I am only too happy to agree that the source of contamination can be something other than leeching or precipitation, and relieved that I'm not crazy that not only should rpavlis water not be responsible, but that I am not making grave errors in preparing it. The explanation about TDS meter values makes sense, as it is doubtless based on conductivity measurement and in addition to temperature correction makes some assumptions in the calculation of TDS regarding expected ionic content. Thanks for that.

I haven't plumbed the Bianca drip tray and really can't see how anything could get from the drip tray into the boiler; see pics below. If it's from milk it should be organic based, and I could collect some and burn it in an oven at high temp to at least see if it's organic or inorganic. Not sure if I'd have enough of it. Other thoughts?

I like the theory about it being burnt milk. You said the wand is wiped clean, but is it solidly purged right after use?

millmountain wrote:Got a TDS meter on the way home, so now we have some data.
<snip>
The steam boiler very clearly has some new additions. Since it only gets used for steam, most of the water may have been in there for months.

I doubt your steam boiler has new additions so much as it has become concentrated over time with use of only the steam wand. You are essentially removing distilled water every time you use the steam wand, leaving any minerals and other dissolved solids behind. This means not just the additives for the rpavlis water, but also any other trace elements that may have been in the water you used to make it. Given enough time enough of something must have built up to precipitate out. I find it unlikely that the tank itself is corroding or otherwise reacting.

I would suggest you drain the boiler as thoroughly as you can, and refill it with distilled water with just a dash of the bicarbonate (much weaker than rpavlis). After this, run some water through the hot water wand periodically to avoid concentrate building up again.

@lessthanjoey, yes I mean that we purge the wand directly after steaming the milk. At the latest after pouring the milk.

Nate42, it's a fair point that we can expect the ppm to be higher in the boiler due to accumulation of potassium carbonate. Nice to hear another confirmation that it's not a corrosion issue. As I have used rpavlis exclusively after the first month (I initially used a different recipe with MgCl₂, NaHCO₃ and CaCl₂, see here), based on distilled water, I don't think this explains the particles as there are almost no trace elements, or not nearly enough to form so many particles. Maybe my distilled water is not as pure as I think, but I suppose the milk theory could "hold water."

Alternatively, if after reaching a certain concentration (due to drawing off steam over months and rarely draining) the potassium and carbonate does precipitate back out, I wonder if the gray appearance could be either a different substance (constituted from potassium, carbonate and carbon dioxide and formed at about 130°C) or just a very thin coating of such a different substance on the surface of particles of potassium carbonate precipitate-enough to change the color from the typical white appearance. The only photo I have so far is below, where I poured the water from the video in the original post through a paper towel.

As I recall, the particles appeared somewhat suddenly several months back, I believe during the summer. Could be they entered the boiler suddenly, as in via some milk incident; could be they built up gradually and I only noticed them suddenly due to lack of using the HWW. I agree at this point that draining the boiler should be a good option. I will try to figure out whether it's possible to analyze the particles. I am awaiting feedback from Lelit and will ask their recommendation on how to best drain the boiler if they agree. Still, any further links or recommendations welcome.

One lesson out of this is definitely to draw a few ounces of hot water through the HWW every day. I remember reading that important tip before getting the machine and taking it seriously, but it obviously didn't stick after I lost interest due to the weird taste of the water.

Meanwhile, the machine continues to make great coffee and lattes.
I definitely want to avoid the boiler filling issue.


Particles from the HWW on a paper towel after drying (40x)

That was a reasonable possible cause - - that the anti-vac hose was sucking up from the drip tray - - considering how very dirty your water is. But looking at pics of the Bianca I see that's not it. The Bianca does have a hose on the anti-vac valve, but like many E61's it appears to terminate (along with safety valve) at a little port well above the drip tray:





It would be good to know where all that dark sediment came from, but you may never know. Draining and flushing the boiler til it's clean as Nate42 suggests, and then keeping it periodically flushed is your next step. For rpavlis water a routine daily flush from the water tap should be sufficient. (Some discussion can be found here: Using hot water tap to manage steam boiler water concentration .)

That part nevertheless does not stay nearly as pristine as in that photo! I will now also be wiping and cleaning it with a bit of water on a regular basis.

Nice resource, your first post in that thread on how much water to draw. That will be fun to measure. I often switch baskets (single/double) or even portafilters and can use the hot-water tap to quickly heat them up. Maybe the taste will even eventually improve, and I can use it for americanos as I originally intended. Would still need to get friendly with the taste of rpavlis water, but I maybe I could play with adding some Mg/Ca concentrate for that.

I'm feeling much better about the situation. Thanks to all, and any further input still welcome.

I don't have a Bianca, but considering you mentioned that you received a "steam boiler fill alarm", I would probably want to pull the fill sensor out and inspect it... Perhaps there was a quality issue with the material of the sensor and it is "dropping" the particulates?

I've never had mine out of my synchronika, but maybe that could also give you a very small window to peer inside the boiler to see what the inner surfaces look like?

Josh

jrham12 wrote:I would probably want to pull the fill sensor out and inspect it... Perhaps there was a quality issue with the material of the sensor and it is "dropping" the particulates? . . . . maybe that could also give you a very small window to peer inside the boiler to see what the inner surfaces look like?

Good suggestion, Josh. This is a good opportunity for me to start learning about about machine internals. I will need to learn how to to remove the fill sensor. This seems pretty harmless and sounds like a good idea, but as the machine is under warranty would it be wise of me to have Lelit confirm what level of activity inside the machine is acceptable?

Nate42 wrote:I would suggest you drain the boiler as thoroughly as you can, and refill it with distilled water with just a dash of the bicarbonate (much weaker than rpavlis).

What is a recommended way of draining the steam boiler as thoroughly as possible? The Bianca doesn't have a drain plug, but this WLL video How To Drain a Dual-Boiler E61 Espresso Machine demonstrates a general procedure for DB machines. It seems just using the hot-water tap is enough for the steam boiler; turning the machine upside down with the mushroom nut off is for draining the brew boiler, which I don't need.

I could make 25 mg/l of the KHCO3 solution, but is there any reason not to use straight distilled for flushing, as long as I drain it all and refill with normal rpavlis water?

As I don't expect a first flush to do a thorough job of picking up particles, my idea is:

1. Drain the boiler
2. Refill with "flush" water
3. Drain the boiler, check color, smell and TDS of cooled drain water
4. Refill with "flush" water
5. Drain the boiler, check color, smell and TDS
6. Repeat flush/fill if necessary until color, smell and TDS are close to input values
7. Refill with rpavlis water, check color, smell, TDS and taste of cooled water from HWW to rpavlis water
8. Check color, smell, TDS and taste at the end of each week for four weeks with daily flushing

Your comments welcome.

[Edit: Forgot to mention that not only has the fill alarm not repeated, the pump run time during startup this morning was short as in the past.]

If you are going to flush more than once straight distilled is fine. I was proposing the dash of potassium bicarbonate for the water you leave in there, deliberately starting out weaker than the standard rpavlis recipe. Not that it's hugely important but you don't brew coffee with the boiler water so it's ok to be low tds. The dash of bicarbonate is to ensure it is not acidic.