When coffee or aquarium people say KH, they are generally referring to alkalinity. In fact all those KH drop titration kits on the market are actually measuring alkalinity (i.e., the acid buffering capacity of the water.)

Also, pH can vary a lot depending on dissolved CO2, and keeping a pH meter calibrated can be work. If you know your alkalinity you can skip trying to measure pH, and you can use pHeq if you want to do an LSI calculation to estimate scaling potential.


From another post a while back . . .

Alk □ pH □ pHeq □ mg/L KHCO3
10 □ 7.5 □ 6.0 □ 20
20 □ 7.8 □ 6.4 □ 40
30 □ 8.0 □ 6.7 □ 60
40 □ 8.1 □ 6.9 □ 80
50 □ 8.2 □ 7.0 □ 100
60 □ 8.3 □ 7.1 □ 120
70 □ 8.4 □ 7.2 □ 140
80 □ 8.4 □ 7.3 □ 160
90 □ 8.5 □ 7.4 □ 180

In the above, the pH is the expected pH (from http://www.aqion.onl/show_ph ) at 25℃ in an open container at equilibrium with atmospheric CO2. The pHeq is the Puckorius pHeq, which might be argued as best for steam boiler scale/corrosivity estimation. (In natural waters you might use pHeq as the appropriate pH for LSI calculations)

1st-line wrote:However, last year, it was an increasing issue with these distilled water plus mineral packets issues or mineral-added formulas.

It might be better not to lump all these together. The simple purified + bicarbonate (e.g. R Pavlis) recipe has much more in common with high quality softened tap water than it does with Third Wave Water. TWW and the Barista Hustle recipes are based on Epsom salt as a magnesium source, which includes sulfate, which might be a factor. But then many tap waters contain significant sulfate, so maybe not. The TWW espresso formula is a little unique in that it contains calcium citrate, which you would not find in tap water. In theory that citrate might be a food source for biofilm, especially in a reservoir that's not emptied and cleaned regularly. But that still seems like an unlikely possible cause of these issues.

1st-line wrote:IMHO, RO, distilled, and any purified water is a NO-GO for me.

Are you OK with RO and remineralizing? There are many areas where the tap water is hopelessly loaded with chloride (or more rarely, silica) and the only practical way to avoid ruining a machine is to use RO or purified, then remineralize by recipe, blend-back, or a remin cartridge.

My take away is 1) that a boatload of these machines are sold (not just the Bianca, but E61 double boilers) 2) A lot of people who buy them use bottled warer. 3) Problems are statistically rare (although amplifed on-line), and not seemingly based on one easy to discern cause. Generally this means that the cause is in each case is a unique conjuncture of screwups (look up "Poisson Process" in your stats texts to get more).

I'm not happy with water recipes that use sulfates or citrates, since they are not a normal component of rain to ocean water. But in all likelihood they are not a main cause of this.

1st-line wrote:Personally, I do like the Lelit water softener with the use of mixture of filtered tap water. IMHO, RO, distilled, and any purified water is a NO-GO for me. Lastly, this is my opinion based on experiences with customers. Water experts and metal experts can chime in as I am open ears, but they better know how these machines work. Yes, we can blame Gicar or the PID sensor manufacturers, but I can guarantee there will be some waters that will just not work.

What does this mean for people with whole-house water softeners? Municipal water is basically unavailable at that point and even though people with whole-house almost always have RO, the standard advice forever has been not to use RO in boilers because it leeches minerals (never mind the other problems being discussed here).

Clutch wrote:Thanks for the feedback Jim. Understand the industry is having issues (seems like every industry is in some sort of turmoil).

Also I understand your experience on resolving these issues and why you believe that there is a film causing the temperature to read low. What doesn't make sense to me is that when I watch my PID readout of the steam boiler during an over pressure event, the temperature is not consistently low. The temperature is erratic jumping from 235 and bouncing around that temp +\-5deg at the sensor or gicar's refresh rate to 300 back to 235 in a matter of seconds. This erratic behavior seems to start and worsen as the boiler heats up. To me, this sounds like a sensor signal issue, not a slow reaction or a poor conduction to the boiler water/vapor. Additionally, I've taken the advice given here and used Brita filtered tap that has been tested and falls well within acceptability and I am also using the Lelit in tank softener.

Clive is supposed to be sending me a sensor. When I install, I will inspect for a film and any corrosion in the boiler. I'll look at electrical connections as well. I'm hoping a sensor is the fix, because if TWW is bad, rpavlis is bad and my Brita water is bad, what water would I use?

Thanks for your understanding of our industry.

I have seen this bouncing a several times in the past year. The readout will be normal as the sensor is reading a temperature lower than the true temperature. The quick bounce happens when there is some type of partial build up in the PID sensor in the boiler. The reading is NOT accurate. I had a customer with a different machine call on Tuesday with an over temp of the steam boiler with steam coming out of the safety valve. We asked him to open the machine and wire brush the sensor. He emailed yesterday and said back to normal.

On the flip side, we had a customer yesterday with the BWT water treatment system with a overheat issue on the steam boiler. BWT is very good, and this was rare, but the customer did not evaluator his water properly and he did not change the cartridge every 6 months or both. Again, we are having a run rate of about 4-5 calls per week versus 4-5 or more calls per day.

From my recollection emailing with the Lelit water expert, the problem with these purified waters + minerals is the even suspension. Again, I am not a water expert. But, he alluded to the mixtures being not equally suspended over time. I may have misunderstood or he may be wrong. I am not sure.

He did recommend the following:

pH 6.5-8, alkalinity of 40-70 ppm CaCO3, a total hardness of 60-120 ppm CaCO3, water free of chlorine and, chlorides below 50 ppm.
Please note he stated that a film can still develop on the PID sensor which can result in an overheat situation. However, as of the time of this writing, this is the formula most desired. Again, I am NOT a water expert, so do not kill the messenger.

The dilemma I forecast is that water that is perfect for extraction may not be the best water for machine reliability and vice versa.

If I had an answer on the 'perfect' water, I would open up a water supply company.

As one of 1st-line's customer's, I truly appreciate Jim's feedback on this forum. As a reseller, they are in a unique position to see the same issues crop up time and again across multiple machines and customers (most of whom I'm sure don't post here). With this data, they may be able to draw inferences about root cause that might not be obvious just by reading posts on this forum. This is not just valuable information for us, but also the manufacturer who can correct defects like this, and I believe there is a defect here.

What is missing from this conversation is a real root cause analysis. I am encouraged to hear that Lelit has hired a water expert, and hope they are engaging other experts to find out what is behind these overpressure events as well as an appropriate corrective action. The message I've heard so far is that many (most?) of the customers encountering this issue have been using certain water formulas, and most of the time (but not all?) it can be corrected by cleaning the temperature probe and changing water. This is a strong clue at the root cause, but this cannot be the end of the story and there is still a lot of work to do here.

I can't help but relate this to my own line of work - I have spent the last few decades as an engineer in the aerospace industry - and our own industry's methods for getting at root cause through the Failure Review Board, or FRB. There are many examples familiar to the public - a failed shuttle launch due to a brittle O-ring launched on a cold Florida day, a space telescope with an optical aberration due to an incorrectly manufactured calibration tool, a Mars orbiter crashing into the red planet due to a mix-up in English vs metric units in the ground navigation software. (Aside... one of my favorite examples from TV comes from the recent miniseries Chernobyl, which unfolds like a mystery where the root cause is finally revealed as... ok I won't spoil it but do yourself a favor and watch this show.) All of these high-profile catastrophic failures were investigated with an FRB, but most FRBs are encountered during ground test, designed to find root cause and correct more innocuous-seeming issues (e.g., an electronics board which occasionally fails to reply to a command) before they turn into a billion-dollar problem. In these reviews, hypothetical root causes are laid out, often in a "fishbone" chart, and systematically ruled out until what remains, like a Sherlock Holmes case, turns out to be the true cause. The corrective action is just as important, mitigating the risk of similar problems in the future.

I'm sure Lelit does not have the resources to conduct a full aerospace-style FRB, but it sure seems in their best interest to do what they can. The repercussions of not fixing issues like this include expensive warranty headaches, lost customers, dealers deciding to drop products, damaged reputation, etc. I would propose the following questions to investigate:

• What are the observables? Clutch stated that his steam boiler temperature probe measurement was jumping erratically, perhaps between a false (low) reading and the true temperature. Are other owners seeing the same behavior, or are some different, such as a reading which is consistently low?
• What hardware do all of these cases have common? Is it the same temperature probe? Same sensor technology (e.g., same NTC thermistor and/or digital converter IC)? Same Gicar controller? How about the boiler? Same stainless-steel alloy?
• Are there any manufacturers or machines using the same parts identified above that do not suffer from this issue? If so, what is different about these machines?
• Water is a prime suspect here, but I haven't seen anything in these 9 pages of posts to draw a conclusion. Are there any ingredients in common to most of these failures (e.g., I've heard Epsom salt recipes mentioned)? If so, I also wonder if Lelit is able to reproduce the issue with the same ingredient(s) and then resolve it by removing the "film" from the probe. I also can't ignore Clutch's experience with his brand new machine and Brita filtered tap water, which does not fit the model here. It could be that his case is unique, and he has a loose electrical connection with his temperature probe.

All of this should ultimately result in a corrective action. If it is as simple as avoiding water formulas with a certain list of ingredients, that would be easy (and should get stamped as a warning on every user manual going forward). If it is traced to a style of temperature probe, then Lelit (and other manufacturers using this probe) should at least conduct a recall or extend warranty coverage to replace this specific probe when it goes bad.

I'm ending my long post here. I am quite interested to see how this resolves, as I plan on owning and enjoying my Bianca for many years.

Interesting to note that in every machine that was "cured" of the problem by removing the thermocouple and cleaning, the only consistent thing was the electrical connector was unplugged and replugged as part of removing it it for cleaning.

This thread is far from conclusive. Various water recipes have been involved/implicated, various machine manufacturers, etc. but a clear cause/effect relationship hasn't been established.

My first Elizabeth began to experience steam boiler over pressure issues within the first 48 hours of using rpavlis water. Other waters I have used include Kirkland Purified Water with minerals added for taste (with additional potassium bicarbonate added), Brita filtered tap water, etc., all with the same result on the original machine.

Jim at 1st line blamed the water immediately and asked me to clean the sensor which I did and this fixed the problem for about 72 hours but the problem returned. I took the machine to him for his evaluation over several weeks. I am not sure what waters he used at his shop but he experienced the same problems after replacing the sensor, other electronic parts and the boiler safety valve. Last i heard he was waiting for a new boiler from Lelit to place in the machine.

Since I was without a working machine for a few months at that point, Jim gave me a new machine which has worked perfectly for two months using Brita tap water and the Lelit in tank ion exchange softener. In my mind, this does not prove water was the culprit since the new machine has all new parts from various production runs, etc. any of which may actually contribute to the problem.

I am not sure what to make of a 'film' causing wild, inaccurate temperature fluctuations. What exactly is the composition of this 'film' and how does this interact with the probe - Potassium? Bicarbonate? Or some corrosive material as a result of using rpavlis water between 60 mg/l and 100 mg/l? What to make of Jim's experience with the machine at his shop as he systematically replaced several parts using whatever water he routinely uses. My machine is finally making great coffee using the in tank softener but that isn't definitive proof water caused the problem. I would nice to substitute rpavlis water now with the new machine as a rechallenge, but I am going to leave well enough alone!

I do know Jim at 1st Line solved my problem and I am a happy customer who is drinking great coffee. Hope someday the cause or causes for the interaction of various combinations of machines and waters becomes known.

Did Jim try replacing the Gicar unit? It could be something as simple as a bad solder joint to the connector.

I confess, the more I hear about the observables here, the less convincing the "film on the probe" hypothesis seems. I'm not certain, but it is probably an NTC thermistor in the probe, which is just a resistor with a coefficient that changes as a function of temperature. The thermistor is sealed off from the boiler water, and only senses the temperature of the probe case. If I understand your post and Clutch's correctly, temperature measurements were 'wild' and 'erratic' which I would read as immediately swinging between values differing by large values. The probe case temperature cannot do this, not even with some exotic film. It sounds electrical. Now if moisture crept inside the probe, causing intermittent grounding, that could do it. But you (bzarycranski) had the same problem after changing the probe, so this seems unlikely too.

dlcain makes a good point above. By taking out the probe and cleaning it, is it possible customers are fixing their problems by unconnecting and reconnecting their probes? The one thing this doesn't explain is the correlation Jim noted between some water formulas and this issue. It is also possible that there are two different issues here with the same symptoms.

Is there any chance the probes have a weakness in the seal of the thermistor to the steam? I don't know how the probes are built, but if any amount of liquid gets to the thermistor is will exhibit erratic behaviour. So I wonder if those exact same probes are used by other manufacturers. I also think it could be a connection issue.. cold solder joint or the like.. but some of these are happening on newer units and usually that takes more time to occur.

What would happen if one simply dried the same probe with heat, without cleaning it.. putting it back in.. and see what happens..

I think what is needed is data as to the exact manufacturer and batch of the probes and start tracking those.. see who else is using them.. and try to find some common characteristics.