It would have to be a preheat-and-cool-down system because the gear pump I'm using isn't designed for hot water. Max temp is 55C at max torque.

I guess we don't know exactly what gas or gasses are coming out of solution in my brew boiler, but it seems to me that the case for it being CO2 resulting from bicarbonate heating is pretty strong. My water typically measures about 200 ppm hardness and 150 ppm alkalinity. My salt-charged softener completely eliminates the hardness by trading Na ions for Ca ions. But the bicarbonate ions (alkalinity) remain at the 150 ppm level, which is pretty high.

I happen to have an Everpure Claris system that I took out of operation a few years ago due to concern that it was producing acidic water. This is a hydrogen ion exchange system that completely removes both the hardness and alkalinity. It has a bypass valve that allows untreated water to mix back into the demineralized water. Typically, I set it to produce about 50-70 ppm of hardness and alkalinity (that was back when the two measured more equal -- nowadays the hardness level is higher.) Anyway, it seems like cutting the bicarbonate level by half should reduce the amount of CO2 produced.

It's not all that hard to change out the softeners. I only hesitate because of the acidic water issue: my Claris output pH measured in the low 6s. The concern was that the low pH can be caused by chlorides in the water. But my water report doesn't show chlorides (don't know if they were measured, but the chlorine level was nil.) It turns out that CO2 dissolved in the water can also lower the pH. It's not a problem for espresso machines because the CO2 comes out when the water is heated. The very fact that I measured acidic water from the Claris suggests that, in fact, I do have CO2 dissolved in the water.

At least, that's the theory. Anybody think I should try the Claris? I can start with very low hardness and alkalinity to test the theory. The espresso might not extract as well, but the test might confirm that the problem is CO2 outgassing.

In the 100-250 psi heating steam boiler world there is a device called a degasifier located in the condensate return to boiler feed water pump suction. It would be a huge headache to mimic this device in a domestic water system. Feeding brew water from a 5 gallon jug after it has been degassed might be the easy solution. Many domestic water systems supply well aerated water which is saturated with dissolved gasses. Heating your water in a pan over the stove will give you an idea of how much gas is in the water and at what temperatures it evolves. I think that is the first thing I would look into. More than likely all the dissolved gas is out of the brew water before you reach 195F @ atmospheric pressure. The initial flow into the group enters at atmospheric pressure and the water temperature is above that which the gas will easily go back into solution.

Not sure the issue is CO2. It is a simple experiment to heat water without the bicarbonate and an equal amount with the bicarbonate and note the volume and temperature of the gas bubbles formed. That will tell you what is going on. CO2 gas is not easily dissolved into liquid water and it does not stay in solution very long when the water is at atmospheric pressure.

The solution to the low Ph of H cation resin processed water is a mixed bed polishing element.

Peppersass wrote: Anyway, it seems like cutting the bicarbonate level by half should reduce the amount of CO2 produced.

The Claris will not drop your dissolved CO2. That's because it replaces your Ca++ and Mg++ cations with H+ ions. Your bicarbonate alkalinity drops, and the carbonic acid goes up (pH goes down), and some CO2 is generated in the process. It looks like this:

H⁺ + HCO₃^- --> H₂CO₃ --> H₂O + CO_2(aq) --> H₂O + CO_2(g)



It is true that at a given pH, a lower alkalinity would give you a little less dissolved CO2, and at a given alkalinity a higher pH would give you less CO2.
You can get an idea about this effect by using this online calculator from Pentair:
http://pentairaes.com/co2-calculator

Using the calculator it appears that if the incoming water is 77F (25C), pH 7.7, and alkalinity of 150 mg/L CaCO3 equiv, you should expect about 12 mg/L [CO2_(aq)], which is about 0.3 mmol/L.

A mmol of gas at STP has a volume of 22.4 ml, so now if you assume that all of this dissolved CO2 is expelled as gas upon heating to brew boiler temp, then your 1.5 liter boiler could give you up to 10 ml of CO2 gas at STP, (which might be closer to 7 ml at brew boiler temp and pressure.)

P.S.
FWIW, One example of a pH adjusting cartridge you can buy that scrubs your carbonic acid (and therefore lowers the CO2) from the water is the Everpure 7NE as discussed in this post:
High sodium, low pH water treatment solutions?
If your pH is already high it would not do much.

Well, the cation-softened water has a pH of about 8. But since the bicarbonate alkalinity is still high, does that mean significant CO2 will be released by heating?

If so, then the Claris, with a pH in the low 6s, plus a 7NE might reduce the amount of CO2 released.

You calculation of the CO2 released assumes a static volume of water. But after the CO2 comes out, the depleted water goes out the group and fresh CO2-laden water enters the boiler, where it's heated and releases more CO2. While very small amounts of CO2 would be released at any given time, wouldn't it tend to accumulate in the boiler? It's certainly the case that the lag seems to be more and more persistent the longer it's been since a drain/refill.

It could be the case that some of the CO2 is going out when the group is run, so the accumulation might be quite slow. Running at a higher pump speed (higher pressure and flow rate) often forces the jump, which might also force some CO2 out of the group, so when I return to nominal pump speed (9 BAR pressure), the lag is temporarily gone. It comes back a minute or two later, after the boiler has reheated.

Peppersass wrote:If so, then the Claris, with a pH in the low 6s, plus a 7NE might reduce the amount of CO2 released.

In theory, yes, but I wouldnt advocate doing that, because I don't think your dissolved CO2 is that high and because I really doubt that it is causing your mysterious pressure jump.

If your pH is actually around 8, then your water utility is likely doing some sort of treatment that raises the pH and lowers the CO2/carbonic acid, probably for anti-corrosion reasons. Using that online Pentair calculator I referenced earlier, at pH=8, temp=25C, alkalinity=150 mg/L, you get a dissolved CO2 of about 3 mg/L (0.07 mmol/L). That's not at all high, nowhere near saturation which would be in the neighborhood of 0.6 mmol/L at room temp and zero gauge pressure.*

Also, in your machine I'm not sure that much, if any, of that dissolved CO2 is gassing off. It is true that solubility goes down with temperature, but it also goes up with pressure. Your incoming water is at room temp and maybe 1.5 bar (gauge) of water line pressure. Your boiler idles at maybe 100C (around 3C above the PID setting) and at maybe 4-5 bar of gauge pressure (5-6 bar absolute). At that temp and pressure the solubility of CO2 is actually a bit higher than it was at room temp and 1.5 bar of gauge pressure. Because of that, and because your water seems to be well below saturation, I think your CO2 is mostly staying dissolved in there.

* Here's where I found numbers for CO2 solubility vs temp and pressure: https://www.nist.gov/sites/default/file ... crd427.pdf

Is there a method to figure out the amount of dissolved gasses?

I can think of two ways that are easy to at home that I'd try:

1. Put an inverted beaker full to the top in a vacuum chamber and lower the pressure gradually (I'd use my vacuum bag chamber for this).

2. Heat water in an inverted beaker. I'd probably put a 3-10w resistor in the base of the beaker and pump a current to heat it up.

Maybe there is an easier way. Maybe some sort of aquarium kit?

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.

How much time is "after a while"? "sooner or later"? Are you talking hours/days?

Completely refilling with fresh water means refilling with cold water with max amount of gasses dissolved. Heating up would then release most of these gasses. So it seems illogical that your "problem" would go away after a complete refill if related to dissolved gasses.

I would rather think it is mechanical related. Also you say it doesn't occur when pulling a shot; maybe the question should be why not? If dissolved gasses are the cause then why not when pulling a shot? On the other hand, if mechanical, I can eg imagine that the "problem" occurs also(!) when pulling a shot but because of the resistance of the puck it disappears much quicker so that you don't even notice it.

Hello there!!

There are a number of different factors involved like the calcification of the boiler and because you're using a sweet plumed in GS/3!

To check if it really is the issue as hypothesized (instead of assuming that it is confirmed as such) I would recommend using bottled water from a reputable source. If the problem persists... (which you can plume in using a pump to generate pressure). It might be mineral buildup on a component that water passes through too.

I read your post... to quote

"Another thing I should mention is that I never see the jump in pressure when pulling a shot or backflushing. For that reason, I'm not sure this is really a problem, but it does concern me that I can't explain it and the initial pressure seems to be getting lower over time"

Eureka! Its very possible that you have overlooked some factor between the boiler and the sensor! Try running Pre-infusion (around 4 bars) then jumping it to 9 bars. Please do report your results. Also, do ask a GS/3 Tech if its actually wear & tear on the sensor itself. Good Luck!