Ben,

This is a good question, and one that has sort of been tackled in other threads but I've never really seen a clear and concise "answer". In my ramble thread, there are no fewer than 30 individual posts hashing this out but no one-liner that ties it all together. That said, I do have a bit of input for this conversation.

First, let's go over the "why water debit?" issue.

Wattbe wrote:When I say 'water debit,' I mean the volume of water that flows from the grouphead with no portafilter engaged over a period of time

pizzaman383 wrote: First, let me say that I hate the term water debit because I just can't get my head around why that term is used instead of flow rate.

Putting these two statements together you get:
"Water debit is the flow rate out of the group head with no portafilter in it."

I use this term because it's two words instead of twelve. Flow rate on its own is insufficient to express what's going on because as soon as you put a puck in the way the flow rate changes.
Caveat: The above is true unless you have a Decent Espresso Machine or some other flow profiling machine that will ramp the pressure up in an attempt to follow a specified flow rate. Such machines don't really have a water debit because the debit is whatever the commanded flow rate happens to be at that time...
So, to Ben's point the water debit really is all about the first 10 seconds. Once the puck starts absorbing water and generating back pressure, it takes over as the primary flow restrictor and the water debit is no longer in play.

The question becomes "how does water debit affect how the puck absorbs water or generates back pressure?". Cliff notes: lower water debit saturates the puck more thoroughly and builds pressure more slowly than higher water debit, with respect to total shot time . From there, you have to look at puck dynamics and understand that the coffee grounds in the basket are doing work when water is flowing through them. It's the work the puck does that generates the backpressure that we all identify as "brew pressure".

The top layer does the least work, because nothing but clean water passes through it. It merely gives off its soluble compounds and some fines and hangs out, being pushed against the layer of coffee underneath it. The next layer has to hold up the top layer and pass not just clean water through it but also any dissolved soluble compounds and fines from the top layer; this layer has more work to do than the top layer. All of this continues on until it snowballs at the bottom of the puck. This bottom layer has the full brunt of the water pressure (over 530 lbs of force) acting on it (by way of supporting the layers above it) along with passing every last bit of solids that goes into your cup through it. The water pressure on the last layer is very nearly zero, but it has to physically hold up the rest of the puck. More specifically, it has to support all the work that the layers above are doing. They guy with the hardest job is the basket. He has to support all the work of entire puck.

Why did I tell you all of that?

Well, imagine holding up all 530 lbs of water pressure that is pressing down on the top of the puck when you are the bottom half of a dry puck of coffee. All the voids in the puck and in the individual grains of coffee are filled with squishy air. When the top half of the puck -which is saturated with water- pushes down, you get a super-tamp that drastically reduces flow.

However, if the bottom half of the puck is filled with water before full pressure is applied, then two major things change. First, the puck is no longer filled with squishy air that lets it collapse, but with incompressible water, which helps support the weight of the puck layers above. Second, and more importantly, by evenly dividing the water pressure from the top of the puck to the bottom of the puck, less of the puck (thickness) is subjected to the full force of the water acting down on the top of the puck and there is less chance of channeling because the physical structure of the puck is under less compressive stress when it is fully saturated than when dry.

Slayer takes this to extreme levels and gets you a very slow ramp up to full pressure during pre-brew. Pre-brew ends organically with slayer when the puck is saturated and pressure rises to the full available brew pressure. This happens even with the needle valve engaged and this point represents the lowest flow of the puck and the point in time when it is the most restrictive to the system. From this point, (when drops coalesce on the bottom of the basket and the cone starts to form) the flow rate begins a steady increase when constant pressure is applied. The Slayer technique is to switch from pre-brew to brew at this point because this is the moment in the process that the needle valve is insignificant in terms of flow. It makes for a seamless transition and makes the tail end of the shot behave more traditionally than it would if we left the needle valve in play.

I bring this up because all machines have this inflection point, when the puck is saturated and the flow into the puck reaches a local minimum. It is this reason that water debit is not called flow rate. You could call it "max flow rate". But then folks might think you have some sort of fancy variable flow rate machine. (Oh wait, Ben does ) The important characteristic of all except a select few machines is that flow rate is dependent on the puck. Water debit is not. As such, flow rate cannot be strictly defined for any machine. Water debit can be easily measured and shared with others.

Cheers!

- Jake

vit wrote:what about the name "Idle flow rate" ?

That's not a bad idea. But there is a breed of machines out there now that make this term problematic. More on this later, but I think "maximum allowable flow rate" may be a winner.

Wattbe wrote: Is it safe to say then, that water debit in machines with some form of Pre-infusion, either manual or automatic is irrelevant?
If the puck reaches saturation point during the pre-infusion stage which, by its very nature will occur at a lower flow and pressure, am I right in saying that water debit plays no role whatsoever?
If the puck reaches saturation point during the pre-infusion stage which, by its very nature will occur at a lower flow and pressure, am I right in saying that water debit plays no role whatsoever?

I don't think so. I hear what you're saying, but it gets complicated because most preinfusion mechanisms are dependent on water debit. That is, increasing or decreasing the water debit changes the dynamics of preinfusion, regardless of the approach taken to preinfusion.

Wattbe wrote: One interesting thing to note is that in the 'low pressure rehash,' it's stated that pre-infusion is set to zero - straight to 6 bar at a 500ml/min flow rate. Surely if the goal is to fully saturate the puck prior to full brew pressure being applied, then a much lower 'pre-infusion' flow rate is the way to go?

I totally agree. I've read everything Strive for Tone has put out and most of it is good reading and seems solidly grounded, but I can't figure out why preinfusion would be somehow incompatible with a 6 bar extraction and there doesn't seem to be any evidence-based reasoning presented that I've seen.

Wattbe wrote: In machines with absolutely zero preinfusion, I can see how water debit could play a role, but most machines nowadays have some form of pre-infusion, either naturally as in the E61 design, electronically as in the LMLM or BDB or manually adjustable on the fly.

I don't think a machine can have "zero preinfusion". Look at the Decent machine. Even with a flat 9 bar profile, it takes over 10 seconds before the machine builds pressure. This is because the max flow rate is (software) limited to 6ml/sec, which is kind of like saying it has a water debit of 180ml/30sec, on the low end of the spectrum. In order for a machine to have no preinfusion, there would need to be zero headspace between the brew valve and the puck. I've yet to see such a machine. Because headspace has to be filled prior to developing any brew pressure, I still maintain that water debit "shapes" the preinfusion of any machine whether or not that machine sports a separate preinfusion method.

pizzaman383 wrote:...with a vibe pump the flow after the preinfusion and pressure ramp also effects the forgiveness factor of machine and the taste of the shot.

I think this is due to the lower water debit of a vibe pump coupled with the more linear pump curve than a rotary pump. Vibe pumps naturally flow less as the pressure increases because their stroke gets shorter when they have back pressure. Rotary pumps are constant volume devices, with any reduction in flow being due to internal slippage. The same can be said of the gear pumps in Slayers and the like. Remember that once once the OPV is open, vibe pumps and rotary pumps behave exactly the same as far as the puck is concerned.

pizzaman383 wrote:So, I think the steady state flow during the shot is also important and this is where different pump pressure/flow rates and gicleurs come into play.

I think "steady state flow during the shot" is a misnomer. The flow starts off at the water debit of the machine, slows down to a minimum at the point of saturation, and then progressively increases as the puck erodes. There's a rather lengthy discussion of this in My long and rambling path to preinfusion/pressure profiling. The post in that link is my bench approximation of the puck pressure, flow rate into and out of the puck over a typical shot of mine, preinfused for ~15s at line pressure before bringing the pump online. Assaf then loosely corroborated my approximation with some telemetry data from Chimera. My point to this is that once the puck is saturated, the type of pump is largely irrelevant as long we're talking about a standard, constant pressure, pump-driven machine.

pizzaman383 wrote: I always bring temperature into the discussion of flow rate because the differential solubility of the various chemicals that impact the taste of espresso. We may control temperature and flow rate but taste is ultimately determined by what chemicals come out of the puck and this is both important and for the most part invisible because it is hard to measure.

I totally agree with you, but again, I don't think that "steady state" is meaningful within the context of a shot. It's a roller coaster. What's important is finding something that works and makes more good stuff come out of the puck and into your cup while leaving the bad stuff in the puck. Personally. I think the HX hump is beneficial in this regard, but I'm open to challenge...

I mentioned at the top of this post about the "new breed" of machines. If you read my ramble thread, you'll see that I'm a bit nuts with respect to all this stuff. After reading pretty much everything I could find on the subject, I decided to try something a bit silly on my machine. What I came up with is effectively a manual implementation of the Dalla Corte Mina. It is exactly the implementation of the Lelit Bianca. The idea is simple. Replace the gicleur with a needle valve similar to the one Slayer uses. Instead of switching a solenoid valve around the needle valve like Slayer, just open it when the puck becomes saturated. This method is dead simple to implement on my machine and, as Lelit demonstrated, is doable on an E61, as well.

What's new breed about these? They are a class of machine that I have dubbed "Variable Water Debit" machines. The DC Mina is the classiest of the bunch, with a tablet interface an a stepper controlled diaphragm valve that precisely hits a handful of predetermined water debit setpoints. Lelit and my Rancilio S20 MIDI CD just use a manual needle valve to haphazardly raise or lower the water debit, on the fly. The result is pressure profiling achieved by dynamically adjusting the water debit. Set the water debit to be less than the flow rate of the puck at 9 bar, and the back pressure generated by the puck falls. Set it above the flow rate of the puck at 9 bar, and you get interrupted flow. The puck sets the upper flow limit, but setting the water debit lower can give you any brew pressure you want. Decent sort of falls into this category. But both the Mina and Bianca are passive; the pump does it's thing and the valves do their thing and you get what you get from a pressure and flow standpoint. Decent is active, so it will bump the brew pressure up to 13 bar to maintain target flow rate if that's what it needs to do. I don't think that's a bad thing, but it's different.

Ok, enough for now.

Cheers!

- Jake