Miyamizu – Heavenly Water – The Gold Standard?

This post describes the Heavenly water, Miyamizu.

Water is the main ingredient in all sake but it usually gets the least attention. Despite getting the least attention, water is important and does play a huge role in the quality of sake. The story that is told to demonstrate this fact is told so often that it has become like a legend.

The legend (no, the real story): Back near the end of the Edo period, 1840, Yamamura Tazaemon owned two breweries. One in Nishinomiya and the other in Uozaki. Tazaemon-san noticed that the sake made at Nishinomiya was always better than that made at Uosaki.

His two breweries were part of the Nada Go-go region or the five sake-brewing towns of Nada. The five districts lie in a line on the coast running west to east: Mishi, Mikage and Uozaki lie in Kobe while Nishinomiya and Imazu lie in Nishinomiya.1 The Nada Go-go region made its fame shipping sake to Edo (Tokyo) by ship, a 20 day voyage. The five districts of Nada produced a little more than 25% of Japan’s sake in 2003. But I digress.

The Nada go-go, i.e., the five sake-brewing districts of Nada
The five brewing districts of Nada, i.e., the Nada Go-go.

Tazaemon-san struggled to figure out why the sake at Nishinomiya was always better. He ensured that they used the same rice, milled to the same level, the same conditions and equipment; nothing helped the Uozaki brewery to produce sake that was as good as that of Nishinomiya. Tazaemon-san even switched the brew masters but not even this helped. One day, Tazaemon-san, had the water used at Nishinomiya shipped to his Uozaki brewery. Sake was brewed using this water and it was as good as the sake made at Nishinomiya. Tazaemon-san had solved the mystery, it was the water!

The take away from this story is that good water is important if you want to make good sake. But what is “good” water for sake? Those in the Nada region would say that it is a hard water that matches the heavenly water, miyamizu. Mineral rich waters produce sakes with quick strong fermentations and are strongly flavored with full bodies. The mineral content of the miyamizu water is:

Miyamizu (ppm)
Potassium 20
Phosphoric acid 5.2
Magnesium 5.6
Calcium 37
Chlorine 32
Sodium 32

This water contains the right combination of minerals that help the yeast to work vigorously. Notice that there is no iron listed. It is important that there is no iron in any water to be used for sake.

While the brewers in the Nada Go-go region had hard water to work with, much of Japan has softer water. In particular the brewers of Hiroshima had very soft water to work with. The Miura-Toji, in 1896, explained the steps needed to produce excellent sake with Hiroshima’s soft water. He suggested a smaller moto be used with less koji than is normal. The koji is made at a slightly higher temperature to bring our more conversion strength.  I don’t have more details at this time. Miura’s paper is “An Account of New Brewing Techniques,” published in Japanese. In 1905 the Hiroshima sake, using Miura-Toji’s techniques, took both first and second place in the national sake assessment. So… it seems clear that soft water, without chemical additions can also be “good” water for sake; you have to use the correct techniques though.

If soft water is used, the sake will tend to have a clean and bright but softer taste that melts in your mouth, taking with it the flavors and aromas of the sake. This is in contrast to hard water which produces a strongly flavored sake.

The recipe that most sake home brewers use has been derived from Fred Eckhardt’s work2 and is designed to work with soft water augmented with additional salts to make the water miyamizu-like.

An example of soft water is that of the Bull Run reservoir:

Bull Run water (ppm)
Potassium 0.2
Phosphorus 0.007
Magnesium 0.9
Calcium 1.9
Chlorine Added 2
Sodium 3.5

You can see that the parts per million here are much lower than those of the miyamizu water. Chlorine is added to the Bull Run water to disinfect the water.

Is Miyamizu the Gold Standard? Can I say, Yes But… Miyamizu is clearly a good, nay, excellent water for brewing sake. However, as demonstrated by the brewers of Hiroshima, their soft water can impart seductively sexy characteristics to sake that are not possible with Miyamizu.

  1. https://web.archive.org/web/20120211075120/http://www.phontron.com/en/nada/about.php Nada map
  2. Beginning with “Sake (U.S.A.)” © 1992 and followed by a serices of recipe versions. His latest version being version 6.2
Click to Enter a Comment

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

12 thoughts on “Miyamizu – Heavenly Water – The Gold Standard?”

  1. Thanks Will,
    I learned to make koji earlier this year and tried my hand at makgeolli as well, so sake brewing was an obvious next step.

    Reading your recipe inspired me to try the full process (half-size batch) rather than take shortcuts to produce a doboroku style beverage, and I really feel the time invested will be worth it.

    I’m on day one of the moromi. The SG is spot on and the aroma is amazing, so I may well be hooked.
    Cheers, Mike

  2. Great information! Using the excellent calculator from the Khymos blog, I was able to get a close match to the mineral balance you quoted.
    My recipe looks like this, starting with soft water (the calculator will balance for any particular water source):
    2.6g Table salt (NaCl), 2.2g Baking soda (NaHCO3), 0.7g Milk of magnesia [Mg(OH)2], 4.5g Chalk (CaCO3).

    Use 200mg of this salt mixture per litre of water.

    I plan on using this with my first batch of sake.

  3. Thanks Will.

    I started with R.O. and calculated what salts to add to get as close as I could to Miyamizu for 2 gallons of water. I then added these to the first 2.5 cups of water for the Moto. The thought process is that I will need up with the right concentrations for he finished product, but by starting out more concentrated and then adding only pure R.O. water, the Moto phase will have all the nutrients needed. I can also play around with flavor impact in the final product (easy to add; impossible to subtract).

    Pretty steep learning curve right now but fun. Lots to read and enjoy.

    I plan to see if I can get a few different styles out of the batch. I will take some out after pressing the lees for Nigorizake. And then try to split the remainder for a bottle of Namazake and the rest Genshu.

  4. Great information about water quality and suitability for brewing sake! My tap water is very hard and highly chlorinated. We have to use bottled water for drinking as the taste is so unpleasant. So I plan to use R.O. water and build up from there like I do with beer.

    I did find Morton salt substitute (MSS) but I have some questions on its use.

    First, I wanted to get an idea of the ingredients and concentrations.

    I was able to find this reference:

    A further sodium-free “Salt Substitute Composition” has been disclosed and claimed in U.S. Pat. No. 3,505,083 by H. C. Miller. The composition is comprised of about 80-99% by weight of potassium chloride and from about 1-20% by weight of fumaric acid. The bitter after-taste of potassium chloride is allegedly overcome by the fumaric acid component of the composition. A commercial version of this sodium-free composition is marketed by Morton Salt Company as “Morton Salt Substitute.”

    Source – https://www.google.as/patents/US4473595

    I have not been able to find anything on the amount of Tricalcium phosphate or monocalcium phosphate in Morton salt substitute so cannot guess how much Morton Salt Substitute is needed to get 5 ppm P in the water.

    Given that, and about 2 gallons of water used in the recipe, then to achieve 20 ppm of Potassium, I would need about 39 mg of KCl per liter (Since 52% of the weight is due to K and 48% due to Cl). For the 2 gallons (7.6 l) total need of water that would be about 0.3 grams total (assuming 99% KCl in Morton’s salt substitute). Did I make an error here?

    The recipe calls for 6 g. Is that calculated to get the K or P levels up where needed? And that amount of MSS will add about 380 ppm Chloride.

    Likewise, I would need .39 g of Epsom salts to get to 5 ppm Mg in 2 gallons (7.6 l) of water. The recipe calls for 0.7 g.

    Should I prepare water ahead of time specifically for brewing?

    Or can I add all the minerals for the whole 2 gallon batch when starting the moto and then just add R.O. water for all subsequent water additions?

    Looking forward to trying this out!

  5. Hello Will,

    I am just starting my second sake homebrew, and I just recently purchased your book, to show my support for the great work that you have done on this site. The brew is going well so far, but while getting ready for the next step I was thumbing through the book again, and was led back to this post to double check: Can this Miyamizu water, which is spring water, really have chlorine in it? This has got to be a mistake in that source-Chlorine is rarely found in nature. Also wouldn’t that chlorine (if it is chlorine) inhibit yeast growth and metabolism. Is it more likely that this is a chloride, not Chlorine? And if so, what chloride? Anyway, thanks for the book and the videos.They are a great resource and inspiration.

    1. Dom,

      I don’t recall where I originally got my values but I have been asked about it several times and when I look around I find new sources and they generally agree. They aren’t the same exact numbers but they are close. We shouldn’t expect the concentrations to remain the same over time. Here are the data from this site and two other sources:


      Springer: Microchimica Acta 1961, Volume 49, Issue 2, pp 217-226 A co-precipitation and spectrophotometric determination of micro-quantities of iron in Japanese brewing water and liquor by Kiyoteru Otozai, Kunihiko Mizumoto

      Site: The values given on this site and in the book Brewing Sake.

      Akiyama: Sake, The Essence of 2000 Years of Japanese Wisdon gained from Brewing Alcoholic Beverages from Rice by Kiroichi Akiyama

      As for the question of Chlorine vs. Chloride, the symbol used is Cl. I have translated Cl as Chlorine.



    1. Joe,

      I don’t recall where I sourced the data in the article. I have seen it in several places; they are usually pretty much in agreement but not 100%. One source you might look at is: Sake: The Essence of 2000 Years of Japanese Wisdom gained from Brewing Alcoholic Beverages from Rice by Hiroichi Akiyama.

  6. Thanks for the great info on water. I never realized what a difference it could make until my wife and I took a tour of the Suntory Yamazaki Distillery in Kobe. They have their own spring there, and a lot of the locals will drive out of their way to get ice cubes for their whiskey and water from there. It is amazing and makes a big difference in their finished product.

    1. Thanks fishruss!

      Suntory Yamazaki Distillery being in Kobe is in the same area as Miyamizu but its water must be quite different; a soft water. Did you try it?

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.