How much does flow rate matter, if brew pressure is same?

Need help with equipment usage or want to share your latest discovery?
dsc106

#1: Post by dsc106 »

Been playing with flow control and getting some fantastic results. It also got me wondering though as some shots I've pulled that I really like, I've pulled with flow rate never exceeding 6g/sec even though pressure was around 8 bar in the group.

ECM Synchronika stock flow is around 11g second. Is this an important variable in how a shot pulls, and if so, what impact? What does it mean if I am grinding pretty fine - I'm at 11 on my niche zero - and don't turn the flow knob past around 6g/sec but still get 8 bar?

I don't think I could go much finer without channeling. On these shots, I'm doing 15 second low flow preinfusion until first drop, then slowly turning to about half way (6g sec flow), then dropping flow slowly for last few seconds and shut off.

Anyway, just got me wondering - what does flow rate matter? Why is pressure the same even at lower, and if pressure doesn't increase, why might I go higher in flow?

Asking because experimenting too much is hard as the only coffee drinker in house and I get too buzzed. Plus coffee is expensive. So I learn slowly over time, hoping to get insights so I'm more efficient as I experiment...

User avatar
Peppersass
Supporter ❤

#2: Post by Peppersass »

Flow rate primarily affects two things:

1. During pre-infusion of very finely ground coffee, a low flow rate for a long enough length of time will fully saturate the puck and open up the spaces between particles enough to permit the shot to flow without choking the machine. You also get extra contact time during the long preinfusion, allowing the water to extract more. This technique, however, is primarily applicable to very light, light and perhaps medium-light roasts, which are hard to extract using "standard" espresso extraction techniques. The low flow rate during preinfusion allows two things that can extract more from the coffee: finer grind and longer contact time.

2. Low flow rate after preinfusion increases contact time, allowing the water to extract more. By reducing the flow rate gradually after maximum pressure has been reached, you can prevent the flow from accelerating as the puck becomes more and more permeable. Some of us use this technique to maintain a constant flow rate after max pressure has been reached. Again, the benefit is longer contact time, which in turn provides greater extraction for lighter roasts that need it.

As for pressure, bear in mind that when pulling a shot the pressure you may be seeing on the boiler gauge isn't necessarily the same as the pressure at the puck. If there's a flow restrictor or needle valve between the boiler and the puck, the boiler gauge shows the pressure before the restriction. As long as the puck is permeable (i.e., you don't have a blind basket in place), there will be a pressure drop after the restriction. This is similar to what happens with electric voltage (pressure), current (flow rate) and resistance (gicleur or needle valve.) On the upstream side of the resistor, you might have 12 VDC. on the downstream side of, say, a 1000 ohm resistor, where it's connected to the negative side of the power source, the voltage drops to zero (equivalent to the zero pressure you measure at the flow coming out of the puck.) By Ohm's law, the current in all parts of the circuit is 83mA. Remember that the water path during a pull is open at one end, albeit with several sources of resistance along the way, including the puck. Like voltage and current, pressure isn't constant throughout the circuit, but flow rate is.

Also note that pressure plays another role after the basket fills. It compresses the puck, which will impede the flow. The longer it takes to preinfuse the puck, the more time water has to completely saturate the puck before pressure reaches maximum and the puck is fully compressed. If the flow rate is too fast, pressure will build too quickly and the water may not have a chance to fully saturate the puck. That can result in channeling, uneven extraction, etc.

What flow rate are you using for preinfusion? Slayer recommends a flow rate of 1.6 - 2 ml/sec. The Decent folks believe this is *too* slow, causing the puck to open up way too much and too soon, resulting in uneven saturation and flow. They recommend a flow rate of 3 - 4 per second. I switched from the Slayer timing to the Decent timing and feel that it does make a difference in terms of even extraction.

The stock flow rate of 11g/sec you cite must be at free flow (no puck in place.) That would be roughly 660ml/min, which is about what an LMLM does at free flow, I think. My GS/3 does around 450ml/min because it has a quite small flow restrictor (gicleur) of 0.6mm.

What's important is the flow rate with a puck in place. 6ml/sec is roughly the flow rate my machine has with a puck ground for normal espresso technique (no flow control) is in place. This produces 9 BAR at max pressure. For light roasts, I use a needle valve to reduce the flow rate to about 3 ml/sec with the gear pump running at the speed that will produce 9 BAR when max pressure is reached. When the first drops fall into the cup, I bypass the needle valve, which raises the flow rate to about 6 ml/sec. There's no change in pressure because it's already at the max (this is the way Slayer machines work, the theory being that running the pump at 9 BAR throughout the shot ensures there's no shock from a sudden pressure change when the needle valve is bypassed.) After max pressure is reached, I reduce the gear pump speed as needed to keep the flow rate between 20-30 ml/min. I could do that by reducing the flow rate with the needle valve, but adjusting via the gear pump is more precise and reproducible.
★ Helpful

User avatar
Jeff
Team HB

#3: Post by Jeff »

With extraction profiling, skipping over the intricacies of basket-fill and soak, the two "bulk" numbers that I track are peak pressure and flow rate are the end of the shot. My profiles decrease in pressure with time to keep the flow rate during high-pressure extraction roughly constant or modestly increasing. I don't know of any proof that is "optimal". It's how I cut infinite possibilities down to something I can wrap my head around.

I think there may be some relationship between peak pressure in the 4-8 bar (in the basket) range and flavor profile, but not enough to say anything definitive. I've definitely have found that the light roasts I pull can go flat or even cardboard if the pressure is too high.

For a given roast level and my selection of profiles, I find flow-rate to be a better metric for me than time. When the fill/hold phases can vary by several seconds and I might be changing ratio as well, "2.5 g/s" measured by a scale is more meaningful to me than "37 seconds". Unfortunately this requires a scale with built-in flow estimation or a connected scale with an app that can do the estimation. (Or a slick setup with a metering pump)

This measured flow rate is very different than the position of an E61-style flow-modulation valve. As Dick pointed out above, the flow in a system like that is a function of the pressure at the pump, the position of the valve, and the condition of the puck. They aren't flow-metering valves, but an adjustable resistance to flow. If, say, 4 o' clock is 2 mL/s with no basket or puck in place, it will be less when there is pressure built up by the puck. In some cases, like a ristretto-style shot, the puck resistance is so high that it dominates and you can open the flow valve a lot and have very little change in flow.
★ Helpful

PIXIllate
Supporter ♡

#4: Post by PIXIllate »

Wow. Dick and Jeff just unloaded a lifetime's worth of great advice, knowledge and experience on you. This should be pinned to the list of must read posts. I wish that I'd have read this when I started. Then again I probably wouldn't have understood or appreciated it back then.

All I can add is how my own use of the e61 flow control device has evolved along with my understanding of extraction theory and best practice. I plan to do a video at some point but below is my current "profile". I only use this for medium light to light roasts. Anything medium or darker I just run at a 7.5ml/sec position and use the in built e61 preinfusion as I have the stock spring installed.

- Start with the flow control in the position that gives me 7.5ml/sec water debit
- Let that run for 3 seconds to fill the headspace and begin to saturate the puck
- Move the control to the 3ml/sec position as the pressure begins to rise
- Wait until the pressure rises just above 2 bar and then close off the control so there is no water flow and pressure holds at ~2bar
- Wait until I get 1-2ml of drips in the cup
- Slowly open the flow control back to the 3ml/sec position
- When the pressure reaches 4 bar open the flow control up until I hit my peek pressure of ~8bar
- Taper off the flow (optional)
★ Helpful

dsc106 (original poster)

#5: Post by dsc106 (original poster) »

Thanks so much for these awesome replies! I have learned so much that I didn't even intend to. Much appreciated.
Peppersass wrote:The stock flow rate of 11g/sec you cite must be at free flow (no puck in place.) That would be roughly 660ml/min, which is about what an LMLM does at free flow, I think. My GS/3 does around 450ml/min because it has a quite small flow restrictor (gicleur) of 0.6mm.
Yes, sorry for the confusion, this is actually what I originally meant in my question. My stock free flow rate is 11g/sec on the ECM Synchronika. Before installing flow control, the machine always produced that flow. Since installing flow control, I have rarely ever spent much time matching my flow control knob to what would be the "stock flow" setting of 11g/sec, instead starting with a low and slow pre-infusion, then turning the knob part way - to what would be the equivalent of a maybe 6g/sec free flow. I did not measure the actual flow of espresso into cup at this setting, I am just referencing the associated free flow rate.

So, my question was - how much does matching my flow control knob (during an extraction) to the machine's stock free flow rate matter? Is it "weird" that I am pulling shots in a way the e61 isn't designed, at only a 6g/second free flow rate? What is the advantage to cranking the flow back up to stock settings? Does it have any impact, really? What I mean is, if the pressure in the basket is holding at 8-bar at a lower setting (yes, I have a pressure gauge added to grouphead), I am essentially just pulling shots slower (via a lower flow) than a stock e61... would this make me prone to over extraction, or creating totally different profiles? Just trying to understand the impact this variable has - or the impact of not ever turning the machine up to a higher flow rate - is likely having on my shots.

I've had some really good results with this, but I am still learning A LOT about tasting straight espresso and comparing, and I only drink a little a day. I've also had some over extracted results recently with beans that were frozen and I am wondering if after freezing, I am extracting them too hard with a low/slow pre-infusion and a longer shot time.

dsc106 (original poster)

#6: Post by dsc106 (original poster) »

Jeff wrote:For a given roast level and my selection of profiles, I find flow-rate to be a better metric for me than time. When the fill/hold phases can vary by several seconds and I might be changing ratio as well, "2.5 g/s" measured by a scale is more meaningful to me than "37 seconds". Unfortunately this requires a scale with built-in flow estimation or a connected scale with an app that can do the estimation. (Or a slick setup with a metering pump)
This is really interesting to me, I do have an Acaia lunar, I will have to pair it with my phone and measure out/track the flow rate as a new metric of tracking my shots (instead of just time/ratio). Thanks so much for all the insights!
PIXIllate wrote:Wow. Dick and Jeff just unloaded a lifetime's worth of great advice, knowledge and experience on you. This should be pinned to the list of must read posts. I wish that I'd have read this when I started. Then again I probably wouldn't have understood or appreciated it back then.
I agree! Such informative and well written posts, it should be included in a flow control primer. Thank you!
PIXIllate wrote:All I can add is how my own use of the e61 flow control device has evolved along with my understanding of extraction theory and best practice. I plan to do a video at some point but below is my current "profile". I only use this for medium light to light roasts. Anything medium or darker I just run at a 7.5ml/sec position and use the in built e61 preinfusion as I have the stock spring installed.

- Start with the flow control in the position that gives me 7.5ml/sec water debit
- Let that run for 3 seconds to fill the headspace and begin to saturate the puck
- Move the control to the 3ml/sec position as the pressure begins to rise
- Wait until the pressure rises just above 2 bar and then close off the control so there is no water flow and pressure holds at ~2bar
- Wait until I get 1-2ml of drips in the cup
- Slowly open the flow control back to the 3ml/sec position
- When the pressure reaches 4 bar open the flow control up until I hit my peek pressure of ~8bar
- Taper off the flow (optional)
Nice! I will try this. It sounds like a riff on the classic lever profile - which as I understand is a high pressure early hit, pause and soak, back to full pressure. Instead, you are doing a SHORT high pressure hit, taper to low pressure, close, open slowly to low pressure, build to full, taper off.
Peppersass wrote: What flow rate are you using for preinfusion? Slayer recommends a flow rate of 1.6 - 2 ml/sec. The Decent folks believe this is *too* slow, causing the puck to open up way too much and too soon, resulting in uneven saturation and flow. They recommend a flow rate of 3 - 4 per second. I switched from the Slayer timing to the Decent timing and feel that it does make a difference in terms of even extraction.
Thanks for this tidbit. I was using a low & slow pre-infusion at around 3g/sec. I didn't realize there was debate over it. I'll have to experiment!

User avatar
Jeff
Team HB

#7: Post by Jeff »

PIXillate wrote: - Start with the flow control in the position that gives me 7.5ml/sec water debit
- Let that run for 3 seconds to fill the headspace and begin to saturate the puck
- Move the control to the 3ml/sec position as the pressure begins to rise
- Wait until the pressure rises just above 2 bar and then close off the control so there is no water flow and pressure holds at ~2bar
- Wait until I get 1-2ml of drips in the cup
- Slowly open the flow control ...
No "high pressure early hit" here. Remember that there's very little pressure in the basket until it mainly fills. Those are instructions for:
* Quick fill
* Hold at low pressure for "a while"
then move to a profiled, high-pressure extraction

Pressino

#8: Post by Pressino »

dsc106 wrote:
Anyway, just got me wondering - what does flow rate matter? Why is pressure the same even at lower, and if pressure doesn't increase, why might I go higher in flow?

Asking because experimenting too much is hard as the only coffee drinker in house and I get too buzzed. Plus coffee is expensive. So I learn slowly over time, hoping to get insights so I'm more efficient as I experiment...
I think the answer to your question depends on where in the brew circuit your "flow pressure" gauge is. In your ECM Synch the gauge measures pressure in the brew chamber just above the coffee puck. That's exactly where you want to measure pressure. The other thing you need to understand is that once the infusion spring opens to let water into the brew chamber the coffee puck is the main thing that resists outflow from the chamber and its resistance varies over time during extraction. In thecondition when the puck completely blocks outflow, the brew chamber pressure builds until it reaches pump pressure...whatever that is set at in your machine. The flow control device is just a needle valve that lets water into the mushroom, and it's located where the e61 gicleur would be, so it's basically a "variable gicleur." It can go from totally closed (no flow at all, with water in the upstream brew circuit maintained at max pump pressure while the pump is running. When you measure the water flow at different FCD valve settings, which you do without a portafilter in place, you are just measuring flow rate (gm/sec) through the needle valve and the pressure reading at the FCD gauge will be zero. If you put a blind portafilter in place and run the pump, the FCD valve will read max pump pressure when you open the FC Valve.It will reach max pressure quicker when fully open, but it will also get there but slower when the valve is less open.

Understanding all of the above makes it easy to understand what happens when you place coffee in the basket and start brewing espresso...just think of the puck as a "semi-blind" portafilter with continuosly variable resistance during extraction. :)

PIXIllate
Supporter ♡

#9: Post by PIXIllate »

Jeff wrote:No "high pressure early hit" here. Remember that there's very little pressure in the basket until it mainly fills. Those are instructions for:
* Quick fill
* Hold at low pressure for "a while"
then move to a profiled, high-pressure extraction
Correct. For an example of a "high pressure hit" you could look at something like a Londinium which delivers something in excess of 20ml/sec for the initial inrush.

Jeff can correct me but I believe you need something like 2 to 2 1/2 times the weight of the dry puck in added water before you reach saturation and the machine will begin to build pressure.

In my example I see pressure start to rise above 0 bar around that 3 second mark (water is beginning to exceed the weight of the dry puck) and by 5-7 seconds I've reached the 2bar plateau that I choose to hold.

This is another area of experimentation, what pressure should you hold at?
I'd be interested in answers from Jeff and others who have experimented enough to generalize the flavor differences between a 1,2,3 or 4 bar hold.

User avatar
Jeff
Team HB

#10: Post by Jeff »

I'm far from an expert on this, so many variables that "hold pressure" isn't one that I change a lot. Even with a machine that runs automated, reactive profiles, there are far too many variables to try to explore all of them. I've got a couple "go to" profiles that tend to work well with the coffees I usually pull (medium-light or light), one with a hold, one that senses basket fill (its pressure-driven "decision" generally seems within a second of first drops) then goes into high-pressure extraction. Between those two and exploring ratio, that's about all I can try out with 250 g of coffee and be able to enjoy most of the bag.

If I were to try to replicate roughly what I'm doing for a hold, it would be flow until 4 bar (it overshoots slightly), shut off flow entirely, wait for the pressure to drop to 1.5 bar, then ramp up. I don't know if that's optimal, it's "works for me" with my water, light-roast coffees, Niche Zero, and DE1 with VST baskets (and a BPlus screen, mainly for cleanliness).

For a sense of timing, with the Ethiopian I'm pulling right now, still getting used to a Bentwood flat, at 5 mL/s into the basket (no matter the pressure), this morning I hit 4 bar at about 7 seconds after pump on, with first drops in the cup at fraction of a second later. At about 15 seconds from pump on the pressure had dropped to 1.5 bar, about 3 g in the cup, and moved to increase pressure for extraction. Different coffees (and probably different roast levels) behave differently so a Kenyan, Colombian, ... will have different timings and different amounts in the cup.

There probably isn't much time to think if you're doing this by hand, so do the thinking in advance and be ready to react to the basket-pressure gauge.