For several years I've been trying to determine the cause of a mysterious pressure lag/jump in my GS/3. Now I'm exploring a theory that it could be caused by CO2 coming out of solution in the boiler water, as described below. I would much appreciate comments from anyone who has expertise in water chemistry and/or fluid dynamics.

I posted about the lag/jump here. The post contains a video of the phenomenon (sorry for the poor quality.) First it shows running at a pump speed that should produce 9 BAR at free flow, but the pressure lags by 1 BAR. Then it shows how an increase in speed to produce higher pressure suddenly causes the pressure to jump. This is accompanied a by a noticeable change in the sound of the water flow. Sometimes just running the pump longer at the lower speed causes the jump. As the video shows, after the jump the pressure at normal free flow is 9 BAR, like it's supposed to be, and the flow has the same sound as it did after the jump. Not shown in the video is that after a few minutes of idle time the lag/jump returns.

An interesting aspect of the lag/jump is that it usually goes away for a while if I completely drain and refill the boiler. But sooner or later it comes back. And the longer the time period since the last drain/refill, the longer the lag.

Our member AssafL has tried to help me think this through. We've explored many possibilities, including expansion/contraction of the pump input and output hoses, bad check valves, bad 3-way valve, etc., but none of them has panned out. Several times Assaf has suggested the possibility of an air bubble. Since bleeding the group has no effect on the lag/jump, the bubble would have to be trapped somewhere in the input water path. I can't imagine where the air would be trapped and where it would be hiding. More important, if this is the case I would expect other GS/3 owners to have posted about the phenomenon.

While thinking about this, I remembered something our member shadowfax told me about water chemistry, namely that carbonic acid in the water can hold CO2 that's released when the water is heated. I have well water and our area contains a lot of limestone and gets a lot of acid rain, the combination of which produces carbonic acid. In further reading, I believe it's actually the case that the carbonic acid quickly becomes calcium bicarbonate, which is responsible for the 150-200ppm hardness in my water. I use a cation softener, which I believe turns this calcium bicarbonate into sodium bicarbonate. But the sources I read indicate that with either bicarbonate, heating above 50C (another source said 200F) will release CO2.

Another data point is that I often hear a bubbling/hissing sound when the heaters come on. I also see a lot of tiny bubble in the water when I bleed the group, but I've thought those might be caused by the hot water hitting cold air.

Could it be that my brew boiler water is releasing CO2 when heated, and that the time it takes to compress the gas is causing the pressure lag/jump? I'm imagining a lot of small bubbles, perhaps adhered to the heating element or trapped in the nooks and crannies of the input path, that get compressed and suddenly hit the wall. One thing I like about this theory is that it seems like it would account for the lag/jump going away when I drain/refill the boiler, but returning after the boiler has built up some CO2.

Here are some questions for the experts:

1. Could gas in the boiler and/or input path cause the pressure lag/jump I'm seeing?
2. Does carbonic acid persist in water or does it turn into calcium bicarbonate?
3. Could carbonic acid or calcium bicarbonate or sodium bicarbonate release enough CO2 to cause the lag/jump?
4. If the answer to #3 is yes, is there a way to purge the CO2 from the water?
5. Has anyone heard of a lag/jump like this?

I am not an expert on the chemistry but I remember that once in a blue moon the water from the faucet in also Gatos (near San Jose,CA) looked like tiny bubbles soda.

But I would guess that any water would have dissolved gasses. In theory, increasing the temp should increase the solubility of the gasses in the water (and increase the pressure in the vessel). But if the machine cycles hot-cold-hot (as in the daily power on-off cycle), over time some of hear gasses would be released into the vessel.

To me it seems that it would make sense to bleed the group every so often to release gasses (and also the random bubble that find their way into the water supply - like when a neighbor redoes his water mains).

What you most likely see when bleeding the group, at least when the water is not cold, is flash boil. Water under pressure boils much later than under normal atmosferic conditions, when I refill my Feamina boiler after the water has cooled a bit I can see vivid boiling through the sight glass.

What I also can see with the lid off is that the heating element causes small gas bubbles, which I explain as mainly being microboiling. The heat of the element is locally causing the water to start to evaporate. Likely because the heat of the element is on/off. All 800 watt is transforming the metal rod into a blistering hot item that is cooled by the surrounding water but it is much hotter than the 100'C or so required. As the water heats there is no uniformity in heat transfer as water further away will heat later, convection will start to increase when the temperature difference gets larger.

I'll leave the pump mechanics/dynamics to the engineers, all I can offer is that a pump may cause cavitation at some speeds but that will not be overcome by draining and refilling the boiler.

The chemistry is in much better hands with rpavlis, our resident chemistry professor, I will say that any chemical reaction is a balance so sodiumbicarb will exist next to the components in dissolved state and temperature and pressure do affect that balance to shift left or right of the equation. CO2 will be formed by heating bicarb in solution but what exactly heat and pressure do with that reaction I can't really say.

Water chemistry is not your issue, the non condensible gasses dissolved in the water are. They are evolved as the liquid water heats up. The gases begin to evolve from the water at 160F or so and continue to 200F or so. It works best to vent the boiler as soon as you see .5 bar or so on the gauge, again at .7 or so and finally when operating pressure is reached and the heating element switches off. The venting process is critical as just winging open the steam valve is not going to do the trick. Just crack open the valve and let the boiler vent into a tall small diameter container. The initial vent continues until the pressure either drops or the rate of change decreases. The next vent continues until you see the condensation and fogging inside of your container. The final vent is the tricky one as you have to be able to actually hear the sound of the venting gases. This requires a low velocity flow or all you hear is the screech of that flow. The non condensible gasses have a solid hiss sound and steam will have a softer hollow sound in that tall small diameter container. A tall form 250ml beaker works very well for this. Each of thee vents is maybe a 10-20 second process.

If the water that enters the puck is over ~212F this is another issue that will drive volatiles out of the coffee as a gas.

Not sure I understand your instructions or how they might apply to a GS/3. It's a double-boiler machine, so the steam valve has no part in the process. The machine is plumbed in, so minimum pressure is 3 BAR. The only way to vent the brew boiler is via a bleed screw in the top of the group.

I've certainly thought that the bubbles I see when venting the hot boiler are due to flash boiling. That said, I've been playing around with tapping the machine and group with a rubber mallet and then opening the bleed screw for a couple of minutes. It's looking like that process temporarily gets rid of the lag/jump, even if it was present before the bleed process. And when it returns the duration seems to be shorter. But I hesitate to call this confirmation that gas bubbles are causing the problem. Need to run more experiments.

I can understand why gases might form from heating, but why doesn't this problem plauge other espresso machines? While my gear pump might take longer to overcome the lag, I don't think that's it. Assaf has a gear pump in his GS/3, which is the same vintage as mine, and he doesn't see it. Also, I saw it occasionally when my machine had the stock rotary pump.

There has to be something unique about my machine or water that's causing this.

What you are describing is dissolved gas in the brew water coming out of solution when the pressure drop as it enters the group occurs. The trick is to figure out how to get it properly vented off. It is possible to degas water before entering the brew boiler but it is not simple or common for coffee machines. A hot water heater set for 180-190F equipped with a automatic gas vent will do the trick.

AssafL wrote:In theory, increasing the temp should increase the solubility of the gasses in the water (and increase the pressure in the vessel).

Increasing temperature actually decreases the solubility of gasses in water. This is why warm soda goes flat more quickly than cold soda. Pipe water can have a lot of dissolved gasses in it. If you ever place a cup of tap water under vacuum, you'll see lots of bubbles form and escape. Environments at high temperature and low pressure favor the escape of dissolved gasses from liquid into a gas phase.

The problem is the OP has to come up with an simple method of degassing the brew boiler feed water.

Jofari wrote:Increasing temperature actually decreases the solubility of gasses in water. This is why warm soda goes flat more quickly than cold soda. Pipe water can have a lot of dissolved gasses in it. If you ever place a cup of tap water under vacuum, you'll see lots of bubbles form and escape. Environments at high temperature and low pressure favor the escape of dissolved gasses from liquid into a gas phase.

The original thought was that oscillating power (on a daily basis) would extract more of the gas. But I could not find a good scientific justification for it. So instead I expostulated the nonsense above...

AssafL wrote:The original thought was that oscillating power (on a daily basis) would extract more of the gas. But I could not find a good scientific justification for it

If the temperature is oscillating and you're able to exhaust gas somehow during the hot cycles, you're right that some of the gas would be extracted. This is the basis of what OldNuc was suggesting for water pretreatment before it gets to the boiler. I can't comment on how easy this would be to implement, but it should work to degas the water.