Good references on water treatment for coffee/espresso - Page 2

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

Postby RyanJE » May 27, 2016, 11:20 am

Thanks! Added to the original post.

Hope you are making headway regarding corrosion.
I drink two shots before I drink two shots, then I drink two more....


Postby Zoey » May 28, 2016, 3:13 am

User avatar
Team HB

Postby homeburrero » Sep 20, 2016, 2:25 am

Inspired by this summer's release of the SCAE water chart report (Full title: The SCAE Water Chart: Measure Aim Treat, by Marco Wellinger, Samo Smrke and Chahan Yeretzian, Zurich University of Applied Sciences), I decided to create a graph similar to theirs that depicts their recommended core zone for hardness/alkalinity along with previous work, including the 2015 Water for Coffee book (Colonna-Dashwood and Hendon), the 2011 SCAA Water Quality Handbook (Beeman Songer, and Lingle), the 2002 ( Jim Schulman's Insanely Long Water FAQ, plus a few points of interest from bottled or formulated water that have been discussed here on the home-barista forum. (Note: I did run my graph by Marco Wellinger, as well as Maxwell Colonna-Dashwood to make sure they were OK with posting my representation of their data here.)

There's nothing new here - it's a simple representation of frequently discussed water specs in one hardness:alkalinity graph. It may be interesting, perhaps educational for analytic types wanting to understand this stuff, and hopefully useful in pointing out the wide variety of waters that have been found to work well for coffee brewing, as well as the areas of agreement across many different recommendations.

I should point out that while a hardness:alkalinity analysis is more sophisticated than, say, a TDS meter measurement, it may still sometimes be an oversimplification. Other ions, especially chloride and to a lesser degree sulfate can be very important in deciding if the water might cause corrosion problems. Also, the proportion of magnesium to calcium may make a small difference in taste and a big difference in scale deposits.

For a definition of the terms hardness and alkalinity, and units and conversions for quantifying those measures see my follow-up post.

Explanation of the shaded areas, zones, and points on the graph are provided in the text below the graph image. Some key references are included at the end of this post.


Pink/Green Shaded Areas
The area lightly shaded in green should be relatively scale free. Within that shaded area it should deposit no scale at temps of 120C or below (appx. 1 bar gauge pressure at sea level).

The area lightly shaded in pink might be excessively scale-prone. At the border of that shaded area, a boiler at 95C might drop about 12 mg of scale per liter of throughput, and at 130C it might drop about 30 mg per liter.

The white zone between the two shaded zones is an area where the water may or may not deposit scale, depending primarily on the boiler temperature.

Both of those shaded areas are based on the scaling calculations from Jim Schulman's modified Langelier/Puckorius equation as described in his FAQ. If your water has significant magnesium (which means relatively less calcium hardness) the borders of these zones would shift upward on the graph.

The best source for understanding and calculating scale deposition rates is Jim Schulman's Insanely Long Water FAQ (see reference section for links).

The hardness=alkalinity Line
In natural waters, for points above this line, "temporary hardness" = total alkalinity. And for points below this line "temporary hardness" = total hardness. Temporary hardness here being the calcium and magnesium that is associated with carbonates and which might precipitate (i. e., form scale) when the water is heated. For points above the line the remainder of the total hardness is "permanent hardness" which is due to soluble calcium or magnesium salts of sulfate, chloride, or possibly a small amount of nitrate.

The hardness = 1.8 * alkalinity Line
This line represents a non-carbonate hardness to alkalinity ratio of 0.8. If the water were composed entirely of calcium and magnesium salts, and if the only anions of those salts were carbonates, chlorides, and sulfates, then this line would correspond to a Larson-Skold index of 0.8. ( ... -Skold.htm) The Larson-Skold index is sometimes used as an indicator of corrosivity, with values in excess of 0.8 being where you might have corrosion concerns with mild steel in water. The SCAE core zone's top border follows this line as a way to keep their recommendation away from potentially corrosive water. Note - this line looks slightly different from the dashed line in fig 6 of the SCAE doc. I think this one is correct (fig 6 of the pdf was evidently distorted a bit in final layout. See SCAE Water Chart is available online )

The C-D & H "Roast Correctable" Zone
The area between these two lines is within a zone discussed in the 2015 Colonna-Dashwood and Hendon Water for Coffee book, chapter 12. It illustrates how a roaster might modify roast parameters to produce a coffee that tastes good when brewed with a local water that is far outside the usual ideal brewing windows, and this roasted coffee might taste better with the water under which it was developed than it would with 'ideal' water.

The C-D & H Ideal zone (appx)
This is also from the 2015 Colonna-Dashwood and Hendon Water for Coffee book, Chapter 10. This zone is their "Ideal brew zone." In their book they also present a larger, and slightly differently shaped "acceptable brew zone" (not shown here) and discuss how they arrived at the borders of those shapes based on a combination of theory and experiment. It's important to note that their recommendation zones are all about water for coffee, as opposed to water for espresso machines.

For those of you who are looking at the Water for Coffee book, you probably noticed that my graph has the axes are labeled "hardness as CaCO₃" and "alkalinity as CaCO₃". In the book, the axes are labelled "General Hardness (GH, [Ca²⁺] + [Mg⁺⁺] in ppm)" and "Carbonate Hardness (KH, [HCO₃⁻] in ppm)". Those labels have caused a little confusion as to whether the units were meant to be measures of ion concentration, or of GH and KH as CaCO3 equivalents. (See "Water For Coffee" book discussion thread for more detailed discussion of this.) I'm confident this graph correctly represents their zone, and did run it by Maxwell before posting it. It's also the same as Marco Wellinger's representation of their zone in the SCAE water chart.

The SCAA "Superior Brew" and SCAA "Adequate Brew" zones and the SCAA target

The "superior" and "adequate" brew zones are both from the 2011 SCAA Water Quality Handbook. For the superior zone, the recommendation for alkalinity is "at or near 40," so the width of that zone on this graph is my interpretation of what values "at or near 40" might be. Note that there is also an online SCAA table for an "acceptable brew zone" that is slightly different from the 2011 "adequate" zone from the handbook. Both recommendations can be found in the table below:

The SCAA guidelines specify a calcium hardness rather than total hardness, so if your water has significant magnesium you would shift all these SCAA zones upward in my graph.

The SCAE "Core Zone"
This is an eyeballed representation from the SCAE water chart report by Marco Wellinger et al., specifically from Figure 8: SCAE core zone as recommendation for espresso machines and hot water boilers.

It's interesting to note the borders of this zone. The left border is roughly where the alkalinity might be getting too low to prevent corrosion due to acidity, and the upper border follows a possible Larson-Skold index, where above the line you might expect corrosion due to chloride and sulfates not balanced by bicarbonate. The right border is where you might be getting into high scale deposit problems, and the lower border follows recommendations from Colonna-Dashwood & Hendon's Water for Coffee book for optimal taste.

(Initially, this report was freely available*, and is a highly recommended read. See the references section below.)
*note: As of Dec 2016 that report was no longer openly available to non SCAE members.

Volvic is a frequently recommended bottled water on this forum. As natural waters go, it is somewhat high in magnesium - the calcium hardness is only 30 mg/l and the magnesium hardness is 33 mg/L (63 mg/l total hardness). So it should not cause scale problems even though it's pretty far away from the green 'scale free' zone. The other potentially objectionable minerals in Volvic are reasonably low. Chloride, at 14 mg/l in this report is just within the Synesso espresso machine max recommendation, which is probably the most stringent out there. This Volvic data is based on a 2012 analysis and report: ... t-2012.pdf

"70/30" water
A formulated mix proposed and used by 5 Senses Coffee in Australia - ... recipe.pdf, and discussed here on HB: 70/30 Water . This recipe uses 70 ppm sodium bicarbonate and 30 ppm Epsom salt, hence the "70/30" moniker, and comes out to a hardness:alkalinity of about 25:42. There is no calcium hardness in this water and it should not scale. There is no chloride, and the sulfate is relatively low - the Larson-Skold comes out to about 0.6 so there shouldn't be a corrosion concern with this water.

100:50 GH:KH
A popular mix by folks on Barista Hustle forums, including Matt Perger. Discussed on HB here: Matt Perger's water recipe for coffee - Is it ok/safe for espresso machines? What do you think?
Most of the people making this formulation are using a mix of Epsom salt plus sodium and/or potassium bicarbonate. This particular mix is all magnesium hardness and no calcium hardness, so should probably not scale despite the high hardness. Although it has a healthy alkalinity, it might be a corrosion concern because of the relatively high sulfate/bicarbonate ratio (the Larson-Skold index on this water is 2.0).

Jim Schulman FAQ
This point is at 90 mg/l hardness and 50 mg/l alkalinity, and is mentioned in a couple places in the Jim Schulman FAQ. It's an example of water that should cause neither corrosion nor excessive scale problems as well as produce a tasty espresso.

Gerber Pure
Is often recommended as a reliably soft non-scaling water, and is one example of a purified drinking water that has a small amount of minerals added to RO or distilled water.

Scott Rao
Both of the Scott Rao books listed in the references below contain this recommendation. His full spec is:
TDS 120-130 mg/L
pH 7.0
Hardness 70-80 ppm
Alkalinity 50 ppm

2016 WBC
The water for the World Brewers Cup and Barista Championship has generally followed the SCAA target. For Dublin, 2016 they changed to a slightly softer water than in previous WBC events.
See ... edited.pdf

Hard (~6 gpg) water after conventional softening
This is just one example of a conventionally softened (strong acid cation exchange) water, as you typically find in home water softeners and in many commercial coffee shops in areas of hard water.

1 mmol/l KHCO3 (rpavlis)
This is a popular water among home barista members, as used by chemistry professor Robert Pavlis (rpavlis on HB). Is purified water with 100 mg/L of potassium bicarbonate added. Is easy to make, will not scale, and has enough alkalinity to avoid corrosion in vintage equipment. Despite having no hardness at all, lots of folks find that it makes perfectly tasty espresso. Dr. Pavlis would use this at as little as half strength (25 mg/L alkalinity) when pulling darker roasts.


Schulman, J., Jim Schulman's Insanely Long Water FAQ, posting, 2002, ... erfaq.html and ... %20FAQ.pdf

Colonna-Dashwood, M & Hendon, C. H., Water For Coffee, Self-Published, 2015,

Scott Rao, The Professional Barista's Handbook, Fourth Edition, Independent Publisher, 2011,

Scott Rao, Espresso Extraction: Measurement and Mastery, Self-Published 2013, ... B00F2VCTP6

Specialty Coffee Association of America, SCAA Standard | Water for Brewing Specialty Coffee, November 2009, online PDF:

D. Beeman, P. Songer, T. Lingle, Specialty Coffee Association of America, The Water Quality Handbook, Second Edition SCAA 2011, ... nt-version

M. Wellinger, S. Smrke and C. Yeretzian, The SCAE Water Chart: Measure Aim Treat, (Specialty Coffee Association of Europe, 2016), (Initially was freely accessible via the preceding URL, as of Dec 2016, appears to be available to SCAE members only: ... ater-chart )
Edit addition Note: If you are not an SCAE member, you can get much of the same information from Chapter 16 of the Craft and Science of Coffee book, available for free download here: ... _Treatment
nínádiishʼnahgo gohwééh náshdlį́į́h

User avatar
Team HB

Postby homeburrero » Sep 20, 2016, 2:41 am

This is a follow-up post to my previous hardness/alkalinity graph. It's here to provide a short and succinct definition of the terms 'hardness' and 'alkalinity,' especially with respect to how they are typically quantified 'as CaCO3.' For a more verbose and instructive discussion, I recommend reading the Jim Schulman Insanely Long Water FAQ, viewing Marco Wellinger's presentation here: ... extraction and/or reading this nice tutorial from the beer brewing community: ... ter_report

Defining and quantifying hardness and alkalinity

The terms 'hardness' or 'total hardness' can be most simply defined as the sum of dissolved calcium (Ca²⁺) and magnesium (Mg²⁺) ions in the water. General hardness (GH) is another term for total hardness.

Sometimes hardness is expressed as calcium hardness and total hardness, with the difference between the two being the magnesium hardness. It also may be separated into 'temporary hardness' and 'permanent hardness,' where temporary hardness refers to calcium and magnesium that is associated with carbonates, and which will precipitate when the water is heated. Permanent hardness is associated with non-carbonate ions, typically chloride or sulfate, and does not precipitate or form scale when the water is heated.

The terms 'alkalinity or 'total alkalinity' are defined to be a measure of the water's capacity to neutralize acid (H⁺ ions). In natural water it is essentially a measure of the bicarbonate (HCO₃⁻) and the carbonate (CO₃²⁻ ) ion content of the water, which neutralize acid per the reactions:

CO₃²⁻ + H⁺ —> HCO₃⁻
HCO₃⁻ + H⁺ —> H₂CO₃ —> H₂O + CO₂

Alkalinity is measured by adding acid to a water sample and counting how many protons (H⁺ ions) are neutralized in the process of getting the water down to a pH of around 4.3, which is where nearly all HCO₃⁻ has converted to H₂O and CO₂. For practical purposes, alkalinity, total alkalinity, and KH are synonymous. "Carbonate hardness" is sometimes used as a synonym for alkalinity, but also has another very different meaning (discussed in more detail below), so I don't use that term as a synonym for KH or alkalinity.

Since neither hardness nor alkalinity is is a measure of a single specific ion, they are both expressed in chemical equivalence units. Some common units are: grains per gallon, French degrees, German degrees, and milliEquivalent per liter. The most commonly used unit is mg/l (or ppm) of CaCO₃ equivalent. In fact, whenever you see a water report that expresses hardness or alkalinity in either ppm or mg/l, you can safely assume that the measure is in CaCO₃ equivalents even if not stated as such.

Expressing both hardness and alkalinity in the same units is always a good idea, because you can then directly compare the two to get an idea about how much of the hardness might be scale-forming. (In natural water, if the alkalinity is greater than the hardness, then the temporary hardness is roughly equal to the total hardness. If the alkalinity is less than the hardness then the temporary hardness is roughly equal to the total alkalinity. )

It's very easy to convert between other units and mg/l (ppm) as CaCO₃.

To get ppm (or mg/L) as CaCO₃ , multiply ...
grains per gallon (gpg, sometimes shortened to "grains") by 17.1
German degrees (dGH, °dH) by 17.8
French degrees (°f, °fH) by 10
English degrees (Clark degrees, °e, °Clark) by 7.02
mmol/l of divalent ion (e.g., Ca²⁺, Mg²⁺, CO₃²⁻ ) by 100
mmol/l of univalent ion (e.g., HCO₃⁻ ) by 50
mEq/l (milli equivalent per liter of any ion or compound) by 50

mg/l (ppm) of [Ca²⁺] ion by 2.5 (to get calcium hardness)
mg/l (ppm) of [Mg²⁺] ion by 4.12 (to get magnesium hardness )
Then add calcium hardness and magnesium hardness to get total hardness

mg/l (ppm) of [HCO₃⁻] ion (bicarbonate) by 0.82 (to get bicarbonate alkalinity)
mg/l (ppm) of [CO₃²⁻] ion (carbonate) by 1.67 (to get carbonate alkalinity)
Then add bicarbonate alkalinity to carbonate alkalinity to get total alkalinity.

Of course to convert the other way, from CaCO₃ equivalent to the unit on the left you divide rather than multiply. Be aware though that if you just know the total hardness there is no way to convert that to [Ca²⁺] + [Mg²⁺] ion, because you don't know how much is calcium vs magnesium hardness. If you have total alkalinity, and want to convert that to bicarbonate [HCO₃⁻] ion, you can usually assume that there is no significant carbonate [CO₃²⁻] ion, and just divide by 0.82. (Until the pH gets above 8.3 or so, there is very little carbonate.)

Because of their nice round conversion factors, French degrees and CaCO₃ equivalents are handy units for water chemists because it's so easy to multiply or divide by 10 or by 100 to get mmol/l, which is the preferred unit for chemists.

Some terminology wrinkles

For drinking/coffee water purposes, ppm and mg/l are the same. That's because a liter of pure water weighs 1 million milligrams. Likewise, epm (equivalent per million) and mEq/l are the same.

GH, general hardness, total hardness, and hardness are synonymous.

KH stands for Karbonathärte, and is typically a shorthand for alkalinity. Sometimes you'll see the statement that alkalinity is not exactly the same as true KH, but that is a quibble. For our coffee water it is essentially the same. In the popular titration kits that measure 'KH', they indeed are using a standard alkalinity measurement method.

"Carbonate Hardness" is often used as a synonym for KH and for alkalinity, and at the moment is primarily defined that way in Wikipedia. But that is a potentially confusing usage — under that definition a solution of nothing but sodium bicarbonate would have "carbonate hardness" even though it has no actual hardness minerals at all. The alternative (and to me, a sensible, useful, and less confusing) definition is the alternative meaning, which is barely mentioned in Wikipedia, and is used in the SCAE Water Chart discussions: Carbonate hardness is the part of the total hardness that is associated with carbonate anions. Viewed the other way around, it also is the part of the alkalinity that is associated with magnesium and calcium cations. So a strong solution of nothing but sodium bicarbonate would have high alkalinity but zero carbonate hardness, and a strong solution of calcium chloride would have high hardness but zero carbonate hardness. Under this definition, carbonate hardness is roughly the same as so-called temporary hardness, because it will precipitate out as scale when the water is boiled. (Note: the Claris/Pentair technical literature also uses this definition of carbonate hardness.)

"Total alkalinity" and "alkalinity" are synonymous. Alkalinity titrations are sometimes done in multiple steps in order to estimate hydroxide alkalinity, carbonate alkalinity, and bicarbonate alkalinity. P-Alkalinity and MO-alkalinity are examples of titration endpoints that are used to make this distinction, which would only be done in very high pH water.

Some people prefer the term 'buffer' when talking about alkalinity, especially when speaking to non-technical audiences. I believe they do that to avoid confusing 'alkalinity' with the term 'alkaline,' which is imprecisely defined and typically used to describe any basic (e.g., pH greater than 7.0) water. For technical discussions it's fine to use the term alkalinity rather than buffer — the term is clearly defined in technical literature (there is a standard EPA procedure for measuring alkalinity, for example) and is fairly well understood in common parlance among aquarium, beer, water utility, and coffee people.

"Hydrogen carbonate" is a more proper term for bicarbonate. I.e., a chemist might say "sodium hydrogen carbonate" while a brewer would say "sodium bicarbonate," and a baker might say "baking soda." Same thing. Water reports still seem to favor the term bicarbonate.
nínádiishʼnahgo gohwééh náshdlį́į́h

Bak Ta Lo

Postby Bak Ta Lo » Sep 20, 2016, 2:50 am

homeburrero wrote:Inspired by this summer's release of the SCAE water chart report (Full title: The SCAE Water Chart: Measure Aim Treat, by Marco Wellinger, Samo Smrke and Chahan Yeretzian, Zurich University of Applied Sciences), I decided to create a graph ...

Thank you, this is great! I am switching over to making my own water using 15 liter bottles of distilled water with a flojet+accumulator, plan to test a few of these. I am starting off with the 100:50 GH/KH, Matt Perger formula, and the formula that "keno" posted.
LMWDP #371


Postby Markant » Sep 20, 2016, 6:39 am

Just what I needed!
Thank you!!


User avatar

Postby CoffeeBar » Sep 20, 2016, 6:50 am

Good Read and Thank you for this post :D

User avatar

Postby icantlactate » Oct 14, 2016, 12:53 pm

Homeburrero is the man!

User avatar

Postby NelisB » Nov 17, 2016, 6:35 am

What can I add to Volvic to make it a little less "light"?

Today I started to add tap water, (1:4), which is very hard here in the Netherlands, (conductivity 44,7 mS/m @20C). It makes the total hardness aprox 80 mg/l, which is good, but the TDS is like 180 ppm. And thats too high, right?

Thanks for your tips!

User avatar
Team HB

Postby homeburrero » Nov 17, 2016, 4:38 pm

NelisB wrote:What can I add to Volvic to make it a little less "light"?

Today I started to add tap water, (1:4), which is very hard here in the Netherlands, (conductivity 44,7 mS/m @20C). It makes the total hardness aprox 80 mg/l, which is good, but the TDS is like 180 ppm. And thats too high, right?

Hard to say without knowing what's in that tap water, and what you mean by 'less light'. If you're like some people, who think that their alkalinity is buffering and dulling the acidity and detracting from taste, you may be wanting more hardness, and with the same or decreased alkalinity. But if your tap water is mostly calcium carbonate, then you will be increasing both the hardness and the alkalinity by adding your tap, which may be contrary to your goal.

At 80 mg/l total hardness, and if it's 80% Volvic, I don't think you're pushing a scale problem because the calcium hardness would still not be that high.

180 ppm on a TDS meter is a tad higher than ideal, but again, depends on what's been added in there from that tap water spike. If the tap water is high in chloride or sulfate ion, and you plan to use it in an espresso machine, then adding it might be a bad idea because of corrosion risks.
nínádiishʼnahgo gohwééh náshdlį́į́h