Water flow rate - any research on effects?

la marzocco, synesso, simonelli, cimbali etc

Water flow rate - any research on effects?

Postby Rich Westerfield on Mon Nov 27, 2006 8:58 pm

A top barista traveling through town noted that our water flow was considerably faster than what she was used to, possibly twice as fast.

We have not experimented with flow as a standalone variable (not in conjunction with grind). We've tried Googling for info this evening, but as of yet haven't found any data on how flow rate would affect the brew.

(In case it needs mentioning, again, this has nothing to do with grinds - this is measuring the rate at which you can fill a cup with H2O through a grouphead. It's not a portafilter or screen issue.)

She suggested the accelerated flow rate leads to a slight underextraction. That makes sense on the surface, but the shots we've been pulling seem OK and taste the same as we've always pulled. We don't see that our crema is any lighter, either.

We respect this person's palate, so we're not discounting her comments by any means and indeed find this info helpful. It's just a surprise to us since it's not something we've ever adjusted. If we're off, it would seem it's a matter of a degree that's all but undetectable to us but apparent to her.

We're going to play with slowing the flow down and seeing how it affects the shot quality, but was wondering if anyone here had any research you might have done on this issue that you'd consider sharing.
Last edited by Rich Westerfield on Tue Nov 28, 2006 12:39 pm, edited 2 times in total.
Rich Westerfield, Co-owner
aldocoffee.com
I'm just here to regulate funkiness.
Rich Westerfield
 
Posts: 565
Joined: Sat May 13, 2006 8:23 pm
Location: Pittsburgh

Postby nick on Tue Nov 28, 2006 12:31 am

Rich, you can't really "slow the flow down." I mean, I don't know what you were intending to do, but what you want is flow-restrictors on the groups.

In an espresso machine brew environment, there are two points of flow-restriction: the plumbing, and the coffee. If you compare the rate of flow straight out of the pump, or straight out of your water source, to the rate of flow out of the shower-screen (without a portafilter), you see the machine's flow-restriction. Then if you compare that shower-screen rate of flow, vs. the rate of flow coming out of the spouts during brewing espresso, the "delta" there represents the coffee's flow-restriction.

So when Lindsay talks about your machine having a "faster" water flow, I'm guessing that your machine has no flow-restriction (a.k.a. "gicleurs").

A small screw with a very small hole through the center can be screwed into the end of the copper tubing that runs out of the "neck" of the group, and into the 3-way solenoid, on each group. This results in a lower overall flow rate without coffee loaded into the portafilter.

That overall flow rate will, somewhat obviously, still be higher than that of the espresso-pour. The improvement that gicleurs provide is that the water pressure is more gradually introduced to the coffee puck. It offers, in many people's opinions, most if not all of the benefits of pre-infusion, without needing actual pre-infusion.

However, I have this weird related theory as well: I hypothesize that because of the flow-restrictor in-line, it allows the coffee puck to devote more energy into the extraction, instead of acting as the main flow-restrictor. It's hard for me to articulate what I mean here, but it's this hunch that I have. It sort of doesn't make sense (9 bars is 9 bars), but it's my hypothesis nonetheless.

What seems pretty clear though, is that flow-restrictors greatly effect the first few seconds of extraction, which, by nature of the gradual build-up of pressure, presumably establishes a better puck-environment for the rest of the extraction.
Nick Cho
nick
 
Posts: 1335
Joined: Thu Jun 23, 2005 8:15 pm
Location: San Francisco, Coffeefornia
full name: Nicholas Cho
company: Wrecking Ball Coffee Roasters
: http://nickcho.com
: http://wreckingballcoffee.com/

Re: Water flow rate - any research on effects?

Postby Andy Schecter on Tue Nov 28, 2006 2:37 am

Yes, there have been many threads about this, including a long one on alt.coffee here:
http://tinyurl.com/srv99

The "Cliff's Notes" answer is that the high flow rate
(1) increases the probability of channeling, and
(2) requires a coarser grind to obtain the proper shot timing.

Number one is an obvious negative and number two possibly explains her suggestion about the underextraction.
Last edited by Andy Schecter on Tue Nov 28, 2006 2:35 pm, edited 1 time in total.
-AndyS
Schectermatic(tm), the oldest, most trusted brand of espresso grinder shnozzola
http://www.flickr.com/photos/andy_s/sets/
Andy Schecter
 
Posts: 449
Joined: Mon Jul 04, 2005 4:59 pm
Location: NY

Postby Andy Schecter on Tue Nov 28, 2006 2:40 am

nick wrote: I have this weird related theory as well: I hypothesize that because of the flow-restrictor in-line, it allows the coffee puck to devote more energy into the extraction, instead of acting as the main flow-restrictor.


Incomprehensible.

nick wrote:
What seems pretty clear though, is that flow-restrictors greatly effect the first few seconds of extraction, which, by nature of the gradual build-up of pressure, presumably establishes a better puck-environment for the rest of the extraction.


Comprehensible and probably correct!
-AndyS
Schectermatic(tm), the oldest, most trusted brand of espresso grinder shnozzola
http://www.flickr.com/photos/andy_s/sets/
Andy Schecter
 
Posts: 449
Joined: Mon Jul 04, 2005 4:59 pm
Location: NY

Re: Water flow rate - any research on effects?

Postby Rich Westerfield on Tue Nov 28, 2006 6:57 am

AndyS wrote:
Yes, there have been many threads about this, including a long one on alt.coffee here:
http://tinyurl.com/srv99

The "Cliff's Notes" answer is that the high flow rate
(1) increases the probability of channeling, and
(2) requires a coarser grind to obtain the proper shot timing.

Numnber one is an obvious negative and number two possibly explains the suggestion about the underextraction.


Andy,
Thanks for the pointer. That thread was exactly what we were looking for (we think - haven't had time to read that whole thread yet).

UPDATE: Read the whole thread... all 104 posts. Wow. Whether the gicleur makes a difference or not, I certainly feel smarter. Or maybe it's that my head hurts... Anyway, that was fascinating.


However, when we toss in a full naked pf there is no apparent channelling. Shots poured with a double spout are still within our desired range with the proper grind.

That's why learning this was a surprise to us - we're not seeing any difference in the pour or in the cup (neither are customers).


Nick,
If we'd been calling them "flow restrictors" all along instead of "gicleurs", I might have read all those earlier posts on 'em. :P

Have to work on one of the groups this week anyway, so will look into this. Just went searching on earlier threads from last year where PeterG and Tacy talked about swapping out to .6mm gliceurs - but Peter cautioned it might not be worth it for the work involved (and the 6mm's were hard to find). Is the .6 what you're using at murky?
Last edited by Rich Westerfield on Tue Nov 28, 2006 1:27 pm, edited 1 time in total.
Rich Westerfield, Co-owner
aldocoffee.com
I'm just here to regulate funkiness.
Rich Westerfield
 
Posts: 565
Joined: Sat May 13, 2006 8:23 pm
Location: Pittsburgh

Postby Michael_Teahan on Tue Nov 28, 2006 9:38 am

Higher flow rates may impact channeling but will have no effect on flow rates through the coffee. Piston machines have no jets whatsoever and have consistent flow rates.

Coarse coffee may increase channeling in high flow rate machines, not decrease it.

High flow rates are more likely to impact the initial distribution of water on some machines, due to the the distribution design of the dispersion plate and/or screw.

Smaller jets not only slow down the flow of the initial splash, they also allow for more gradual increases in pressure over the filter bed, mimicking pre-infusion.

Once saturated, and assuming the grind and quantity are correct, the jet not longer impacts the extraction as pressures will always seek equilibrium, which takes just a few pounds of pressure.

Michael
Michael_Teahan
 
Posts: 18
Joined: Fri Oct 14, 2005 3:05 pm
Location: Los Angeles

Postby James Hoffmann on Tue Nov 28, 2006 10:01 am

Michael_Teahan wrote:
Coarse coffee may increase channeling in high flow rate machines, not decrease it.


I don't suppose you could please explain this a little more? It is something I want better to understand.
James Hoffmann
 
Posts: 619
Joined: Mon Jun 27, 2005 8:34 am
Location: London, UK
full name: James Hoffmann
company: Square Mile Coffee Roasters
: http://www.squaremilecoffee.com
: http://www.jimseven.com

Postby Michael_Teahan on Tue Nov 28, 2006 12:34 pm

As the coffee expands with moisture in increase in resistance forcing the coffee to find other pathways (dry) to pass through the filter bed. Large partical size means less resistance and the greater potential for the water to pass through (channel) through the bed before it seeks out other pathways.

Smaller particals increases the resistance of the filterbed, increasing the pressure in the chamber above the puck and, ironically, forcing a more even pressure across the surface area earlier in the process. Coarse coffee will delay this effect.

Because the flow rate through coffee is always less than that through the jet, the effect of the jet once the bed is saturated approaches zero.

Michael
Michael_Teahan
 
Posts: 18
Joined: Fri Oct 14, 2005 3:05 pm
Location: Los Angeles

Postby nick on Tue Nov 28, 2006 12:43 pm

Michael_Teahan wrote:Because the flow rate through coffee is always less than that through the jet, the effect of the jet once the bed is saturated approaches zero.

To expound on what I wrote before, my head agrees with your posts 99%. There's a lingering 1% in my mind though.

On an un-restricted machine, the effective pressure at the portafilter/puck is fairly consistent with the pump pressure reading (i.e., 9 bars on the gauge is just about 9 bars of real pressure on the puck).

However, what about on a flow-restricted machine? Is the saturated pressure on the puck truly the same as on an unrestricted group? Have people out there put a portafilter-pressure-gauge on a 0.6mm gicleur'ed group? Is there really no pressure-drop across the jet?

Again, I can't really explain it... but I have this tiny feeling in the pit of my stomach that there's actually something more going on with flow restrictors than what's conventionally understood... and I'm 99% sure that I'm wrong about that. :wink:
Nick Cho
nick
 
Posts: 1335
Joined: Thu Jun 23, 2005 8:15 pm
Location: San Francisco, Coffeefornia
full name: Nicholas Cho
company: Wrecking Ball Coffee Roasters
: http://nickcho.com
: http://wreckingballcoffee.com/

Postby Michael_Teahan on Tue Nov 28, 2006 3:01 pm

The jet has no effect on the ultimate pressure inside the group, only the rate at which it rises. The pressure on the surface of the puck is a factor of the supplied pressure on the pump side of the jet, less the loss of pressure through the filter bed.

The pressure inside the group is always slightly less than the rated pump pressure.

Think of it this way:

You fill up a balloon with an air tank using a pressure regulator set to 5psi. The tube used to fill up the ballon is 1mm across. If the tube is smaller or larger, it will change the rate at which the balloon fills, but the pressure is never more than 5psi.

Now, picture a small whole in that balloon. If the whole is larger than the tube that fills it, the balloon will never inflate. Espresso leaving the puck is acting as that small leaking hole. If the rate of extraction allowed by the coffee is slower than the rate of flow of the jet, the difference between the two is the actual pressure in the group. As the difference increases, the effective drop in pressue over the puck approaches zero.

Inside the group, that pressure will rarely be 9 bar unless the restriction of the puck is so great as to virtually stop. It is always less; how much so is dependant upon the difference between the size of the jet and the extraction rate.

Sounds like I just complicated everything.

Ultimately, jets will not affect extraction rates if the grind and quanitiy is correct.

Michael
Michael_Teahan
 
Posts: 18
Joined: Fri Oct 14, 2005 3:05 pm
Location: Los Angeles

Postby nick on Tue Nov 28, 2006 6:30 pm

Michael, thanks for engaging me in this. If I might continue the dialogue...

The thing is, everything you've said makes sense, and relate to things that I've thought of.

However, if I might approach the issue at a slightly different angle:

I have a LM 3EE Linea, that at about four years of age, needed a de-scale (bad water... don't ask :cry:). The main issue that I had was that we have 0.6mm gicleurs on each of the three groups, and two of the groups had low-flow issues, which were caused by scale and particulate buildup in the gicleur orifice itself.

So to step away for a second, the assumption is that all flow-restriction does is to delay the introduction of full-pressure, but it does not affect the full-pressure. So in theory, if you had two groups, one with a flow-restrictor and one without, one might take a little longer to get to full-pressure, but with all other factors held as constants, once full-pressure is reached, the flow rate should be the same.

That's not what happened.

This is certainly an extreme extrapolation, but I'm not proud to admit that this Linea was in fairly poor condition (until it was descaled). Shower-screen (no portafilter) flow was down to a little less than half the normal flow rate on the "good" group. When I would pull a shot with this group, two things happened: it took FOREVER for even the first drops to emerge from the group (about 30 seconds), and... the pour never achieved "normal" flow rates.

Now I know what you're thinking: this is like your example, where the hole in the balloon is too big... or, more accurately, the incoming air is so slow that the same diameter hole results in no inflation. The rate of water flow above the coffee is too low; lower than the espresso would pour. In this case, we're sort of "off the charts," and the extraction won't happen properly.

But here's where this leads me: (I can't believe I'm doing this... but this is fun! I'm such a nerd! :P)

Let's make a graph. X-axis is the diameter of the gicleur orifice. Y-axis is the rate of flow, with the unit of measure being seconds that it takes to extract a 60 mL double, AFTER the first drops emerge (or, more specifically, after full-pressure is reached). Let's assume all other elements to be constant (grams of coffee, basket shape, grind, etc.). After I created the graphs, I realized the flaws of using that particular unit for the Y-axis... but you get the point... bear with me. :-)

Graph 1.
Image
Until your last post (above) in this thread, one might guess that this is what you meant: that the effective rate of flow does not change, no matter what the gicleur is, once, of course, full-pressure is reached. This can't be true, because at a 0.0mm orifice, you wouldn't have the same flow as a 1.5mm orifice. You clearly clarified this issue in your last post. However, let's continue anyway...


Graph 2.
Image
This can't be right either. This might be what someone could assume I mean in this discussion, but it's not.


Graph 3.
Image
This might be a graph that you can agree with. You contend that the effective rate of flow does not change due to gicleur orifice, hence the flat and constant curve at 0.4mm+. Let's arbitrarily estimate 0.4mm as the gicleur diameter that creates that zero-pressure-drop point. In other words, at 0.4mm, the flow of water above the coffee is equal to the flow rate after the extraction. Going right-to-left, starting at that point, there is a curve upwards towards infinity, as the X-axis approaches zero.


Graph 4.
Image
This is what I'm claiming. I don't believe that above a certain gicleur diameter, that the flow-rate would be a constant. Because of the necessary curve left of 0.4mm, I have to believe that there is somewhat of a curve at plot points right of 0.4mm as well.

Therefore, I believe that there is a measurable difference in effective flow and/or pressure-drop between heavily flow-restricted groups vs. unrestricted groups... not just in pressure ramp-up. Just how much of a difference? That's the question.

On a related note, I do have a Linea 3EE that we use strictly for training, with 0.6mm restrictors on each of the three groups. It would be fairly easy for me to compare a 0.6mm gicleur to a 0.8 (or 1.0) to an un-restricted group.

Thoughts?
Nick Cho
nick
 
Posts: 1335
Joined: Thu Jun 23, 2005 8:15 pm
Location: San Francisco, Coffeefornia
full name: Nicholas Cho
company: Wrecking Ball Coffee Roasters
: http://nickcho.com
: http://wreckingballcoffee.com/

Postby Andy Schecter on Tue Nov 28, 2006 7:31 pm

Michael_Teahan wrote:The jet has no effect on the ultimate pressure inside the group, only the rate at which it rises.

<snip>

Inside the group, that pressure will rarely be 9 bar unless the restriction of the puck is so great as to virtually stop. It is always less; how much so is dependant upon the difference between the size of the jet and the extraction rate.


These two paragraphs, of course, contradict each other.

Michael_Teahan wrote:Ultimately, jets will not affect extraction rates if the grind and quanitiy is correct.


Not sure why you put "ultimately" in there, it pretty much takes the meat out of your argument.
-AndyS
Schectermatic(tm), the oldest, most trusted brand of espresso grinder shnozzola
http://www.flickr.com/photos/andy_s/sets/
Andy Schecter
 
Posts: 449
Joined: Mon Jul 04, 2005 4:59 pm
Location: NY

Postby Andy Schecter on Tue Nov 28, 2006 8:19 pm

[edited a couple times to try and eliminate my dumb stuff]

nick the geek wrote:On an un-restricted machine, the effective pressure at the portafilter/puck is fairly consistent with the pump pressure reading (i.e., 9 bars on the gauge is just about 9 bars of real pressure on the puck).

However, what about on a flow-restricted machine? Is the saturated pressure on the puck truly the same as on an unrestricted group? Have people out there put a portafilter-pressure-gauge on a 0.6mm gicleur'ed group? Is there really no pressure-drop across the jet?


Yes, there is a pressure drop across the gicleur. It is small for realistically-sized (unplugged) gicleurs, but it is not zero. Obviously the pressure drop is a lot greater for a 0.6mm gicleur compared to a 1.0mm.

When I measured it I got a range between about 0.1 bar and 0.5 bar. After the initial preinfusion period, the pressure drop gradually increased as the shot progressed. I believe that it increased because (at least when I tested it) the puck steadily eroded as material was flushed away. This reduced the puck's restriction and increased the flow rate. Increasing the flow rate across the gicleur results in a higher pressure drop (the pressure drop is proportional to the square of the flow rate).

nick wrote:Graph 4.
Image
This is what I'm claiming. I don't believe that above a certain gicleur diameter, that the flow-rate would be a constant. Because of the necessary curve left of 0.4mm, I have to believe that there is somewhat of a curve at plot points right of 0.4mm as well.

Therefore, I believe that there is a measurable difference in effective flow and/or pressure-drop between heavily flow-restricted groups vs. unrestricted groups... not just in pressure ramp-up.


This graph is pretty much jives with what I measured and with the theory as I understand it.

nick wrote:I can't really explain it... but I have this tiny feeling in the pit of my stomach that there's actually something more going on with flow restrictors than what's conventionally understood.


To recap, I believe the pit of your stomach may be on to something. I believe that gicleurs have at least three readily identifiable effects:
(1) they regulate the initial soaking of the coffee grounds ("preinfusion")
(2) they provide a small pressure drop which increases as flow rate increases. this means that the pressure your puck sees is a little lower than you may think
(3) the pressure drop may slightly increase as the shot progresses

Some or all of these things are ones that a sensitive barista might possibly sense in the pit of his stomach or even on the tip of her tongue.
Last edited by Andy Schecter on Thu Nov 30, 2006 3:06 am, edited 1 time in total.
-AndyS
Schectermatic(tm), the oldest, most trusted brand of espresso grinder shnozzola
http://www.flickr.com/photos/andy_s/sets/
Andy Schecter
 
Posts: 449
Joined: Mon Jul 04, 2005 4:59 pm
Location: NY

Postby Michael_Teahan on Wed Nov 29, 2006 11:06 am

Andy,

The two paragraphs aren't contradictory, they refer to two seperate issues.

If measuring the pressure differential effect of the jet ising a blind filter, the differential approaches zero as the group pressurizes. The pressure drop on the output side of the jet under extraction, therefore, is a function of the loss of pressure as the water flows through the espresso. The rate of drop at that point is affected by the jet size only to the extent that it can or cannot exceed the flow rate of the water through the coffee.

If the flow rate is excessive, jet size matters under extraction. If the flow rate is appropriate, jet size matters little. To the extent that you have measured small pressure differences with different jets begs the question of whether or not a difference in extraction was noted as well.

Most Italian machines spec 8.5 to 9.0 bar depending upon the coffee used, so a small degree of variation is to be expected without consequence to the coffee. Additionally, the flow rate for single shots (yes, people still pull single shots in the real world) would dramatically affect the pressure differential. Qualititative differences between singles and doubles are not likely associated with this differential, but due to other factors if at all.

Jet size DOES matter under pre-infusion conditions.

If the jet is sufficiantly clogged to the point where the flow through the coffee is competing with the flow through the jet, you will see differences in extraction rates. The theories posed from a jet design standpoint assume a properly running machine.

Everything Andy refers to regarding the effect of jets on the group, with minor quibbling about the effect under loaded extraction, is accurate.

Sorry if I responded to Andy's comments before getting into your post, it always happens when I get 'called out' on something.
Michael_Teahan
 
Posts: 18
Joined: Fri Oct 14, 2005 3:05 pm
Location: Los Angeles

Postby Andy Schecter on Wed Nov 29, 2006 5:26 pm

Michael_Teahan wrote: To the extent that you have measured small pressure differences with different jets begs the question of whether or not a difference in extraction was noted as well.


I have not detected differences in taste related to small pressure changes. Extensive testing with better trained palates might detect systematic differences, I don't know.

I have noted large (~8 sec) differences in shot timing based on changing from the equivalent of a 0.6mm to a 1.0 mm gicleur. This is due to the change in the initial coffee wetting, not to any pressure drop in the subsequent extraction.

Michael_Teahan wrote:Jet size DOES matter


I know that. In fact, I get half a dozen emails a day from kind folks offering to help me do something about it.
-AndyS
Schectermatic(tm), the oldest, most trusted brand of espresso grinder shnozzola
http://www.flickr.com/photos/andy_s/sets/
Andy Schecter
 
Posts: 449
Joined: Mon Jul 04, 2005 4:59 pm
Location: NY

Postby Mark Prince on Wed Nov 29, 2006 10:29 pm

Andy the Gear Head wrote:To recap, I believe the pit of your stomach may be on to something. I believe that gicleurs have at least three readily identifiable effects:
(1) they regulate the initial soaking of the coffee grounds ("preinfusion") and
(2) they provide a small pressure drop which increases as flow rate increases. this means that the pressure your puck sees is a little lower than you may think
(3) the pressure drop may slightly increase as the shot progresses


Much of this discussion is way over my Gr.10 level physics learnin' but I'll say this - I've been very much witness to all three points above since moving to a 0.6mm gicleur on my La Marzocco. But it's only noticable when the pack and prep is spot on.

I'm very much an empirical guy - empirical evidence (believe what you witness) - as such, I also am much more in Nick and Andy's camp on all of this discussion. Flow restriction has improved my own shot performance, and going back to what was probably the best shot of espresso I've ever had (I have a bit of a, uh, photographic memory for specific tastes), it came from an Elektra Micro Casa lever machine, and things just start to make so much more sense these days - I always thought it was just a fluke with the stars aligned, that kind of thing. But given how the preinfusion works on spring lever machines, and also going with Jim Schulman's theory on how a solid "block" of water falling on a puck instead of a jetted (even flow restricted jet) of brew water may improve overall shot quality, maybe it wasn't so much of a fluke after all.

Mark
Mark Prince
Just in it for the espresso and coffee, Vancouver BC
Mark Prince
 
Posts: 1064
Joined: Mon Jun 27, 2005 10:13 am
Location: vancouver bc
full name: Mark Prince
company: CoffeeGeek.com
: www.CoffeeGeek.com

Small Correction

Postby Michael_Teahan on Thu Nov 30, 2006 10:13 am

I do not want to imply that jetting doesn't affect the quality of the shot or the extraction process; it matters a lot. It just matters more on the ramp up and pre-infusion than on the later end of the shot. If a large just results in uneven distribution (Marshall's machine) it will affect the later end of the shot because the puck started out unevenly saturated.

Keep in mind the the column of water that lands on the coffee in a piston lever machine is a very low pressure initial wetting, building pressure gradually up to the pressure indicated on the boiler pressure guage.

Just like a jet; just different.

A little.

Pistons also have a declining pressure curve from the initial high pressure extraction. That would be interesting to look at, and I think Jim has been working on it.

Piston lever machine work great, though if you believe the temperature Nazi's, these machines can't make espresso.

Probably Italian magic beans.

Probably.
Michael_Teahan
 
Posts: 18
Joined: Fri Oct 14, 2005 3:05 pm
Location: Los Angeles

Re: Small Correction

Postby Marshall on Thu Nov 30, 2006 4:36 pm

Michael_Teahan wrote:I do not want to imply that jetting doesn't affect the quality of the shot or the extraction process; it matters a lot. It just matters more on the ramp up and pre-infusion than on the later end of the shot. If a large just results in uneven distribution (Marshall's machine) it will affect the later end of the shot because the puck started out unevenly saturated.


Michael is referring to an adjustment he did that I reported on alt.coffee: http://tinyurl.com/y7en4n. I'm happy to say 15 months later that the front channeling never reappeared after the gicleur reduction and that there has been no clogging of the gicleur.

[Ignore my shameless sucking up to Michael in the alt.coffee thread. I was extremely grateful.]

Marshall
Marshall
 
Posts: 505
Joined: Thu Oct 06, 2005 10:21 pm
Location: Pasadena, CA
full name: Marshall Fuss
company: Marshall R. Fuss, Attorney at Law
: http://www.linkedin.com/in/fusslaw

Postby Mark Prince on Thu Nov 30, 2006 8:45 pm

I was hesitant to even contribute to this thread because to be honest, I really value everything Michael writes anytime his contributions appear online - I don't necessarily always agree with his theories and ideas - but I do definitely take them with a heavy weight and serious consideration - I think he's brilliant, and definitely knows his stuff. The only reason why a "no tamping" philosophy even enters my thinking spectrum is because of what Michael has written about it over the years. I do take to heart and mind what he thinks about that aspect when thinking about the tamping process, and to be honest, its his viewpoints that actually lead to new avenues of thinking about how passive preinfusion, gentle pressures and lower dispersion screens vis a vis the placement inside a filter basket when a PF is locked in... all of that came from Michael's views on tamping.

What I'm trying to say is, I really appreciate the point / counterpoints in this thread... even if it's taking me some time to wrap my limited brain cells around it all.

Mark
Mark Prince
Just in it for the espresso and coffee, Vancouver BC
Mark Prince
 
Posts: 1064
Joined: Mon Jun 27, 2005 10:13 am
Location: vancouver bc
full name: Mark Prince
company: CoffeeGeek.com
: www.CoffeeGeek.com

Postby tim on Fri Dec 01, 2006 2:11 am

Interesting thread.
But I am still curious about the difference in taste.
A couple of years ago I equipped a 3 group Linea with 3 different flow rates. 1 was 0,5mm, one 0,6 mm. and one free flow (1mm)

Interestingly enough I cannot recall exactly the outcome, but I did actually decide that the free flow (1mm) was the best for my espresso as I like it bright and fruity. It seemed that the restrictors created a more ashy and almost burned flavour in the cup.

There is another factor here that would affect the result (one I did not know of then) the temperature variations between the 3 groups.

Anyone else did this ?
Tim Wendelboe
tim
 
Posts: 93
Joined: Sat Jun 10, 2006 1:59 am
Location: Norway

Postby Jim Schulman on Fri Dec 01, 2006 12:54 pm

tim wrote:Anyone else did this ?


I think there's something to this; perhaps it has to do with dwell time. My informal take is that machines with short dwell times have brighter, less muddy shots. On the other hand, a really short dwell, like on the 1mm Lineas, tends to be a severe challenge for us mere barista mortals.
Jim Schulman
coffeecuppers.com
Jim Schulman
 
Posts: 265
Joined: Tue Apr 11, 2006 10:56 am
Location: Chicago

Postby gscace on Fri Dec 01, 2006 7:34 pm

Hi:

Proprietary research has been conducted on this effect. Dunno what is in the public domain.

For simplicity, think of the effects of flow restrictors in terms of a garden hose. If you put your finger over the end and successfully block off all flow, the pressure will be constant all the way down the hose. As soon as your grip loosens and a little water leaks out, then there is incremental pressure drop per unit length down the pipe. In order for the water to move, the pressure upstream has to be greater than the pressure downstream. If only a little water is leaking out, like with a very small gicleur or with a LM Linea specially neglected by Nick, there is little pressure drop between the pump and the gicleur. However, on the downstream side of the restriction, the pressure will be much less. In the hose analogy the pressure on the opposite side of your finger is zero (relative to atmospheric pressure). If you could somehow get an auxiliary device to block the water flow from the hose that was downstream of your finger, then your finger wouldn't have to work so damn hard to squeeze off the flow. The pressure drop across your finger would be zero if the downstream device successfully blocked all flow, and the drop would increase as the device began to leak.

Think of the coffee cake as a big flow restrictor. Once brewing commences, pressure on the upstream side of the cake is somewhere between atmospheric pressure and the pressure of the water exiting the pump. There's a big honkin' pressure drop across the cake. Pressure drop across the gicleur is zero immediately before brewing, and jumps up to almost the entire difference between the pump's pressure setting and atmospheric pressure (the whole 9 bar nut) in the instant after the pump is activated, when air is being squeezed out of the volume between the group solenoid and the coffee cake. Pressure drop across the gicleur reduces as water is absorbed by the cake, and as any pre-infusion chambers(such as spring-loaded chambers) become filled. Eventually the pressure drop becomes more or less constant as flow through the cake becomes more or less constant.

While The previous paragraph describes the effect of a gicleur on water pressure experienced by the coffee immediately after brewing commences, it doesn't discuss the effect of different gicleur diameters on pressure profile. Obviously a bigger hole flows a whole lot more liquid through it for a given pressure drop. Think big diameter garden hose vs small diameter one. The effect of a big, or non-existent gicleur is that the period of pressure rise at the cake, and subsequent reduction of pressure drop across the gicleur is greatly reduced. The coffee cake experiences high pressure almost immediately after brewing commences. Smaller gicleur diameters produce slower pressure rampup at the cake.

Small gicleurs mask inconsistencies in grind fineness, dosing volume, and distribution. For a given pressure drop, the big hose flows way more water than the tiny one. Consider that varying the parameters mentioned above will all have an effect on the ability of the coffee cake to resist pressure. For example, an underfilled basket will provide less resistance to pressure than a properly filled basket. Given this condition, the pressure drop across the gicleur will be greater than desired and more water will flow through the system. The effect is greater with large diameter gicleurs than with small diameter ones. This is why an automatic Linea without gicleurs seems relatively intolerant of inconsistent technique, whereas the visually apparent effect of sloppy technique is less obvious once 0.6 mm gicleurs are installed. It don't mean that things are totally hunky dory though. In actual fact, the "steady state" water pressure immediately upstream of the cake will be lower than you would expect if a basket is underdosed and small gicleurs are used. I think Andy's number of up to half a bar is in the ballpark. What's the effect on coffee taste? I'm presuming it's tastable but I haven't done any tests correlating variation in pressure above the cake with taste. Perhaps Michael Teahan can weigh in on that.


-Greg
gscace
 
Posts: 102
Joined: Fri Aug 12, 2005 7:15 am
Location: Laytonsville MD
full name: Gregory Scace
company: Espresso Research

Postby tim on Sat Dec 02, 2006 3:42 am

Interesting !
I don't remember he exact results of the difference in taste, but I do remember that it was a big difference ! It was also necessary to adjust the grinder for suiting the various flow restrictors.

Again, temperature might have played me a trick too..
Tim Wendelboe
tim
 
Posts: 93
Joined: Sat Jun 10, 2006 1:59 am
Location: Norway

Postby Klaus on Sun Dec 03, 2006 4:08 am

We installed 0.6 mm gicleurs on our LM GB5/EE and had quite a possitive change in the taste profile. The biggest difference was a significant increase in body and mouthfeel. Shots seem heavier and with a better texture. Also somewhat increase in sweetness. We have the flow restrictors in our machine at the shop and are quite happy with it. We were actually just yesterday talking about putting them in the machine at the office too, because we feel it's so much better.
I know that Jens-Henrik (former LM importer here in DK) had one of the WBC Trieste machines here for a while and it had the 0.6 mm installed, but after a while he said he had problems with it. Shots ran on the sour side or something - but there might have been other issues at hand too (like temps consistently running too low at beginning og shot).

Klaus Thomsen
The Coffee Collective
Blog here
Klaus
 
Posts: 197
Joined: Fri Sep 30, 2005 5:13 am
Location: Copenhagen, Denmark
full name: Klaus Thomsen
company: The Coffee Collective
: http://www.coffeecollective.dk
: http://coffeecollective.blogspot.com

Postby erics on Wed Dec 06, 2006 3:43 pm

Orifice calcs tend to be a little fuzzy unless you're the orifice builder and tweak your correction factors BUT . . .

Based on a flow of 75 ml in 25 sec (60 in the cup & 15 in the puck), 9.0 bar pump discharge pressure, 200 F water, and this site's (http://www.okcc.com/) info on page 24 of their catalog, a 0.6 mm gicleur gives 8.1 bar to the puck, a 0.8 mm gives 8.7 bar to the puck, and a 1.0 mm gicleur does 8.9 bar.

A 0.6 mm orifice should give about a 95 ml water debit.

Now where is that baby accumulator I was looking for? :D

Eric S.
erics
 
Posts: 1
Joined: Sun Mar 05, 2006 7:40 am
Location: Silver Spring, MD

Next

Who is online

Users browsing this forum: No registered users and 1 guest

cron