Published: Jul 29, 2019

DIY Water Recipes Redux

A couple of years ago, we published a method for making your own water recipes, using concentrated mineral solutions, diluted with deionised water, to make a range of waters with different hardness and alkalinity. This post updates that, with new recipes that allow you to easily target a specific GH and KH in your water.

We’ve also added in a calculator that allows you to work out what will happen if you add minerals to existing water, rather than just deionised water. This will be useful to anyone in soft water areas who would like to remineralise their tap water, for example.


The Concentrate Recipes

You’ll need the following before you start:

  • Baking Soda – NaHCO3, Sodium Bicarbonate (not to be confused with baking powder)
  • Epsom Salts – MgSO4.7H2O, Magnesium Sulphate.¹
  • Deionised/Distilled/Ultra-pure water
  • Scales (accurate to 0.01g)
  • 3 x ~1L water containers (preferably glass, and odour/residue free)

The Buffer Solution

Dissolve 1.68g of sodium bicarbonate in 1L of deionised water. This creates a solution with a KH (as CaCO3) as close as we can get to 1000ppm.

The Hardness Solution

Dissolve 2.45g of Epsom salts in 1L of deionised water. This creates a solution with a GH (as CaCO3) of 1000 ppm.


Create Your Own Water Recipe Using Deionised Water

Using these two solutions with deionised water is very simple. To get your desired KH and GH, you can simply use that number of mls of each solution, then make the total volume up to 1L with deionised water.


Mineralise Pre-existing Water

If you prefer to add hardness or buffer to existing water, perhaps to take advantage of the calcium or other minerals already in the water, then you can use this calculator to work out what the final KH, GH, and TDS of your water will be. This is helpful if just want to boost the hardness of your favourite bottled water, for example, or to mineralise your tap water if you live in a soft water area.

To use this calculator, measure the KH and GH of your existing water (and optionally your TDS), then put these numbers into the calculator along with the amounts of each solution you plan to use.

You’ll see that adding 10ml of a solution doesn’t simply increase KH or GH by 10, like it does with distilled water. This is because the solutions themselves dilute the water that you started with. By tweaking the amounts of each solution that you use accordingly you should be able to work out what amount will get you your target GH and KH.


Why Are These Recipes Different?

Our previous solutions were designed to give you 1g/L of magnesium or bicarbonate ions, respectively. However, this is not the way that we usually measure hardness and alkalinity.

Rather than give the concentration of the ions directly, both hardness (GH) and alkalinity (KH) are usually measured in calcium carbonate equivalents. In other words, it tells you how many parts per million of calcium carbonate you would have, if all the hardness or alkalinity was due to calcium carbonate alone.

It’s measured this way because simple drop test kits can’t distinguish between calcium or magnesium ions, so it’s easiest to assume it’s all calcium. This means that when we start using magnesium or sodium salts to tweak the water, we need to do a bit of maths to convert those amounts into CaCO3 equivalents.

To make it easier for you to experiment with different hardnesses, we’ve altered the recipes to aim for a specific KH and GH instead. This means that you can easily target any GH or KH you like, with no conversion required.


The Water Recipes

These are the original Barista Hustle water recipes, updated for use with the new concentrates.

Recipe 1 – Melbourne

  • 11.5g Buffer
  • 23.7g Mg
  • 964.8g DI water

This is a close approximation to Melbourne water. This is very “soft” water, low in mineral content, and useful for those long filter brews or cuppings drawn out over five to ten minutes. Would also help with those darker espresso roasts that don’t need as much help extracting out flavours. 

Recipe 2 – WOC Budapest

  • 40.1g Buffer
  • 51.2g Mg
  • 908.7g DI water

This is in the target range for the World Brewers Cup in Budapest (51 mg/L total hardness as CaCO3, 40 mg/L alkalinity). In Budapest the total hardness would come from calcium as well as magnesium, leading to a different flavour outcome — competitors beware …

Recipe 3 – SCA

  • 40.1g Buffer
  • 68.6g Mg
  • 891.3g DI water

This is the official SCA specifications from the SCAA 2009 handbook. Similar to Budapest only the total hardness has gone up slightly. The specifications state a range of total hardness as low as 17 mg/L as CaCO3 up to 85 mg/L as CaCO3. So you could keep your buffer here constant at 40.1g and go as low as 17g of Mg solution or as high as 85g (don’t forget to subtract the total concentrates used from your DI water!).

Recipe 4 – Barista Hustle Water Recipe

  • 40.1g Buffer
  • 80.7g Mg
  • 879.2g DI water

The original Barista Hustle water recipe — where it all began. Add an extra 4.3g of the Mg concentrate and you’re at the top limit of the SCA specifications.

Recipe 5 – Rao Water

  • 50.1g Buffer
  • 75.7g Mg
  • 874.2g DI water

This is close to Scott Rao’s recommended water chemistry for brewing flavourful, balanced coffee. Slightly higher than the SCA in both total hardness and buffer, with a little more buffer than the BH recipe.

Recipe 6 – Hendon Water

  • 30.8g Buffer
  • 99.9g Mg
  • 869.3g

This is close to the centre of Christopher Hendon’s and Maxwell Colonna-Dashwood’s “Ideal Brew Zone”. If you’re inclined to “dial in” some water for a particular roast, this is a good starting point.

Recipe 7 – Pretty Hard

  • 35.1g Buffer
  • 126.1g Mg
  • 838.9g DI water

This begins the ascent up in water “hardness”, probably better suited short brew times for filter coffee. This is starting to grab a lot out from the coffee so brew recipes would need some adaptation. This rips everything out from the coffee. So either slow down or speed up the brew time via grind adjustments, and shorten or increase your beverage weight. Dependent on the roast somewhere along those two spectrums you’ll find something tasty. Or not.

Recipe 8 – Hard dot AF

  • 45.2g Buffer
  • 176.8g Mg
  • 778g DI water

This is a fairly high point with pushing mineral level where you’re basically cranking the amp up to 11. Your brew parameters from the earlier water recipes would need to change a lot here.


A Note About TDS

Note that the TDS given by this calculator might not be exactly the same as the TDS you would see if you measure the resulting solution with a TDS meter. This is because TDS meters assume a certain ratio of all the ions in the water, and by adding these solutions, we’re messing with that ratio.

¹ The “.7H2O” part refers to the fact that water forms an intrinsic part of the crystal form of this salt that you can buy in the shops, which is the clear crystals called Epsom Salts. Each magnesium sulphate molecule is surrounded by seven water molecules in this type of crystal. We’re specifying this here, as the weight of the water in the crystal affects the calculations.

Lastly, a warning: don’t put remineralised water in your espresso machine unless you’re very sure what you’re doing (in which case, you probably don’t need this post!). If you create super-hard water and scale up your kettle, it’s fairly easy to fix — but descaling an espresso machine is no fun at all.

October 2021

Mon Tue Wed Thu Fri Sat Sun
  • CN: 咖啡师认证
  • CN: 咖啡师认证


  1. rocky

    Thanks for such a clear and helpful resource!

    Maybe a little off-topic, but have you had any experience/seen any data around whether pushing water hardness upwards can help offset the reduced extraction achieved with lower brew water temperatures (eg, at high altitudes)? I imagine the effects of “temperature” and ions might differ with respect to specific molecular interactions, but maybe it’ll help a bit.

  2. 5ku77

    Another question:

    Can I use “Demineralized water” for these recipes?

    Thanks again

  3. 5ku77

    Hi, thanks for this article.

    Justo to clarify: In the Recipe 4 – Barista Hustle Water Recipe, the “Buffer” is the “Buffer solution” (Sodium Bicarbonate+DI water), but what is the “Mg” referring to? Is it the “Hardness Solution” (Epsom salt + DI water) or only “Epsom salt”?


    • 5ku77

      Ok it refers to the Hardness solution, the recipes are all for 1L of water.

      • BHLearn

        Hey 5ku77, yes, you’ll end up with 1L of usable water, but please note, the recipes are asking you to add specific amounts of the hardness and buffer solutions to a specific amount of deionised water. (So you’re not measuring out 1L of pure water and adding the solutions to that.)

  4. aslan.sobati

    Hello BaristaHustle.
    Please tell me which one of these recipes is suitable for light roast profiles? Also for medium roast specialty beans?

    • BHLearn

      Hi aslan.sobati, thanks for commenting. Trial and error with several different water recipes will give you the best result. But as a rough guide, if you ever find a particular coffee tastes a bit too sharp for your palate, you may find using a water recipe with a higher KH value will help. Just don’t overdo it. We generally find a KH above 70ppm mutes the taste of most coffees. BH

  5. mobomobile

    Can someone here rebuttal this video? , this guy sells espresso machines and stresses how using distilled or RO water with added minerals is very bad for machines with PID sensors, but would probably be bad for non PID machines also. I guess you need to know chemistry, so I’m trying to understand. If I get natural spring water checked and found out exactly which minerals, and how much of the minerals are in it, and then get distilled water and add those exact minerals and exact quantities, what would be the difference between the two? As many who have been researching water probably ran into Jim Schulman’s Insanely Long Water FAQ, he summarizes at the beginning to use a certain brand of natural spring water. I’m currently using water from an RO system that adds magnesium and calcium (before this I was using tap water through a Brita with a resin pouch inside). The RO has a TDS of 60, PH of about 7.5. I haven’t checked other parameters.

    • BHLearn

      It’s hard to fully evaluate the claims in the video, since they don’t name the specific mineralisation treatments they’re talking about. Limescale can form on temperature sensors (‘PID sensors’ are not a thing) and water level sensors and they should be cleaned periodically as part of normal maintenance – but a small amount of scale is normal.

      In general, you’re right that adding the correct minerals to distilled water can make an absolutely identical product to tap or softened water. Remineralisation kits often include different minerals though, which may have different effects on machines – you can see some examples in this post. Similarly, we don’t recommend our own water recipes for espresso machines as they don’t include calcium.

      If you want to check your own RO and be sure it’s safe, then you’ll need to do a bit more testing. You’ll need a calcium test kit and an alkalinity (KH) test kit – you can find both of these in pet stores that sell aquarium equipment. Depending on how the mineralisation part works, you may also find that the mineral content varies depending on how heavily you us the system – so make sure you check that out too.

      You can then plug the numbers into our Water Calculator and check that your LSI is in a safe range. If so, and everything tastes good, then you’re golden.

  6. tonycosma

    Good day,
    Can I use this recipe to make my own water but use RO water instead of distilled water?


    • BHLearn

      Hi Tony, we’ve found pure RO (that has not been bypassed) has a TDS of about 11ppm. So yes, it’s a good base water so long as your RO gives you a way to isolate the water coming directly our of the RO membrane. But we certainly advise you to take GH and KH readings of the base water first, before you use it in your machine.

  7. tadeoklapp

    Thanks for this post guys, really useful.
    I’ve tried Rao’s recipe over distilled water. V60 brew time went from6 minutes, almost irrespectively of the grind size used (things barely changed even when grinding much coarser). I’ve searched online extensively for this and didn’t find anything – and pretty sure I got the weights right (using an Acaia lunar). Did you ever experience this? What could be the cause?

  8. ke.cai.1024

    I don’t understand how you obtained the amounts of salts required for 1000 ppm buffer and hardness solutions. In other words, how did you work out 1.68g and 2.45g per liter for baking soda and Epsom salt respectively? Here is my calculation:

    The chemistry/math is simple. Take hardness solution (MgSO4) as an example. The percentage of Mg+ in MgSO4 by weight is the atomic mass of Mg+ (24) divided by the atomic mass of MgSO4 (atomic masses of Mg, S, O being 24, 32, 16 respectively, this comes out to be 24+32+16*4=120). Therefore, the percentage of Mg+ in MgSO4 is 24/120=20%. Then taking any amount of Epsom salt, to work out how many grams of Mg+ there are, we just need to multiply by 20%. Therefore, there is 2.45g * 20% =0.49g of Mg in 2.45 g of Epsom salt. Dissolved in 1 liter of water gives us a solution with Mg at 490 ppm, not 1000 ppm. A similar calculation for the buffer solution also yields a different concentration than 1000 ppm. Am I doing something wrong? I have not touched chemistry in 15 years.

  9. Massimo

    I love this post! I’ve read and re-read it multiple times.

    What should the recipe be to achieve:
    ~95 mg/l Magnesium Sulfate
    ~45 mg/l Calcium Citrate
    ~10 mg/l Sodium Chloride – (so 10 mg/l of chloride)
    ~40 Alkalinity


  10. mattyagnew

    Hi, I would really love to learn to remineralise the water in my local area as it just seems more sustainable for me. Firstly, if remineralising your local tap water do you still need to use de-ionised/distilled water to make up the concentrates? Secondly, I have my water stats here but I’m just not sure what my GH, KH and TDS are as I think we have different names for it in Ireland!

    Water Stats:
    Calcium Hardness (mg/l) 27.9
    Magnesium (mg/l) 0.6
    Potassium (mg/l) 0.3
    Calcium (mg/l) 10.2
    Total Hardness (mg CaCO3/l) 11.2
    Clark English Degrees 2.0
    French Degrees 2.8
    German Degrees 1.6
    NI Hardness Classification Soft

    I would really appreciate any guidance. Thanks in advance!

    • BHLearn

      Hi Matty, thanks for the question.

      Regarding your stats, it looks like your general hardness (GH) is 28ppm (which is 1.6 degrees). My guess is what they’ve called ‘total hardness’ there is actually your KH at 11ppm. ‘Total hardness’ normally means GH, I think this is a mistake.

      I hadn’t realised this before, but if you just use your supply water to make both your concentrates and dilution, you’ll get the desired hardness – plus whatever is already in your water. So if you follow our recipes, but add 28ppm less hardness and 11ppm less buffer when you mix it, you should end up with the results you’re looking for.

      I’d recommend a carbon filter to remove chlorine and other odours though – and it’s a good idea to test your water each time you mix and not rely on municipal testing, since water can be variable, and test kits are not that expensive.

  11. marlon.flechsig

    Hey guys thank you for the updated recipes. I was using the Barista Hustle Recipe all the time. Now I found out about this updated version.

    I‘ve used this water recipe for approximately 3 years with my Bezzera Unica and had no problems with scale/ corrosion so far.

    Will these „new“ versions be safe to use in an espresso machine?

    Thank you on advance!


  12. markpetrig

    Hi All,
    I am hoping someone can shed a little light as I have been getting a lot of conflicting information from various sources.

    I have an RO system with the following filters:

    Stage 1 – Sediment Pre-filter
    Stage 2 – Granular Activated Carbon (GAC) filter
    Stage 3 – Ultrafine TFC RO membrane (300 litres per day 1)
    Stage 4 – Carbon polishing Post-filter

    Does the base water need to be deionised (effectively distilled water) by adding another stage 5 Ion Exchange filter. Before creating the above water recipes?

    Thanks a lot.

    • brygmanden

      Hi Mark. Thanks for posting the question! I bet there are many users out there trying to figure out what is appropriate filtration and what is good enough.

      Although the recipes mentioned in the post are based on ‘pure’ water (0ppm) and using ‘calculator 1’, you can easily mix recipe water with an ‘almost pure’ RO product water. My personal RO system’s product water is around 12ppm, where GH is 12ppm (mostly calcium).

      If you’re able to carry out a diluted titration test for GK and KH, you can figure out what your RO product water contains, too.

      With that number in hand, head back into the post and look for the paragraph ‘Mineralise Pre-existing Water’. You’ll find a calculator called ‘Remineralisation Calculator 2’, which is designed to work with an existing mineral content in your water. Here, in my case, I would be subtracting the amount of calcium I have (12ppm) from the target amount of magnesium.

      If you want to re-create the exact water recipes mentioned in the post, then yes, you need to start with a ‘pure’ water.

      To get your RO system producing pure water, I would recommend you contact a local water filtration expert and get their advice on what type of filtration you need. Unless you distill your water, it is very difficult to produce ‘pure’ water.

  13. cltrejo

    Hello, thank you for the article, can you please let me know if Im doing this right? If I wanted to make Hendon’s recipe for one gallon, is the method correct as followed?

    Hendon Water Recipe as listed in the article for 1 liter
    30.8g Buffer
    99.9g Mg
    869.3g DI Water

    Gallon Recipe
    (1 Gallon = 3.78541 Liters)
    Buffer: 30.8 * 3.78541 = 116.59g
    Mg: 99.9 * 3.78541 = 378.16g
    Water: 3290.66 (this is calculated as 1 Gallon = 3785.41 g/ml, 3785.41(1 Gal) – 116.59 (volume of Buffer) – 378.16 (volume of Mg) = 3290.66 ml of water after displaced volume)

    116.59 + 378.16 + 3290.66 = 3785.41 (1 gal)

  14. bruceh1

    Rather concerned that the 2nd calculator doesn’t work and is outputting incorrect data. It is incorrectly varying the final calculation based on the size of how much of the existing water is added.

    For example, if we input that no hardness or buffer solution is used, and simply vary the amount of existing water to add, it is suggesting that the output ppm of the resulting water is changing, which is clearly incorrect.

    • BHLearn

      Hi Bruce, thanks for the question, good spot. That field (‘Amount of water to add’) was accidentally coloured blue but isn’t actually supposed to be editable – it’s just supposed to tell you how much water to add to make it up to 1L after you add a certain amount of concentrate. If you leave that cell unedited then the calculator should still give you correct results. We’ll update the calculator so that line isn’t blue to make it clear.

      If you want different quantities to 1L, then you’ll need to multiply your recipe accordingly. Hope that helps!

      • eric-4424

        Ran into this same issue…thought the spreadsheet was wrong, but luckily spotted this exchange in the comments. So just a reminder to make that change to the spreadsheet!

  15. tim.scongack

    This article is great thank for you putting this together, as someone new to this I had a few questions:

    For the above Budapest WOC mix do you mix the hardness and buffer solutions or the buffer and epsom salts?

    Once the concentrates are made how many ml do you add to 1 gallons; should you target 150 PPM?

    After creating the buffer solution I had to add more baking soda to get the PPM to 1000; are store bought TDS detectors accurate or should I stick to the concentrations outline above?

  16. robin.carey

    Would scaling these recipes up to multiple liters/gallons be as simple as multiplying the buffer, magnesium, and DI water up to your desired total? Or do the amounts need to be adjusted further for larger quantities of water?

    • BHLearn

      Hey Robin, for sure these recipes scale up and down. Just note that we don’t advise you leave very large quantities of water sitting around in storage for reasons of hygiene.

  17. jamie-2902

    Is it possible to make the hardness solution using calcium sulfate (gypsum) in place of the epsom salt. If so, how many g/L?

    • BHLearn

      Hi Jamie, In cases where you have corrosive water that is very soft, working with Gypsum might help to create a slightly positive LSI index. But there is a major risk of causing too much limescale buildup with this approach so you need to proceed with caution. To get at better understanding of working with LSI, we go into detail about it in the Water Course. We also recommend the Water Geek app for android.

  18. Leonardy Huang

    Hi, thanks for the recipes!

    Quick question, do I need to shake the solution to make sure it is mixed properly, or can I just wait to let it dissolve by itself? Thanks.

    • BHLearn

      Hi Leonardy, Epsom salts and bicarb of soda are readily soluble in water but it’ll speed up the process to give your concentrates a shake when you are preparing them for sure.

  19. imagin4ry

    can any one help me how to measure the kh and gh of existing bottle water/tap water?

  20. fel.blake

    Hey BH, just wandering what”the book” is? It’s referred to in the original post under the “method” section. I’d like to read further about this. Super interesting stuff!

    • fel.blake

      Also, can I keep the “buffer”, “hardness” and a bottle of the BH mixed water in the fridge for any length of time? Or should it always be done fresh everyday.

      • BHLearn

        What about the freezer? You could dose what you need into a little ice tray and add a cube as and when you require.

  21. chris

    Hey, i have a knot in head.
    What should be the final ppm of the crafted water?

    • BHLearn

      Hi Chris – don’t try to measure the ppm with a conductivity meter. These assume a normal ratio of ions in (tap) water, and won’t give you an accurate reading for this recipe.

  22. Buster

    Quick question. Baking soda is made up of sodium and bicarbonate at 27% sodium the rest bicarbonate. So if we have 50g of the buffer in 1 L of water isn’t it only 36.5ppm of bicarb and 13.5ppm of sodium. Is this correct thinking? This also applies to the epsum salt also.

  23. Scott Manley

    Hi Gordon – Deionized water by definition has a pH of near 7 at 25C (77F) initially. When it comes into contact with the atmosphere it will absorb C02 and form carbonic acid, which will lower the pH. It sounds like you are using either distilled water; or RO water that naturally has low pH. Note that measuring pH can be tricky and is highly temporal. Also, the neutrality point of 7 is only true at 25C/77F. Meaning that as water temperature changes, so does the neutrality point. For example, the neutrality point of water at 200F is 6.14 pH. So at 200F, a pH of 6.14 would not be acidic.

  24. Joe

    Hi Ghazi,
    Did you find an answer to this question, was wanting to do the same thing.

    • ghazi

      Hi Joe

      You first need to find out the form of the mineral. For potassium bicarbonate , manufactures never state, after talking with a lot of people it seems that it is crystaline and natural no added water so “anhydrous” . There is maths included in another BH post –
      this nice post will tell you how much to add to 1 ltr DI water for your conecntrate. I have found with the maths included:
      100.115 divided by 50,000 x 1000 = 2g per litre

      Hope this helps


  25. Chris S.

    I’ve begun crafting water for espresso using the above as a guide and, so far, I’m pleased with the results. It truly is amazing how great an impact the water chemistry has on the extraction!

Submit a Comment

Copyright © 2021 Barista Hustle, All Rights Reserved!

You have Successfully Subscribed!