Using water debit to adjust espresso brewing parameters

Beginner and pro baristas share tips and tricks.
User avatar
Jake_G
Team HB

Postby Jake_G » Feb 08, 2019, 2:01 pm

This is not a conversation about preinfusion. But only because preinfusion is a misnomer. This topic isn't remotely new.

Anyone who has followed the LMLM threads knows that different sized gicleurs provide different water debit flow rates and that folks who prefer classic comfort blends likely love the shots produced by the higher water debit of the stock gicleur, while those who prefer lighter roasts will prefer the characteristics of the smaller 0.6mm gicleur.

Water debit is a common measurement to describe the flow from the shower screen of an espresso machine when there is no back pressure on it. In general, rotary pumps have higher water debit measurements than vibe pumps. Equipment manufacturers generally equip machines with restrictors often known as gicleurs to adjust the water debit down from what it would be without one. Smaller gicleurs result in lower water debit measurements. The large majority of Home Baristas accept their machines for what they are and live within the confines of the manufacturer's gicleur selection. Small changes in water debit are inconsequential to the results of the espresso produced from a given machine. However, there are now more and folks using machines that can alter their water debit in the fly.

So what. Who cares?

By now, most of us should be familiar with the charts below from Espresso 101: How to Adjust Dose and Grind Setting by Taste:
Image
And
Image

These are both wonderful, but they are 2-dimensional.
Dimension 1: Dose.
Dimension 2: Grind

Enter dimension 3: Water Debit.

3 dimensional charts are tricky. The charts above are a "slice" of the 3d chart with water debit held constant. Since most machines feature a fixed water debit, there has been no driving force to explore what happens when you change it. Certainly little mention of how to use it to control the flavor of your shot. So options for a quick and easy tool are to hold grind constant and put together a "Dose vs Water Debit" chart, hold dose constant and put together a "Grind vs Water Debit" chart or to build a 3d "Rubik's Cube Chart" showing all 3. I like Rubik's Cubes, but I think they make most folks' brains hurt. These charts would be no different, but the beauty of a 3d chart is that it puts it all together in such a way that you can experiment with dose, grind and water debit in a fashion that yields predictable results with minimal waste.

I dont have the charts ready yet but I can begin the discussion with results that have been observed from a handful of folks with machines capable of tweaking water debit at will.

Neutral Water Debit
This is somewhere between 4 and 8 ml/s. With a "normal" dose that yields roughly 2mm of headspace between the tamped puck and shower screen, and a "normal" grind this will act like a standard E61 with respect to pressure ramp at the puck, dwell time to first drips, etc.

Decrease Water Debit
As you decrease the water debit below 4 ml/s, you start developing puck saturation before pressure builds and the flow rate of the shot increases. Decreasing water debit takes longer to get first drips, but it does the same thing to the overall timing of the shot as lowering the dose or using a coarser grind. Moreover, it allows one to use a finer grind with a given dose, or use a higher dose with a given grind than a higher water debit would allow. It's important to note that with the current rising tide of flow profiling machines, water debit can be adjusted along a continuum, just like grind and dose.

Increase Water Debit
As you increase water debit above 8ml/s, you begin to compact the puck faster than water permeates through the puck. Pressure builds at the puck very quickly and the shot time will [b]increase[b]. Increasing water debit can increase or decrease time to first drips, depending on the grind and dose, but the thought is that more fines migrate to the bottom of the puck and the resulting pour will generally be slower, much like increasing the dose or tightening the grind would. Likewise, increasing the water debit allows one to use a coarser grind with a given dose or a lower dose with a given grind than a lower water debit would allow.

In and of itself, I'm not sure that water debit does much to impact the flavor of the coffee. What it does is allow you to access combinations of grind and dose that would otherwise be outside of your normal brewing parameters. It also gives you an added level of control in case your dose is a bit high or low or if you accidentally forgot to adjust the grind tighter after your last gusher, you can just bump your water debit up or down accordingly and enjoy a great shot.

This is already too long for most reasonable folks to read.

Cheers!

- Jake

RyanJE

Postby RyanJE » Feb 08, 2019, 2:33 pm

This is a very interesting read Jake, thank you for the information! In the end you state that you are not sure water debit affects the flavor, but that conflicts with the title of the thread.

Aside from being able to potentially correct a shot that might be headed off course, what do you think adjusting debit does then? When I read posts from those with ability to flow profile (eg. Jim S and the Bianca) it appears it opens a "new dimension" of extraction that cannot at all be obtained with Pressure, Grind or P.I.
I drink two shots before I drink two shots, then I drink two more....

User avatar
Jake_G
Team HB

Postby Jake_G » Feb 08, 2019, 2:38 pm

Good call.

I adjusted the title accordingly.

I didn't give this statement nearly enough emphasis:
Jake_G wrote:What it does is allow you to access combinations of grind and dose that would otherwise be outside of your normal brewing parameters.

I think this is what myself, Jim and others have discovered.

Cheers!

- Jake

nuketopia

Postby nuketopia » Feb 08, 2019, 4:13 pm

I take issue with the comment about the LMLM and coffee preferences and orifice selection. I own an LMLM and have used with both the 0.8mm and 0.6mm orifice. My net takeaway from that experiment is that the orifice is largely immaterial to the performance of the machine.

I left the 0.6mm in the machine because it is a royal PITA to change and opening the brew boiler to change it risks damaging a special O-ring that costs about $50 to replace.

My conversation with LM about it is they chose the 0.8mm is to reduce the chance of machine failure due to scaling in homes where the water quality is not well controlled.

The LMLM is it's own animal really, in terms of pressure ramping and water flow, and very much different than the e61.

The larger orifice in the LMLM ramps pressure faster, which I could observe by video recording the SCACE-II pressure with an iPhone. But not by very much. It ramps very quickly in both cases.

User avatar
Jake_G
Team HB

Postby Jake_G » replying to nuketopia » Feb 08, 2019, 5:20 pm

Thanks Larry,

I will clean up the first post be more clear and rely less on urban myth that doesn't appear to be accurate.

Cheers!

- Jake

User avatar
lancealot

Postby lancealot » Feb 09, 2019, 1:27 pm

Bear with me, I am trying to get my mind around this.

Could water debit be considered a way to set or adjust maximum flow, in the same way that setting your OPV would set or adjust maximum pressure?

User avatar
Jake_G
Team HB

Postby Jake_G » replying to lancealot » Feb 09, 2019, 1:41 pm

Yes, but...

It's a complicated relationship between water debit and the way the shot pours. This morning I tried the following with a fixed fine grind in a Rancilio precision 18g basket:

Shot 1- 19.5g in, ~2ml/s water debit, 30g out in around 35s after first drops.

Shot 2- 16g in. ~8ml/s water debit, 30g out in around 40s after first drops

If I had been thinking clearly, I would have done a better job stopping these shots at the same 1:1.5 brew ratio. Instead, the second shot was a tad bit less strong.

I'll do more testing and grab a couple shot videos to get the overall timing more precisely recorded. Point being, faster water debit allows for a slower shot with less dose and slower water debit allows for a faster shot with a higher dose. They all work together...

Cheers!

- Jake

User avatar
lancealot

Postby lancealot » Feb 09, 2019, 5:17 pm

Thanks Jake,
I think I understand the mechanics of it. I am just trying to understand how it effects the taste of the beverage.

Mesmer

Postby Mesmer » Feb 09, 2019, 5:18 pm

Trying to get my head around measuring water debit. Say you have a tank fed espresso machine and you can measure water debit pulled from the tank and also water debit returned to the tank (from OPV for example). Would the difference between the two equal the actual debit of coffee in the cup? Thinking there's nowhere else for water to go to and it shouldn't get compressed.

nuketopia

Postby nuketopia » Feb 09, 2019, 5:31 pm

Jake_G wrote:Thanks Larry,

I will clean up the first post be more clear and rely less on urban myth that doesn't appear to be accurate.

Cheers!

- Jake


No worries, all in good humor.

The 'water debit' topic has certainly come up many times over the last 10 years. "Water debit" is just an odd term to express volume flow rate with no basket present. Ie, zero-pressure maximum flow.

You measure by simply letting the machine run without the portafilter for a set period of time and measure the total volume at the end of the period. Some of the older threads suggest 30 seconds.

Much of the numbers running around are based on particular types of machines, most generally, e61 style machines. It's been oft-discussed and an "optimal" number sort of developed on the 'net.

The LMLM appears to be its own animal here. I measured it and it produces a prodigious amount of water if left to run. Much more than an e61 with a rotary pump and crazy more than vibe-pump e61. But - that is the way it is designed. It is capable of moving a lot of water - which isn't a bad thing if you wander through the math and physics of it all.

I experimented with it by changing the orifice to 0.6, since it is used on LM pro machines (though I didn't opt for the expensive synthetic ruby orifice). The "water debit" was reduced. However, it seems to be immaterial. The observed results show a slight reduction in pressure ramp at flow start. I can't measure it, but a smaller orifice would further reduce the meager amount of pre-infusion water, as the pre-infusion is the result of the release of brew boiler pressure (and line pressure if the machine is plumbed) through the orifice integrated over 1-second that it is open.

It all gets into the dynamics of fluid flow and pressure through an annulus and a particle bed, which is a bit hard to sort out.

The zero pressure flow rate (aka debit) can probably be compared between machines of similar designs. But of very different hydraulic design, not so much.

The relationship between volume and pressure and flow is a bit complex. They're all inter-related and depends a lot on the availability of volume from the pump, the regulation of pump discharge pressure, the size of the plumbing, presence of flow restricting devices and so on.

The LMLM for instance, has a parallel water path between the pump and brew boiler, so the effective pipe is twice as wide. Then there's the brew boiler chamber, the orifice, a solenoid valve and a long, convoluted path through the bottom of the brew head to the shower screen. So, very, very different water circuit than say, an e61. The same differences would likely hold for other machines with unique hydraulics.

Very hard to compare the two. I doubt one "water debit" value is applicable across different machine designs.