misterdoggy wrote:Volume of water used is the MOST important factor at my house.

Makes sense. My input water is about 140ppm, too, but the family was using the pitcher.

Ken Fox wrote:You are asserting that these same people (or other, similarly gifted people) have taken various sorts of hard water and treated these hard waters with all reasonable options available, then used the water to make espresso and have taste tested these in a blind fashion? Is all hard water the same, or does it differ in constituents and hence in the way it might respond to various treatments? They studied this also?

I'm in agreement with Ken on this one. There can be huge variation in water composition from one region to another, and it's not at all clear which of the myriad possible chemicals/concentations have detectable impact on the taste of espresso prepared with said water. It seems to me that the only way to know is to blind test espresso made with treated and untreated water, or water produced by two different forms of treatment, from the same location. Even then, the ability of the tester to detect differences will likely come into play. If, for example, I can't tell the difference between espresso made with water from my location at 0 ppm hardness and 100 ppm hardness, then it makes no sense to expose the machine to scaling.

To be clear, I'm referring to treated water that contains substances other than pure H2O, such as one would get with a cation system, a Claris system, ZeroWater+tap or a Brita filter. I believe there's a fair amount of evidence that pure H2O isn't suitable for making espresso.

Marshall wrote:Great, you're already the GS/3 drainage expert (not being sarcastic).

Nah. Just read the manual. It's always amazing to me that many people don't bother. That used to tick me off decades ago when I wrote technical manuals.

Marshall wrote:But, I wouldn't expect a manufacturer to tell you which descaler to buy. If LM recommends one, they risk ticking off 5 competeing suppliers, and each descaler supplier may have somewhat different directions for use. I would check with your dealer for advice.

It seems to me that LM has a responsibility to recommend and document all owner maintenance procedures that may be required to preserve the integrity of the machine, especially if there's any danger that failure to perform the procedure, or performing it incorrectly, may result in damage. If that requires recommending a particular descaler, then so be it.

I appreciate your point about LM ticking off descaler suppliers, but note that they do supply a particular brand of detergent with the machine for backflushing and they thoroughly document the procedure (though, to my knowledge, the new automatic backflush procedure in the 1.13 firmware has only been documented by users on Gs3cafe.) I see no reason why they can't do the same with regards to descaling. The current recommendation to drain the boilers monthly isn't sufficient.

I did discuss water composition and descaling with Chris Coffee when I bought the GS/3, and the recommendation was to keep the hardness at around 3 grains (about 50 ppm), and the machine won't need descaling for several years. The recommendation was to return the machine to them for treatment at that time. Didn't go into any detail with them on what that would entail.

Dick, I hope when you said "0 pH" you meant "neutral (7) pH"!

My understanding from the Water FAQ is that typical pH readings are not necessarily a good indicator of the chemicals at work in your water, as (e.g.) very soft water will be prone to absorbing CO2 from the air, which will shift a pH reading strongly to the acidic side (I've seen this before when testing carbonated water). However, equilibrium pH, which is calculated using the alkanity of water (carbonate hardness), is a much more stable indication of pH. According to Jim's table, water that has ~30 ppm ALKALINITY will have an equilibrium pH of around 6.7, or very weakly acidic. However, you guys that are blindly shooting for 30 ppm TDS have no idea what is making up that number. Your alkalinity might be MUCH lower than that, which could well mean that your equilibrium pH is lower (maybe 6.0).

I will say--I don't understand how even that could cause significant leaching, but I don't see the point in speculating OR testing it out on your $4000+ machine. Get a carbonate hardness test kit, and formulate your water so that you have ~50 ppm alkalinity, giving you neutral equilibrium pH, and get a general hardness test kit, and make sure that your mineral hardness (at 50 ppm alkalinity) is at or below about 30 ppm. If it's higher, then you'll need to drop your alkalinity a bit more (use the table labeled "Maximum Non-Scaling Hardness by Temperature and Alkalinity" in the Water FAQ as a guide).

Again, a TDS meter is probably not a very useful tool if you want to "surf" the "zone" where you have sufficient minerals, etc. in your water to prevent leaching but don't have so much that you get scale. You need a GH and KH test kit. They're pretty easy to use.

Peppersass wrote:I did discuss water composition and descaling with Chris Coffee when I bought the GS/3, and the recommendation was to keep the hardness at around 3 grains (about 50 ppm), and the machine won't need descaling for several years. The recommendation was to return the machine to them for treatment at that time. Didn't go into any detail with them on what that would entail.

IMO you can file Chris' suggestion as "usually but not always true" and most definitely an oversimplification of the issue. Use Jim's table of actual calculations for something that will be much more accurate and useful when used with your water's measured properties.

GC7 wrote:The idea that a Brita filter treated water at 30-35 ppm tds can cause weld failures makes absolutely no sense to me.

Stainless depends on a passivating layer forming on the surface to protect the metal. If the passivating layer does not form, there are many exotic ion combinations that can attack the metal.

Sometimes the alloy or the weld method is wrong for the application.
http://en.wikipedia.org/wiki/Corrosion

Ian

That's all of the most basic stuff required to get an idea of what is potentially possible in this situation. However, I stand by my idea that a simple batch of 35 tds water should not cause problems. Exotic ion combinations and improper weld alloys are not standard operating procedures I would hope.

Perhaps something exotic comes out of desalinization of sea water in Israel but I am pretty confident in my water piped in from aqueducts in upstate NY. Unfortunately its soon to be well water in northern westchester county.

shadowfax wrote:Dick, I hope when you said "0 pH" you meant "neutral (7) pH"!

Yeah, I forgot my basic chemistry for a moment there...

shadowfax wrote:However, you guys that are blindly shooting for 30 ppm TDS have no idea what is making up that number... Get a carbonate hardness test kit, and formulate your water so that you have ~50 ppm alkalinity, giving you neutral equilibrium pH, and get a general hardness test kit, and make sure that your mineral hardness (at 50 ppm alkalinity) is at or below about 30 ppm.

I hope you're not referring to me As you know from our correspondence, I have hardness and alkalinity test kits from Hach and API. Trouble is, they don't give me the same readings. The API GH test reads 40%-60% higher than the Hach hardness test, and the Hach alkalinity test reads 40%-60% higher than the API KH test. I suspect the problem has to do with the different methods used by the kits. The Hach kits use color changing dyes and a separate titration reagent, whereas the API kits just use a titration reagent. With the API kits, it's not clear to me that the first drop isn't required just to get the base color from which the color change takes place. I need to get some 10 ml test tubes so I can double the resolution of the API tests. Perhaps that will resolve the discrepancy.

According to the Hach kits, the water I'm formulating now, which has a TDS of 30, consists of 34 ppm hardness and 50 ppm alkalinity. Nearly perfect according to your recommendation above. But according to the API kits, it's the other way around: 54 ppm hardness and 36 ppm alkalinity. So, my water is somewhere in the range of neutral to slightly acidic. If the Hach test is right, my water probably isn't etching the boiler welds, but it may be ever so slightly too hard to avoid scaling in the steam boiler, according to Jim's table. But if I dilute it with more ZeroWater, then the water won't scale in the steam boiler but it will drop into the slightly acidic range, possibly threatening the boiler welds. I suppose I could add some baking soda to correct that. If, on the other hand, the API kit is right, my alkalinity and hardness levels should avoid scaling in the steam boiler (though the numbers are right on the borderline.) Unfortunately, in that case the water is slightly acidic, possibly risking damage to the boiler welds.

This is why, for plumbing in a GS/3, a cation system makes more sense to me than the Claris system. A Claris unit would present the same dilemma as my current ZeroWater+tap pourover system. Basically, with the hardness and alkalinity in lockstep, it may not be possible to get the best balance between the two, and I can't add baking soda to correct a deficiency in alkalinity. If I understand correctly, a cation system removes all the hardness, but leaves the alkalinity alone. In my case, with tap alkalinity in the 143 or 190 range (depending on which kit you believe), a cation system would avoid descaling and produce a pH completely safe for the boiler welds. The downside would be having to recharge the system on a regular basis.

One question I have is whether cation systems eliminate 100% of the hardness. I've heard figures ranging from 0-20 ppm. I would think at my high alkalinity levels even a fairly small amount of residual hardness might produce scale in the steam boiler. Unfortunately, Jim's tables don't go below hardness of 25 ppm, so it's hard to tell.

Of course, none of this addresses the issue of whether the hardness minerals affect taste. It seems that if one wants to avoid descaling completely, then the hardness levels are going to be 0-30 ppm, well below what the "experts" recommend. That's OK with me. Right now, I'm more concerned with the long term health of my $4K+ machine than slight differences in the flavor profile.

Peppersass wrote:According to the Hach kits, the water I'm formulating now, which has a TDS of 30, consists of 34 ppm hardness and 50 ppm alkalinity. Nearly perfect according to your recommendation above. But according to the API kits, it's the other way around: 54 ppm hardness and 36 ppm alkalinity. So, my water is somewhere in the range of neutral to slightly acidic. If the Hach test is right, my water probably isn't etching the boiler welds, but it may be ever so slightly too hard to avoid scaling in the steam boiler, according to Jim's table. But if I dilute it with more ZeroWater, then the water won't scale in the steam boiler but it will drop into the slightly acidic range, possibly threatening the boiler welds. I suppose I could add some baking soda to correct that. If, on the other hand, the API kit is right, my alkalinity and hardness levels should avoid scaling in the steam boiler (though the numbers are right on the borderline.) Unfortunately, in that case the water is slightly acidic, possibly risking damage to the boiler welds.

I am not sure how it's possible to have a lower TDS than a mineral hardness (or alkalinity for that matter). I understand that those two numbers aren't additive, but seems like the larger of the two should define a bottom bracket for TDS. As for discrepant tests, that is frustrating. I believe these drop-based tests are only accurate to +/- 1 drop, so for these low values they can be relatively inaccurate. I definitely think now that a 10 mL test is well worth the while for this kind of thing. It certainly ought to help you narrow down the apparent discrepancies. You might check out this article about alkalinity test kits for some more enlightening info on test discrepancy and best test practices. I'm certainly throwing out my test kit (it's very well over a year old!) and getting a new one before doing any more testing.

This is why, for plumbing in a GS/3, a cation system makes more sense to me than the Claris system. A Claris unit would present the same dilemma as my current ZeroWater+tap pourover system. Basically, with the hardness and alkalinity in lockstep, it may not be possible to get the best balance between the two, and I can't add baking soda to correct a deficiency in alkalinity. If I understand correctly, a cation system removes all the hardness, but leaves the alkalinity alone. In my case, with tap alkalinity in the 143 or 190 range (depending on which kit you believe), a cation system would avoid descaling and produce a pH completely safe for the boiler welds. The downside would be having to recharge the system on a regular basis.

I'm really, really not going to believe that a pH of ~6.5 could damage SS welds without an "exotic" mix of dissolved substances. I think attributing the issues that LM mentioned with the Israeli machines exclusively to excessively low TDS is an oversimplification of the issue at best. Maybe they've got an exotic ion combination? Who knows. Without more details, it's hard to take the story as any more than FUD.

One question I have is whether cation systems eliminate 100% of the hardness. I've heard figures ranging from 0-20 ppm. I would think at my high alkalinity levels even a fairly small amount of residual hardness might produce scale in the steam boiler. Unfortunately, Jim's tables don't go below hardness of 25 ppm, so it's hard to tell.

My thought is that you're not going to completely prevent some scale buildup with this approach. I think you're right that some minerals get through a cation system, and with 150-200 ppm of alkalinity, those minerals will certainly precipitate in your boilers.

Anyhow, right above Jim's table is the formula that he used to calculate the maximum non-scaling hardness at a given alkalinity:

"Max Hardness Allowed = Alog{44.01 - 14.58*log(T+273) - 2.739*log(A)}"

Where A is the alkalinity (IIRC in ppm) and T is the temp in Celsius. You can use this for low levels of alkalinity if needed (and also tailor the calculation to the specific steam boiler temperature you want to use).

Of course, none of this addresses the issue of whether the hardness minerals affect taste. It seems that if one wants to avoid descaling completely, then the hardness levels are going to be 0-30 ppm, well below what the "experts" recommend. That's OK with me. Right now, I'm more concerned with the long term health of my $4K+ machine than slight differences in the flavor profile.

I have to say, I am with Marshall on this one... Why waste $4000 of your hard-earned money on such a wonderful machine to handicap it with water that's known to produce inferior taste? Just because you don't feel like flushing the steam boiler twice a year with a mild citric acid solution? The GS3 steam boiler is one of the easiest boilers of any machine out there to drain, which is the the only potentially challenging part about descaling (that is, on machines that don't feature a drain).

another_jim wrote:As Marshall already said; this thread is nuts.

It's certainly got some nuts mixed in...

With all the good considerations and points mentioned above
one thing is also a fact:

a commercial machine set up in a commercial environment
such a drive through, coffee shop etc etc

IS NOT an easy task to descale,

so water treatment as a prevention
( we should say a DELAY rather then prevention)
is the best tool to fight scale build up which eventually will happen