Optimum flow rate w/o portafilter (water debit)

(Alexander Buzath) #1

It seems 200-280 ml/30 s is seen as the ideal flow rate w/o portafilter (water debit) - at least these values are recommended by @MattPerger and @Michael_Cameron

I’m wondering if there’s a potential downside of going even lower (maybe 80-100 ml/30 s) by using a very narrow restrictor (but keeping the pump pressure in the 6-9 bar range)? Wouldn’t the result be quite similar to what a Slayer does?

Thanks, Alex

(Matthew Perger) #2

If it’s too low, you’ll have a problem with low pressure in the puck. You’ll have to grind so coarse that pressure cannot build up.

@Michael_Cameron is probably the best man for advice on this right now. He’s in the thick of experiments.

(Michael Cameron) #3

I think the best way of thinking of this is, if you keep dose, grind setting, grind temperature, brew water temperature, volume of water (volumetrics), pump pressure, and tamp pressure constant (and that’s a big “IF” …) then a lower water debit will result in a longer shot time. And you’d probably grind coarser to compensate. Depending on what other variables you kept constant if you wanted shot time to be static, there’d be a point where water debit would be too low to get favorable results in the cup.

I’m all about bell curves at the moment, so I’d say that 200/280ml / 30 sec is somewhere at the top of an optimum water debit curve. 80/100ml I think would be on the wrong side of that peak.

(Imile de Villiers) #4

What happens when you are above the 200ml/30sec rate? I have a client whose refurbished GB5 is pushing about 3 times that amount of water while the pump pressure gauge remains around 9bar.

(Alexander Buzath) #5

Thanks a lot, Matt and Michael, for your explanations.

I think I can put it all together now:

With a high water debit there’s a strong force hitting the coffee in the portafilter and a high risk of channeling and uneven extraction. By reducing water debit the wetting of the puck takes place slower and extraction becomes more even and forgiving. But eventually you reach a point where you’d have to grind too coarse to maintain flow.

(Chris Moore) #6

The giggler (sometimes spelled gicleur or gigler) may be missing or otherwise damaged on the refurbed GB5. Check page 22 in the manual linked here. That piece is responsible for regulating flow rate to the grouphead, and not having it could produce the experienced condition.

From what I can tell from the diagram, it is possible to install the grouphead without the giggler.

(Maximiliano Chang) #7

Hello everyone!

I´ve got a Sofia Bianchi and have different flow rates in each group. group1 220ml/20s and group2 241/20s.

Any ideas on how to fix it?

(Maximiliano Chang) #8

another question, in order to get 200-280ml/30sec i need to go below 8 or 7bar pressure. is that okey?

(nicolas) #9


My store runs a La Marzocco GB5, the first group had a water debit of 310g/30s, the second is 262g/30s and the third is 257g/30s.

From what I’ve read on here this means that the first handle is more susceptible to channeling because of the amount of water hitting the puck.

Can someone further elaborate on this for me as I really have no clue, but I feel like it’s something that has to be addressed in store


(Michael Cameron) #10

Hey Nico, sounds like you’ve got an “overuse of one group head” problem, allowing scale to build up in the group with the largest water debit. It could be solved with a service of the flow meters - you’d need to have your coffee tech come in and have a look. But it’s a common enough problem, you usually find the group closest to the grinder is the one used most often, and it doesn’t take long for a water debit difference to start showing up with each group. Ideally they should be a few grams apart at most - so yeah, get a tech in to have a look.

(David) #11

On the GB5 the group jet or gicleur can be in one of two places depending on the serial number break. Refer to this TSB:

The serial break is at 5679 for AV and 5680 for EE models. The gicleur was installed under the cap necessitating removal of the entire cap to clear or change the restrictor. With the new design the gicleur ruby is incorporated into a very small grub screw that can be removed simply by unbolting the 3-way solenoid body from the cap. The procedure is simple in either case but there are some safety considerations that must be followed to prevent serious injury so hiring a technician is best. Good luck
Ed: This is the Piero cap starting with s/n 2075, before that it was the Linea-style external plumbing and the gicleur was located in the tube leading from the flowmeter to the left fitting on the group neck.

(Troy) #12

Perhaps I’m misinterpreting the meaning the term “water debit” (have to love coffee jargon :neutral_face:). Is it not just the flowrate through a group with no portafilter in place?

If so, I would expect a more heavily scaled group (and thus greater resistance to flow) to present with a lower water debit (i.e. smaller flowrate), not a larger one…

(Scott Burlington) #13

Is 200-280ml/30sec at 6 bars? Or does it not make a difference?

(Alexander Buzath) #14

That´s an important question! According to the Hagen–Poiseuille equation the flow rate decreases proportional to a decrease in pressure. So it DOES matter if you set the water debit to 200-280 ml/30 s before or after regulating down to 6 bars.

(David) #15

Here’s the guide for gicleur orifice sizes as equipped on La Marzocco machines as of Jan. 2014 and their effect on freeflow (water debit).

(Troy) #16

I generally find it much more intuitive to think of pressure as the result of flow, than the other way around. In fact, you’ll notice the Hagen-Poiseuille equation (and most other flow equations) are typically given with pressure drop as the dependent variable (and flowrate or velocity independent).

An analogy is driving a car - would you say that your speed increases proportional to the drag force, or that the drag force is proportional to your speed?

Pressure drop down a tube can be thought of in much the same way - resistance to flow as the result of friction (i.e. drag)*. The greater the resistance, the greater the pressure drop for a given flowrate. In an espresso machine, it is the pressure drop that determines the backpressure at the pump (i.e. the pump discharge pressure).

I personally think the emphasis on pressure in espresso has overshadowed the importance of flowrate/velocity, and wonder if the “flow depends on pressure” perspective is part of the reason why.

*That’s a really simplified model, and it’s always important to remember that all models are wrong, but it’s certainly much more useful than the ubiquitous garden/fire hose analogies (which just lead people down the garden path).

@latte911 - interesting info in that guide from La Marzocco. Rather round numbers for the flowrates with the various different gicular sizes. Comparison with (predicted) pressure drop across just the gicular (i.e. excluding the rest of the system) is interesting:

(Mark Stahlwood) #17

Does anyone have any thoughts on trying to measure water debit on a pressure profiling machine?
My cafe uses a single group modbar, and I’ve been getting great extractions at lower pressure settings, but found it next to impossible to measure water debit. Even when I set my pump pressure down to 3bar or lower, I’m getting about 300ml in 20s, and that doesn’t shift much regardless of pump pressure. Set at 12bar or 3bar and water debit is only different by ~20ml.

Pump pressure reads at 0bar regardless of what it’s set at when there is no portafilter in the group (or even an empty portafilter in the group), and the pump sounds like it spins up to a speed higher than any it brews with. This leads me to believe that on the modbar, the indicated pump pressure is a calculated or measured value of pressure at/near the coffee puck. I’m also assuming that since the modbar doesn’t use flow restrictors and has a gear pump that without something (coffee in a portafilter) to provide resistance to the water, it’s just pushing water through the group at roughly maximum possible output.

So I guess what I’m asking is, does anyone else with a modbar or strada ep or similar machine find it tricky to measure water debit? Am I missing something simple that would allow me to measure water debit?
If it’s not possible, then it’s not possible, but I’m curious if I arrived at a similar water debit to @Michael_Cameron and @MattPerger’s recommended values after dropping my brew pressure/temp/tamp pressure/grind size and extending my shot time (which has vastly improved the consistency and quality of my extractions).

(Alexander Buzath) #18

:+1: I think you´re completely right.

In the end it´s not the pressure but the water moving through the puck that extracts the coffee. What you want is a reasonable flow rate and an adequate grind size.

Changing your water debit (whether with restrictors or by altering pump pressure) is one of many options to reach that goal:

Higher water debit & everything else unchanged -> finer grind for same flow rate
Lower water debit & everything else unchanged -> coarser grind for same flow rate

While there are other ways to change grind size while keeping flow rate steady (dosing, tamping etc.), lowering water debit has the obvious advantage of wetting the puck more slowly and thereby enhancing evenness of extraction.

So my conclusion is that the ideal water debit allows for a slow saturation of the puck while maintaining the grind size fine enough to extract the coffee. In the end that´s probably what pressure and/or flow profiling is about :wink:

(Matthew Perger) #19

On a strada, it’s impossible because of the way the machine is designed. It has no restrictors because for them, pressure was the variable to be played with.

(Matthew Perger) #20

“wetting the puck more slowly and thereby enhancing evenness of extraction.”

Sounds counter-intuitive to me. Slower = one end having more water for longer, no?