The chemistry of scale in espresso machine boilers [FAQ]

Water analysis, treatment, and mineral recipes for optimum taste and equipment health.
User avatar

#1: Post by rpavlis »

I hate to make posts that seem to be chemistry lectures, but perhaps this summary of the chemistry of scale might be helpful. This is an attempt to put the chemistry of scale into a single post.

Scale formation is a serious problem with many (maybe most) water sources. The "scale" is usually either CaSO₄, CaCO₃, or MgCO₃.

Most scale is calcium carbonate, CaCO₃. Often it is contaminated with MgCO₃.

Many rocks are basically solid CaCO₃ and/or MgCO₃. These rocks slowly dissolve in ground water because ground water is in contact with air which contains carbon dioxide. The following series of reactions occurs.

CO₂ + H₂O ⇌ H₂CO₃
H₂CO₃ + CaCO₃ ⇌ Ca²⁺ + 2HCO₃⁻

Ca²⁺ + 2HCO₃⁻ ions are stable in water solution. However, solid Ca(HCO₃)₂ is unstable. Attempts to produce it by concentrating solutions results in the reactions shown above proceeding in reverse.

Since both of the above reactions are reversible we have a problem! When we heat solutions containing HCO₃⁻ and Ca²⁺ reverse reactions result in loss of CO₂ which results in separation of CaCO₃. Thus we get scale.

Adding water solutions of weak acids to CaCO₃ will result in the following reaction, shown with acetic acid, which chemists tend to abbreviate as HOAc.

2HOAc + CaCO₃ ⇌ Ca²⁺ + 2⁻OAc + CO₂

The carbon dioxide, CO₂, is a gas so it is lost from the solution driving the reaction forward. This is the descaling reaction when acetic acid is used, other weak acids such as citric and formic behave similarly.

Magnesium compounds in rocks give similar reactions, and descaling reactions are also analogous.

When concentrations of Ca²⁺ and/or Mg²⁺ are low enough relative to carbonate the solid carbonates will not separate. However, steaming concentrates ions in the water, so even water that appears to have low concentration of these ions can produce scale when a lot of steaming is done.


There are calcium sulphate minerals. Calcium sulphate is weakly soluble in water, about 2 grams dissolve per litre. Unlike most compounds, CaSO₄ is LESS soluble in hot water than cold. When milk drinks are prepared, single boiler machines remove steam which leaves the dissolved ionic pollutants behind. Everything in the water is concentrated. Furthermore the boiler is HOT, and soon we find CaSO₄ precipitating inside the boiler. When the machine is used without dealing with the problem the amount of calcium sulphate continues to grow. The more steam is produced, the more solid forms.

Weak acids like acetic, formic, and citric do not produce enough H₃O⁺ to dissolve CaSO₄.

Sulphuric acid, H₂SO₄, is a strong acid, meaning that when you dissolve it in water it reacts with the water almost completely by the reactions:

H₂O + H₂SO₄ ⇌ H₃O⁺ + HSO₄⁻
H₂O + HSO₄⁻ ⇌ H₃O⁺ + SO₄²⁻

When CaSO₄ is treated with aqueous sulphuric acid solutions the SO₄²⁻ ion is partly converted into HSO₄⁻ by the second reaction above proceeding backward. Thus one can dissolve CaSO₄ in aqueous sulphuric acid. However, the H₃O⁺ in aqueous sulphuric acid also is capable of dissolving a LOT of other things, such as the zinc in brass!

Zn + 2H₃O⁺ ⇌ H₂ + 2H₂O + Zn²⁺

There are several compounds that can form complexes with calcium that can remove the CaSO₄, but many are hazardous and none are very convenient.

Certainly the safest way to remove CaSO₄ is to take advantage of the fact that 2 grams of it dissolve per litre of room temperature water. If one use pure distilled or good reverse osmosis water for making espresso, the solid CaSO₄ will slowly dissolve. This needs to be done until all the scale has been eliminated which will likely require several weeks.

When the pure water descale is in progress, it becomes saturated by CaSO₄. It will thus be very hard, and people who do not drink sulphate contaminated water all the time may notice this water tastes somewhat like like wall board!

The way to avoid having to deal with calcium sulphate scale is simple: Be sure the water does not contain much sulphate, and if calcium sulphate scale begin to appear, act right away, rather than wait until the boiler is a mass of solid.

My 1978 La Pavoni Europiccola had a LOT of CaSO₄ scale in it when I got it. It took about two months of using straight distilled water to eliminate it completely. When I finally got all of this scale out of the machine I noticed the water rapidly developed a thin tan precipitate! I discovered someone had dropped an M6 steel nut into the boiler. When the scale surrounding it got removed the nut rapidly reacted with the water because it was then in contact with the copper boiler. After I removed the nut, everything was fine.

User avatar
Supporter ♡

#2: Post by JohnB. »

rpavlis wrote: Weak acids like acetic, formic, and citric do not produce enough H₃O⁺ to dissolve CaSO₄.
I've found undiluted white vinegar (acetic acid) quite effective at dissolving the typical scale found in espresso machine boilers. It might take a few days of soaking to remove a heavy build up but it works better cold then hot citric acid solutions. It certainly is better then having to drink coffee made from distilled water for several months while you wait for the scale to dissolve into your shot.

User avatar

#3: Post by rpavlis »

Calcium and magnesium carbonate scales are easily removed with weak acids. If you had scale that dissolved without a LOT of changes over a lot of time it was not calcium sulphate scale. Calcium sulphate is about as soluble in pure water as it is in weak acid solutions. Calcium sulphate scale usually occurs when people steam a lot of milk.

The bad thing about using purified water to remove calcium sulphate scale is that the water is terribly hard from dissolving the calcium sulphate during the process. The flavour of sulphate hard water can be unpleasant to some people. When the machine is not filled with pollutants, distilled water tends to extract most coffee components better, but most tap water contains bicarbonates which tends to reduce acidity, and also favours extracting alkaline things from the beans a bit better. There is FAR more calcium and magnesium in coffee beans than in the water used to make espresso. Adding potassium or sodium bicarbonate to pure water increases the ionic strength less than equivalent amounts of calcium or magnesium, so allows better extraction, and it will not produce scale. You cannot do this when removing calcium sulphate scale. It must not have ions that form insoluble salts with calcium in it, or ions that can produce such salts. 100 ppm sodium or potassium bicarbonate gives close to the concentration of bicarbonate that you find in hard water. This small amount of bicarbonate also, by coincidence, makes the pH of the water near the value of greatest stability toward copper oxide coats which protect copper from air oxidation.

When one notices that weak acid scalers have no effect, the cause is likely that the scale is calcium sulphate, not carbonate.

User avatar
Supporter ♡

#4: Post by JohnB. »

rpavlis wrote:When one notices that weak acid scalers have no effect, the cause is likely that the scale is calcium sulphate, not carbonate.
That's when owning a bead blast cabinet comes in very handy!


#5: Post by noidea »

Thanks that's a very informative post, I sort of wished I had taken more notice in my chemistry classes 30 years ago....
It also makes me wonder what the inside of my steam boiler looks like (Dalla Corte Mini) as there is no hot water tap and I make a lot of milk drinks with it and have been wondering about scale buildup in the steam boiler. I only feed it bottled water as our local tap water is very hard, but I bought it used and are not sure what the previous owner did, he was a coffee machine tech so I hope he did the right thing and fed it decent water as well.

Regarding weak acids for descaling CaCO₃ scale, what are your thoughts on sulphamic acid over citric acid? Also do you know any thing about commercial descaling products? A lot of them state that they are activated or contain inhibitors but reading the MSDS for them doesn't give much away apart from them containing either sulphamic, citric or lactic acid and "other substances not deemed hazardous"

User avatar

#6: Post by rpavlis »

The ionisation of sulphamic acid is very high, it has a pKa of around 1. (i.e. It is on the border between being considered a strong acid and weak acid, and it is on the strong side of that border!) Thus it is not a good thing to put in brass boilers or boilers with brass fittings because the high acid concentration can "dezincify" the brass. One should not use sulphamic acid on espresso machines. Lactic acid is a stronger acid than acetic, about like formic.

If one can look into one's espresso machine's boiler, one should do so frequently to be sure there is not scale forming. Carbonate scale and sulphate scale have a similar appearance. However, carbonate scale will dissolve fairly quickly in white vinegar or citric acid solutions. Sulphate scale dissolves no better in white vinegar or citric acid solutions than it does in purified water. There are commercial sulphate scale removing products. Some have EDTA. There are other materials too. I have always thought that when people steam a lot of milk drinks it is best to use sulphate free water. It is always best to use chloride free water. The thing that is really good to have in espresso water is bicarbonate ion--as explained earlier, it makes the water slightly alkaline, right at the pH of greatest copper oxide coat stability, and it reduces the acidity of the product. By coincidence the pH that is best for boiler stability tends to be excellent for flavour.


#7: Post by jonr »

I'm interested in knowing more about inhibitors and how they can protect the metal while still removing scale.

User avatar

#8: Post by rpavlis »

Corrosion inhibitors are materials that tend to provide a protective film over things. Benzotriazole and some of its derivatives are famous material for using on boilers and the like. I do not think it is approved for food usages.

Phosphates are good corrosion inhibitors for copper and brass in low concentrations. Phosphate inhibition works best in slightly alkaline water. Even 1 part per million of a phosphate in water has a strong anti corrosion effect. That is just 1 MILLIgram per litre!!!!

Descaling solutions are acidic and tend to remove protective copper oxide films. Distilled water is inexpensive, it costs only about twenty five cents a litre in US grocery stores. It dissolves some carbon dioxide from the air and becomes weakly acidic, though most of the carbon dioxide is driven out when the boiler is heated. It is NOT violently corrosive. It does NOT "want" to react with everything. Its redox potential results in its NOT reacting with water. Its lack of electrical conductivity reduces electrolytic corrosion. Oxide coats will form on copper using it that are fairly protective, however, the oxide coats are somewhat more stable in slightly basic water, especially when that water contains traces of phosphates! Copper needs protection not from water, but from OXYGEN. Water that removes the oxide coats can result in corrosion unless the copper be kept in an inert atmosphere.

Tap water that is fairly clean--that is that has mostly Ca and Mg bicarbonates as the only contaminants is thus not very corrosive, except when there is too much of either Ca or Mg, one gets scale. The descaling solutions remove the protective oxide coats. As I pointed out before, there is much much more Ca and Mg in espresso beans than in the hardest water that would be used to extract them. The weak ion water softeners that remove bicarbonate remove this good contaminant!!!!! One can go to any grocery store and purchase a box of sodium bicarbonate. If one weigh out 4.2 grams of it and dissolve it in 100 mL of water one has a 4.2% w/v solution of it. If one add 2 mL of this bicarbonate to each litre of distilled or good reverse osmosis water, you end up with 84 mg/litre, which is 1 millimole per litre. This is about what is in slightly hard water, but it will not form scale. If you add a trace of phosphate to this, you will have it also inhibited. Because sodium is monovalent, its presence does not raise the ionic strength as much as Ca or Mg would, so extraction will be slightly better.

(I normally use potassium bicarbonate, and make a 10% solution of it, and add 1 mL of it. The effect is basically the same.)

Water compounded like this has no sulphate nor chloride. Both are present in the beans, however. Chloride IS the real villain for corrosion, not just with copper but with many other metals as well. To me it makes fantastic tasting espresso too. You can also adjust the bicarbonates for really acidic coffees, giving another element of control.

Bicarbonate buffers pH, and that can have a substantial flavour effect, especially on light roasts. Most of what some attribute to hardness is really due to bicarbonate's slight basicity and its buffering action that can improve extraction of acidic components in the beans, at least a bit.


#9: Post by OldNuc »

On an industrial scale I see Sodium Hydroxide in a 30-50% solution used to remove calcium sulphate. The inference was it is a fairly rapid scale removal process and not a hazard to brass or copper. It was on one of those sites where you have to give them 2 Jacksons to get the entire document. Do you have any further info on this method of sulfate removal?

User avatar
Team HB

#10: Post by another_jim »

Thanks for the run down
Jim Schulman