Using water recipe to slowly "descale" an espresso machine?

[johyphenel]

I'm finally taking the plunge into water. I own a BDB machine, first unit eventually got replaced when I sent it in for repairs a few years ago. When I got the brand new replacement I was still reticent to dive all the way into water, so instead I tried to use bottled water to hopefully be an improvement on the really hard tap water I have.

The good news is that I picked Crystal Geyser at random, even better that I live in California and get the Mt Shasta / Weed source for CG in our stores; the bad news is that when it was out of stock during the pandemic we went with Safeway Refreshe brand as a replacement... When I checked that label again today it says "Source: from a municipal water source" ... So, my machine has some amount of scale now[3].

I've been reading the forum a bunch trying to understand more about water, and was hoping some of you with more experience could help validate my understanding in a few areas:

1. If I were to get a new machine (had my eye on the DE1XL recently..) would CG from Weed still be a good pick as-is for a no-scale solution? (CG water report 2021, screenshot attached[2])

I tried to enter this into the sheet myself (screenshot below [4]), but wasn't sure if I should use "Calcium" or "Hardness, Calcium" value..

Making rpavlis water & making water to slow descale the machine

2. For the base water, should a TDS meter read "0" before I add my potassium bicarb? (Is the ZeroWater TDS meter "good enough" for this check?)

IIUC, the final goal is to have 0.05g - 0.1g (50 - 100 milligrams) potassium bicarbonate (KHCO₃) per liter of water[1], with most people using the high side of that range. Since that's hard to measure accurately, two ways (of many) I've seen described on HB to make this reliably:

• Mix 10g KHCO₃ with 1L water for concentrate. Add 10g concentrate per 1L distilled water for final dilution.

• Mix 10g KHCO₃ with 100ml water for concentrate. Add 1g concentrate per 1L distilled water for final dilution.

3. Assuming that's correct for a solution that will not cause scale, is it also correct that reducing that ratio would help remove scale that is already there (perhaps faster?)? For example, if I add 25-50 milligrams KHCO₃ per 1L water would that be a good compromise for a slow-descale without losing all flavor? (Or am I misunderstanding how this all works..? )

My purchase list:

• Water: https://primowater.com/self-service-refill-water/ (alternative, tap + ZeroWater)

• KHCO₃: Saw someone here bought Pure brand, found on Amazon: https://a.co/d/6aZ9QKu

Please let me know any feedback you have


[1] "You can simply add from 250 to 500 milligrams of potassium bicarbonate to a 5 litre container." - 1977 La Pavoni Europiccola with green deposits/scale
[2] CG water report 2021

[3] Scale buildup I currently have:

[4] Scale & corrosion sheet for CG sourced from Weed, CA, in 2021

[homeburrero]

Compliments on a very clear and complete post!

1. If I were to get a new machine (had my eye on the DE1XL recently..) would CG from Weed still be a good pick as-is for a no-scale solution? (CG water report 2021, screenshot attached[2])

I think it would do very nicely. In a 130 ℃ steam boiler it is very slightly on the positive side in an LSI calculation, meaning that it might scale. And if you use the steam wand the water will become more concentrated. This can be alleviated by occasionally flushing and refilling your steam boiler with purified (distilled, RO, or de-ionized) water in your reservoir.

I tried to enter this into the sheet myself (screenshot below [4]), but wasn't sure if I should use "Calcium" or "Hardness, Calcium" value..

You got it right, note that if you mouse over that little upper right triangle it pops up "Ca** as CaCO3". I'm reluctant to trust anonymous spreadsheets, but did an LSI and verified that this one gives you a good calculation. Can't agree with the text about water being corrosive.

2. For the base water, should a TDS meter read "0" before I add my potassium bicarb? (Is the ZeroWater TDS meter "good enough" for this check?)

Yes, you can give it a sanity check on known water, like your tap water to verify that it's working. For this purpose any water that reads below 10 ppm or so should be good enough. For the Zerowater filters they tell you to replace the filter when it gets up to 6ppm, which is good advice - these can get fishy resin taste and odor when they are depleted. You can also use that meter as a sanity check on your finished rpavlis water. With the water at 25 ℃, 100 mg/L KHCO₃ should read in the 60 ppm ballpark on that meter.

3. Assuming that's correct for a solution that will not cause scale, is it also correct that reducing that ratio would help remove scale that is already there (perhaps faster?)? For example, if I add 25-50 milligrams KHCO₃ per 1L water would that be a good compromise for a slow-descale without losing all flavor? (Or am I misunderstanding how this all works..? )

Your recipes look fine. The standard espresso machine water advice is to keep the alkalinity above 40 mg/L for corrosion risk reasons, which would be 80 mg/L of KHCO₃ . In the case of R Pavlis water, which has zero chloride and sulfate, I think you can go lower than that without worrying. The lower alkalinity would give you a lower pH and would theoretically dissolve more scale. But in reality I think the descale effect difference would be negligible. Some people (not me) notice a taste improvement for some coffees if you use 50 mg/L rather than the full strength 100 mg/L. That would be only 25 mg/L (as CaCO3) alkalinity, but should fine with this water - - Dr. Pavlis had a lot of corrosion expertise and sometimes recommended the lower bicarbonate recipe for taste reasons.

[johyphenel (original poster)]

Thanks Pat! Glad I'm on the right track

Your recipes look fine. The standard espresso machine water advice is to keep the alkalinity above 40 mg/L for corrosion risk reasons, which would be 80 mg/L of KHCO₃ . In the case of R Pavlis water, which has zero chloride and sulfate, I think you can go lower than that without worrying. The lower alkalinity would give you a lower pH and would theoretically dissolve more scale. But in reality I think the descale effect difference would be negligible. Some people (not me) notice a taste improvement for some coffees if you use 50 mg/L rather than the full strength 100 mg/L. That would be only 25 mg/L (as CaCO3) alkalinity, but should fine with this water - - Dr. Pavlis had a lot of corrosion expertise and sometimes recommended the lower bicarbonate recipe for taste reasons.

If I understand this correctly, it sounds like the full strength rpavlis water recipe of 100mg/L KHCO₃ should already dissolve some scale (even at steam boiler temps), and going down to the half strength 50mg/L KHCO₃ would only be negligibly faster, is that correct?


Thanks for the fast response, excited to try this out over the next year and see what happens!

[homeburrero]

If I understand this correctly, it sounds like the full strength rpavlis water recipe of 100mg/L KHCO₃ should already dissolve some scale (even at steam boiler temps), and going down to the half strength 50mg/L KHCO₃ would only be negligibly faster, is that correct?

Yes, I'd call it negligible because I think either would be slow as molasses in dissolving hard limescale. My reasoning about that is here: Descaling using distilled water? .

Just looking at the pH difference between 50 mg/L and 100 mg/L KHCO3, I think it would be only around 0.3 or 0.4 pH points. But pH is a log10 scale so the difference in [H+] between the two is a factor of 2.5 or so.

[johyphenel (original poster)]

Thanks for the links! Collecting some quotes that stood out to me (with some added emphasis). Hopefully they will help anyone else who finds this topic.

Descaling using distilled water?

Using distilled to periodically flush a steam boiler in order to prevent problems due to concentration of minerals caused by steam wand use is a common and recommended practice: Espresso machine water tastes really bad

Espresso machine water tastes really bad

When you steam, you lose water, but not minerals. So the mineral content in the boiler builds up. You also can get some leached metals. The result is a brackish taste. This is not a factor in the coffee taste, since the espresso water passes through the HX, not the boiler.

However, mineral build up can cause scale and will spoil Americanos. Once a week to once a month (more frequently if you are in a hard water area), fill your tank with distilled water, and open the hot water tap. This will drain the boiler and replace the highly mineralized water with distilled water. This is called "blowing down" the boiler and should be done routinely with espresso machines.

Descaling using distilled water?

I'm pretty sure your machine has a boiler level sensor probe that requires some minerals to work, so pure distilled would not be a good idea.

The next best thing would be to use distilled that has a small amount of bicarbonate added. Add a tiny pinch of sodium bicarbonate or potassium bicarbonate, in the neighborhood of 0.05 to 0 0.1 gram per liter of distilled. It will tend to dissolve limescale, albeit more slowly than would pure distilled water, which is slightly acidic. But since it has that alkalinity you can safely use it as your routine water and you have less need to know when/if it has cleared out limescale in your machine.

What was said about the water level sensor -- it will not function properly with distilled. Don't ask how I know this!

I guess an important thing for folks with machines like the BDB that have a water level probe, is to not use 100% distilled water for this flush! Good to know!

--------------------

This also stood out (emphasis mine):

Descaling using distilled water?

I wondered how much scale will dissolve in distilled water, do you have any idea/guess? Heat will help, yet using Bicarb will indeed slow things down, just wondering if the proces takes on paleontological periods or more humane...

Good question. I think it would be very slow, but maybe not paleontological. As a rough guess, we know that RO water very slowly trickling through a calcite filter might dissolve at a rate of 30 mg/L. That calcite is finely ground, not hard deposits, but if you use that as an upper guess, and assume you have 1 liter per day of throughput and 5 grams of limescale deposits, it still would take over 6 months to dissolve. Water with a little bicarbonate will have less carbonic acid and would be slower, heat would also make it slower (calcium carbonates become less soluble at high temp), and of course the small surface area of the limescale deposits would make it slower. I'm guessing that it might take a decade to remove a few grams of limescale.

So it sounds like, in the conditions of an espresso machine, scale has a much easier time being developed than removed. Yet another reason to get your water right early I guess - glad I'm figuring this out before my next upgrade!!

Thanks again! Really appreciate you taking the time to help me understand