Trapped Air & Spongy Pulls (Again)

A haven dedicated to manual espresso machine aficionados.
John Michael Hauck

#1: Post by John Michael Hauck »

Very much has been written here on the subject of trapped air and spongy pulls. I am indebted to the many posts here by the late Robert R. Pavlis rpavlis and Pat Moore homeburrero where they discuss flash steam. I am looking to reignite that conversation with my current understanding here.

This is about manual lever espresso machines with boilers. Many of these have a cylinder in the group head to hold water for pulling a shot. A piston in the cylinder can be moved up and down by a lever. When the piston is in the down position, it prevents water from exiting the group head (left image). To brew espresso, the portafilter is removed, filled with finely ground roasted coffee beans and re-attached to the group head.

The operator begins the brewing process by raising the piston (right image). This movement has multiple effects. One, is that the upward movement forces some water above the piston back into the hotter boiler. Another effect is that rising piston creates a vacuum underneath it which is readily filled with air drawn through the dry porous coffee grounds. Finally, when the piston reaches the top, a hole is exposed on the side of the cylinder, allowing water to enter the area below the piston.​

​Ideally what would happen next is that all the air under the piston is driven back through the dry coffee grounds, and the area under the piston is filled with just enough water at the perfect temperature for brewing espresso. Then, the forceful lowering of the piston drives that water through the ground coffee to extract a perfect espresso.



However, many users of direct lever machines will occasionally experience what is called a "spongy pull". That is, as they force the piston down with a lever, they notice the lever springs back up if they let go. This is due to some (compressible) air trapped in the area below the piston. This is not an ideal situation because not enough water will be available for pulling the shot. Also, when the portafilter is removed, any remaining compressed air may make a mess of things.

It is important to understand that it is steam, not liquid water that effectively drives out the air underneath the piston. When the water enters this area, it undergoes a rapid depressurization. That is, the pressure of the water in the boiler is higher than the pressure of the air under the piston. The rapid depressurization results in a small percentage of that water turning to steam, known as "flash steam". While the percentage of water that turns to steam is small, the volume that steam occupies is huge. To better understand this, we can calculate the volume occupied by water that turns to steam, and the volume occupied by the water that remains liquid. We can then compare these two volumes as a ratio.

By examining this chart, we can see that at 105C, the water entering the area under the group head occupies about 15 parts of steam to 1 part liquid water. In other words. at 105C, enough steam is generated to fill the area under the group head 15 times over. At 100C or below, none of the water entering the area under the piston turns to steam. This then is the reason why introducing cooler water into the area under the piston results in spongy pulls. The remedy of course is to increase the water temperature by, for example, maintaining a higher group head temperature.



Here is my Google Spreadsheet showing the calculations.

However, brewing espresso benefits from water at a temperature in the range of 91-96C to keep it from over-extracting the coffee, which yields a bitter taste. This intricate knife-edged trade-off of these types of lever machines makes them endearing to many.

User avatar
TigerStripes

#2: Post by TigerStripes »

Thanks for the interesting post John - let's see if I understand correctly with some theoretical numbers.

1. A 1 bar boiler with 60c group head results in 80c water hitting the puck. (hot boiler / cool group head)

2. A 0.5 bar boiler with 90c group head results in 80c water hitting the puck. (cool boiler / hot group head)

Are you saying I would expect to get less spongy pulls from scenario 2, where I'm using a lower boiler temperature and a hotter group head to achieve the same final temperature hitting the coffee puck?
LMWDP #715

John Michael Hauck (original poster)

#3: Post by John Michael Hauck (original poster) replying to TigerStripes »

Here are some assumptions that help explain my thinking:

A1) The flash steam generated by the water as it undergoes a rapid depressurization when it enters the area under the piston is what drives off the air and that is what reduces the spongy pull. The higher the temperature of that water, the more flash steam is generated. I'll call this location of rapid depressurization the "pin hole".

A2) The flash steam generated by the first drops of water passing through the pin hole is more effective at driving off air than the last drops. One reason is that the porosity of the ground coffee decreases as it is wetted, preventing the escape of air. Another reason is that the larger quantity of water in the area under the group head also acts as a barrier to the escape of air.

A3) The temperature of the water going through the pin hole increases over the time it takes to fill the area under the piston. This is because the initial drops of water come from the group head and the last drops of water come from the boiler. (http://www.youtube.com/watch?v=eQuEZDY6Awo).

A4) Both your scenarios indicate "80°C water hitting the puck". I'll take this to mean the average temperature of the water under the piston immediately after it is filled is the same in both scenarios.

A5) Your scenario 1 indicates a "1 bar boiler" and your scenario 2 indicates a "0.5 bar boiler". From this we I will infer the boiler water temperature in scenario 1 is hotter.

Given the assumptions, what I need to answer is which of your two scenarios would see a higher temperature of the first few drops of water passing through the pin hole.

Because the last few drops of water passing through the pin hole in scenario 1 are hotter and the average temperature of the drops passing through the pin hole in both scenarios is the same then the first few drops passing through the pin hole in scenario 1 are cooler than scenario 2.

Therefore, I would conclude that scenario 1 would yield a spongier pull.

However, I am not very confident in the truth of A2. It may be that the scenario with the hottest drop of water at any time during the filling process is the one with the least spongy pull.

jpender

#4: Post by jpender »

John Michael Hauck wrote:This intricate knife-edged trade-off of these types of lever machines makes them endearing to many.
While I would never wish to take away the joys of such complications I can't help but wonder if an exit port just opposite the water inlet that would give air an escape route other than through the puck would solve this dilemma. It would be wet and would require a drain though. Maybe it wouldn't work. And of course it isn't there so it is just a thought.

John Michael Hauck (original poster)

#5: Post by John Michael Hauck (original poster) replying to jpender »

I've always wondered why La Pavoni added this little void area in the generation 1 group heads. Maybe they were thinking the same thing as you?


mikel

#6: Post by mikel »

John, any thoughts about the effect of locking in the portafilter after raising the lever part way?
After seeing videos of the puck getting lifted during the lever raise, I've been doing it to minimize the chance of disturbing the puck, but could it help with spongy pulls too?

John Michael Hauck (original poster)

#7: Post by John Michael Hauck (original poster) replying to mikel »

Mike,

Let's say you raise the piston to just before the "pin hole" is revealed and then attach the portafilter. At this point, the area under the piston and above the coffee grounds is still full of air. So, no, I do not believe your practice will help with displacing that air when the water is introduced.

Now you have me wondering about how it is that air passing through dry the grounds can disturb the puck (while raising the piston). How can this happen? Is it because the grounds are wetted by residual water in the portafilter? Are they wetted by a leaking group head that drips on them once the portafilter is attached? Or can truly dry grounds be so fine and so tightly packed they prevent air from passing though? Regardless, if the puck is relatively air tight, then it seems spongy pulls will be inevitable.

Your thoughts?

baldheadracing
Team HB

#8: Post by baldheadracing »

I'm struggling to see how water, steam, and air could interact any differently in a manual lever vs. a spring lever - assuming both are dipper-fed.

In a spring lever, boiler pressure and group temperature directly determine the characteristics of the air trapped between the bottom of the piston and the top of the water column that immediately forms over the puck once pre-infusion starts. The mass of this trapped air is fixed by the piston/cylinder/basket/puck depth combined bore and stroke, as air is lighter than steam or water.

Regardless, my Europiccola just arrived at Customs an hour ago, so in a few weeks I will know where I'm wrong :D

P.S. Thank-you for the awesome models of La Pavoni's. The group comparison video in particular convinced me that I wanted a Gen 1 - the other generations' groups just didn't feel as "right" in their design to me.
-"Good quality brings happiness as you use it" - Nobuho Miya, Kamasada

John Michael Hauck (original poster)

#9: Post by John Michael Hauck (original poster) »

baldheadracing wrote:I'm struggling to see how water, steam, and air could interact any differently in a manual lever vs. a spring lever - assuming both are dipper-fed.
Craig, I think we agree on this. That is, the basic interaction is the same regardless as to whether the extraction is powered by muscle or spring. Where there is a difference is in the operator's perception of trapped air; that is, a spongy pull. It seems to me that in a direct lever machine, the operator feels this, and in a spring lever machine the operator may not be fully aware of the situation. Do you now think we are in agreement?

baldheadracing
Team HB

#10: Post by baldheadracing »

I think that we are in agreement.

The one "if" I am wondering about - and will check once I get my machine - is if the three holes that are in the v1 group cylinder are at the same height. If one of the holes is a little higher, then the air and steam that would otherwise be trapped between the bottom of the piston and top of the water/puck during pre-infusion could also expand and rise to occupy the space above the piston, leading to less air being trapped once the piston starts its downwards travel and those three holes are blocked off.

In true gen 1's this effect - if it exists - could be hidden by the dual-switch operation. I am assuming that a dual-switch machine would be less spongy just due to the lower boiler pressure at the low power position. (Mine has a pressurestat so I won't have that advantage.)

If my hypothesis is correct, then that might also explain the mysterious well in the top of the gen 1 group cylinder. Or not :wink:.
-"Good quality brings happiness as you use it" - Nobuho Miya, Kamasada