No one using a city water system is going to maintain a constant output, because the input can vary considerably through the seasons and as the utility changes its mix of supply sources. At the time we had our paddle o-rings replaced, our GS/3 was two years old. Our regular alkalinity tests of the Claris showed results from 30 to 50 mg/L, with the average being probably around 40 to 45. At the time of the two-year service, the brewhead was clean as a whistle. YMMV.

another_jim wrote:Is the water allowed to aerate between the Claris and the machine? In some ion systems, this is very advisable. The alkalinity calculations in the water FAQ are based on equilibrium (i.e no more exchanges of O2 and CO2 with the atmosphere). You can get 2 pH swings away from equilibrium if the water is depleted or overly gassy.

I don't believe it's typically allowed to aerate. Do you think this matters much in a boiler that's 198-260°F? I was under the impression that the heat would force most of the dissolved gases out of the water. Of course, that doesn't solve the issue of where they would go. I assume in a steam boiler they would bleed off with the steam. I imagine there's much less of a chance for that in a dedicated brew boiler.

Marshall wrote:No one using a city water system is going to maintain a constant output, because the input can vary considerably through the seasons and as the utility changes its mix of supply sources. At the time we had our paddle o-rings replaced, our GS/3 was two years old. Our regular alkalinity tests of the Claris showed results from 30 to 50 mg/L, with the average being probably around 40 to 45. At the time of the two-year service, the brewhead was clean as a whistle. YMMV.

From what I'm hearing, "clean as a whistle" is not necessarily what you want. The main question on such a machine is, do you see any bright copper tubes on the banjo tube, etc. in the inside of the boiler? They should generally be somewhat oxidized in the presence of 'safe' water. If you're not seeing this, you need to increase the alkalinity a bit. It sounds to me like you're very close to where you should be with Claris, if perhaps a little on the acid side at least some of the time. I would turn your setting up one notch and check again in a few months, myself, but certainly when you're that close to 'ideal' I'll bet any damage will take a very long time to exhibit itself.

As you say, Marshall, YMMV. I can't guarantee that you'll have corrosion if you run the alkalinity too low, but I can definitely say that you could. If you're not keeping an eye on it closely, you're taking a major gamble. The fact that the water supply varies seasonally is all the more reason to monitor this, especially if you're 'cutting it close' on the setting. Again, with corrosion there is no going back-just leaks and parts replacements.

shadowfax wrote:I can now say firsthand that Claris can and will damage your espresso machine if you use it wrong. My advice based on what I see is to forget trying to do anything with Claris other than hitting 50 ppm alkalinity (KH) or higher on your output water.

Great post, Nicholas. I think you may have hit the nail on the head. But it's possible that an even more conservative recommendation is in order.

Your post inspired me to dig up the water measurements I made while using the Claris in 2010 and 2011. My previous post said that I set the Claris to produce about 68 ppm hardness and 70 ppm alkalinity. That's not quite true. The 70 ppm alkalinity figure was from the most recent measurements, taken after I changed to a new filter in December of 2011, shortly before discovering the low pH and discontinuing use of the Claris.

My older measurement records show, however, that the alkalinity ran lower than 70 ppm -- sometimes well below 50 ppm. Here are the measurements I emailed to you when I first installed the Claris:



My recollection is that I ran the Claris at setting 3 for a couple of months, then found that the hardness level had crept up to about 80-90 ppm. So I moved the Claris to setting 2, which brought the hardness down to around 68 ppm. I don't have measurements taken at that time, so I don't know if the alkalinity remained at 55 ppm or moved down to 45 ppm as I had originally measured at setting 2.

I ran at setting 2 for about a year, until I noticed a significant increase in hardness in mid-2011. Then I lowered the setting to 1. I knew that this meant the Medium-size cartridge was nearing the end of it's life, so I replaced it with a Large-size cartidge in September of 2011. Here are measurements taken that year:



The bottom line is that the alkalinity level of the water coming from my smaller Claris cartridge was at or below 50 ppm for over a year. At the end of that period, I discovered that the chrome plating on the underside of the group cap had been eaten away almost entirely. I have photos of the group cap, taken in July of 2010 after nine months of running with a cation system, that show the plating intact. I also have photos taken in July 2011, after one year of running with the Claris system, showing the plating badly corroded and mostly dissolved away. I was able to wipe off the rest, down to the brass, with a paper towel. (I shudder to think how many tiny particles of plating made it into my espresso...)

I ran the Claris for about three more months, with the alkalinity measuring about 70 ppm. That's when I discovered the low pH -- 6.24.

Given the low pH reading at 70 ppm, I think one could legitimately question whether 50 ppm provides enough headroom to avoid corrosive water.

Further, and perhaps of even more concern to those who get water from a well, like I do, my results show variation in alkalinity over time. This is probably caused by variation in the amount and composition of the rain we get and how it percolates through the soil to the water table.

Over the past three years, the hardness readings of my tap water have generally varied between 150 ppm and 170 ppm, mostly hovering around 160 ppm, while the alkalinity readings have generally varied between 120 ppm and 160 ppm, mostly hovering around 150 ppm. Sometimes the variation is more extreme. I recall once getting an alkalinity reading of 100 ppm and redoing the test several times because I couldn't believe it had dropped so low.

I'm sure I initially set the Claris when my alkalinity was at its most frequently-observed level of around 150 ppm. But what if it dropped to 120 ppm or 100 ppm sometime after that? I think this explains why the readings fell to 30 ppm in August of 2011.

I also wonder if the mixing valve in the Claris head can be relied on to mix in tap water uniformly all the time, but that's another question.

All of this argues for a great deal of caution when using a Claris system, and possibly any RO-based mixing system. If you get your water from a well, or the mineral composition of your municipal water varies a lot, I highly recommend measuring the Claris output water frequently and with an accurate measuring system, such as the Hach kits. In other words, ditto what Nicholas said.

Peppersass wrote:All of this argues for a great deal of caution when using a Claris system, and possibly any RO-based mixing system.

In my experience, RO-based mixing systems are the water treatment of choice at high end coffee bars. I do not think your machine's damage is common in the industry. If it were, the systems would have been abandoned long ago.

I agree, I think that the warning is not as applicable to blended RO. With blended RO, you don't have the easy ability to reduce alkalinity far more aggressively than any other property in the water. I use RO water blended with tap water to 80-120 ppm TDS, all run through a carbon block filter before going to my machine. After a year of this, my brew boiler is in perfect shape, with no scale formations and a healthy layer of oxidation on my brass and copper parts.

I think monitoring is important with RO... If your blend setup malfunctions and you get straight RO, that would be bad. But all in all, I'd say much 'safer' than an un-monitored Claris setup.

My point was that if your water tends to vary in alkalinity, you need to take that into account when setting an RO-blending system like the Claris. You want to make sure that the alkalinity doesn't drop below safe levels, which Nicholas says is around 50 ppm. This requires regular monitoring.

Marshall wrote:In my experience, RO-based mixing systems are the water treatment of choice at high end coffee bars. I do not think your machine's damage is common in the industry. If it were, the systems would have been abandoned long ago.

Sure, but most high-end coffee bars are located in cities, and it's possible that, on average, there's less variability in the mineral content of municipal supplies than in well water in locations like mine (very rocky soil, varying concentrations of acid in the rain, etc.)

Unfortunately, I wasn't able to find data on the variability of mineral content in US municipal supplies. Best I could find was this study on the mineral content of municipal supplies versus well water. Interesting, but not quite what's needed to prove or disprove my point.

Peppersass wrote:Sure, but most high-end coffee bars are located in cities, and it's possible that, on average, there's less variability in the mineral content of municipal supplies than in well water in locations like mine (very rocky soil, varying concentrations of acid in the rain, etc.)

Unfortunately, I wasn't able to find data on the variability of mineral content in US municipal supplies.

L.A., like many other cities, mixes its water from a variety of sources, including wells, larger aquifers and rivers. The mix varies with the seasons and rain conditions. As you can see from our most recent table (published 7/31/12), the hardness is quite variable:



As I said, the water treatment of choice for high end coffee bars here is soften>RO>remineralize. There is an environmental movement to ban RO (and salt-based softening), because of the waste effluents it produces and some governments have banned them.

Source: https://www.ladwp.com/ladwp/faces/ladwp ... 1912009000