above could be all wrong, but having the data gives me boost of confidence in what I believe in, and I would love to hear other peoples thought on this. I would also love to be proven wrong, if anyone has better explanation please please comment!!

** For those of you without an Aurelia where you don’t have any interchangeable dispersion screen maybe try playing with different basket size, but keeping the dose same.

Pencil Icon

Teflon coated….(Long post)

Plethora of things are teflon coated these days. It’s on the frying pan, fabrics and recently it has been sprayed on our portafilters. It was a mission to clean the dirty dried up remains of an espresso shots, and we can’t clean our baskets and portafilters after every shot pulled. After coating the surface with teflon, only thing i do now is rinse with hot water, nothing else.

Then i came across this: 

James Hoffmann did an interesting experiment with teflon coated baskets, with a video.( Check that post first before reading on this post if you haven’t done so)

I thought by coating the basket with teflon; flow rate and extraction from coffee bed would remain the same, or would have correlation to non teflon coated basket.

I was wrong; the teflon somehow changed all things where i thought it would remain unchanged, and the video shows that.

This made me thought ‘Teflon coated portafilter is enough’, until this showed up

Yes, it is a teflon coated dispersion block for Nuova Simonelli Aurelia. This was sent to Andrew by Paul from Coffeehit (Many thanks Paul) for us to play with and test on. So on a quiet Saturday evening, i made my way to Bank, where i got to play with this little green thing 1 on 1. 

My first impression was ‘just a dispersion block with teflon’ nothing special really, except for the fact that; on the hand, it felt really smooth and silky, unlike the rough brass dispersion block.

Side by side, it looks like this

For those who does not know the function of the dispersion block on the Aurelia, put simply, hot water is jetted out from the mixing cavity at an angle; this is done so when it hits the dispersion block, equal amount of water by passes the 8 holes on the block itself, so water is distributed evenly, hence achieving good extractions from all areas of the coffee bed.

I thought having teflon on dispersion block is a good idea, as it is very very hard to clean! So i quickly fitted the teflon block on one of the group head, and started to pull some shots through and start my experiments..

Experiment 1.

Firstly, i set the grinder using our house blend, through the group head with no teflon dispersion block. When i started to get good shots through, i measured the dose, time of extraction and the brew weight on both group heads ( one with teflon one without)

So have a look below:

 So for each shot, i tried to get these variables constant:

Dose : 21g

Time: 25sec

Tamping technique 

Few things that i want to point out, or you guys already noticed is that, the teflon shots were pulled from a naked portafilter, where as brass had spouted portafilters, also, i can’t guarantee that each group head was at equal temperature, but i would imagine that it was within ± 0.5℃.

Even with these factors did contributed to the fact that; my experiment was not entirely ‘Uniform’ or ‘Accurate’ giving fair results, the results did surprise me.

17g difference in brew weight is quite big. Very big i would say. Also, when i was observing the shots coming through the group with teflon, it seemed to me that the water was not removing much soluble solids from the coffee bed, to me, it seemed to be channeling rapidly even though the shot was aligned at the centre,

here is a picture:

With my eyes only, i couldn’t explain what was happening, so i decided to let my palates take charge. The shot from the brass dispersion block tasted of everything i desired; floral notes along with apricot notes coming through, giving way to sweet chocolate notes finished with very nutty after taste lingering in my mouth.

On the other hand, the shot from the teflon dispersion block was quite unpleasant.

It felt to me as if i was drinking sour americano, flavors were underdeveloped, it was very watery in my mouth, so much dilution, i could only taste slight nutty notes on that shot.

After tasting each shot, I’ve realised that, somehow, the teflon dispersion block was not doing it’s job. It wasn’t letting water pass through the 8 holes equally, or evenly so even extraction could happen, instead channeling was occurring, not much of the water was going through the coffee bed removing soluble solids.

I had no clue why this was happening, i repeated the same experiment over 10 times but i had the same results as above.

Experiment 2

I decided to take a look on the spent coffee bed. What this tells me is how much soluble solids have been readily removed from the coffee bed. So if i pulled; nice evenly extracted espresso, the spent coffee bed should have nice graduation of color. The top of the coffee bed( where it has first contact with water) should be lighter than the bottom, and inside the coffee bed, should have nice color distribution( Light to dark from top to bottom).

Here are some pictures of the spent coffee bed:

 The spent coffee bed from teflon is still quite dark, where as coffee bed from the brass is noticeably lighter, showing that soluble solids has been removed from the top of the bed.

 From the second picture, i wanted to show the color inside the coffee bed, this picture, just like the first shows that the coffee bed from the teflon is still dark, no change in colour between coffee bed, not much solids were removed.

I don’t know, I just don’t. I know that teflon dispersion block causes or contributes majorly to channeling, therefore producing under extracted, sour shots. But i just can’t figure out why…

If i was to take an educated guess, i might say that; maybe there is something to do with friction of the dispersion block. Brass, even though on hand it feels smooth, on microscopic view of the surface will be very rough looking, so causing friction when water is flowing on it’s surface.

But if we cover that surface with teflon, are we taking away all that friction or most of it? Making the surface ‘slippery’ for the water?

I don’t know, i guess it’s up to the scientist for this one… it was enjoyable though, in the mean time, i will make sure i give these brass friends nice polish.

Comments and suggestions are welcome!! I would like to hear what other people think about this alot!

Sang ho