Water Profile for Brewing in Mountain View, California

Introduction

Some homebrewers go to considerable lengths to fine tune the mineral content of their water. I am not an expert in either brewing or water chemistry and will not express an opinion for now on whether that effort is worthwhile since I have yet to do any experiments of my own, brewing the same beer with and without water modifications. My concern is that if the mineral content of water coming out of the tap changes day to day, that invalidates any precise one-off lab analysis I get done. I originally performed this experiment when living in Colorado. I lived between the two reservoirs that served the city and as the local water authority advised me, I saw big changes from day to day as my supply was drawn from different processing plants. A write-up of those results may be found at Water Profile for Brewing in Boulder, Colorado.

Now I am in Mountain View, California, I repeat the analysis here. The city water authority provides a map of which areas of the city are fed from different sources and shows that my water is bought from the San Francisco Public Utilities Commission (SFPUC). That is fed primarily from Hetch Hetchy reservoir in Yosemite but does also use small amounts of locally collected run-off. The expectation is therefore that my water should generally be very soft, coming from snow melt and a granite watershed. Water occasionally drawn from local reservoirs is potentially much harder. The SFPUC and local Mountain View annual reports contain the same numbers and match expectation. The average hardness and alkalinity are fairly low, but the total range of variation through the year is large. Table 1 shows numbers from the 2016 and 2017 annual reports, the latest available at the time of writing, but numbers are consistent over the preceding five years.

Compound	Year	Minimum	Maximum	Average
Alkalinity (total as CaCO₃)	2016 2017	7 6	112 131	39 52
Hardness (total as CaCO₃)	2016 2017	8 7	76 82	44 51
Calcium (as Ca)	2016 2017	2 2	18 31	10 16
Magnesium	2016 2017	0.2 0.2	6 11	3.6 6.2
Sulphate	2016 2017	1 0.9	30 34	16 17
Chloride	2016 2017	<3 <3	16 17	8.8 9
Sodium	2016 2017	2.6 2.3	17 31	11 18
Potassium	2016 2017	0.2 0.2	1.0 2	0.6 1
Total dissolved solids	2016 2017	<20 <20	95 122	63 76
pH	2016 2017	8.2 7.4	9.8 9.8	9.4 9.2

Table 1. pH and concentrations in mg/l or parts per million for most the ions commonly cited as influencing beer taste and quality as reported by SFPUC 2016, 2017 reports. Some of the ions that a reputed to affect beer flavours differ by almost a factor of ten between the minimum and maximum.

Measuring Hardness & Alkalinity

I use the same methods as were described and validated in Water Profile for Brewing in Boulder, Colorado. Briefly, this is total hardness and alkalinity drop-count tests using the "API GH & KH Test Kit" intended for fresh water aquaria. To improve the test resolution on soft water I typically test 20–35ml of water per sample rather than the manufacturer's suggested 5ml. pH measurements are obtained with an Oakton EcoTetr pH 2 (not the new updated pH 2+) calibrated using Biopharm Inc. pH 4, 7 and 10 buffers immediately before each measurement. The EcoTestr is fairly stable as long as it is kept in electrode storage solution (not just water) so there is no real need to recalibreate every time.

Results

Figure 1 shows the measurements of my tap water over the first year in the new house. Seeing just these data it would seem reasonable to suggest the outliers indicate the test is unreliable, but repeated measurements of the known water samples show it is actually very reliable.

Figure 1. Results from my Mountain View, CA tap water. Overlaid horizontal lines mark the published minimum and maximum values from the local water company's annual report. Note how hardness and alkalinity measurements tend to cluster near either limit rather than fully populating the range between. Error bars on Hardness and Alkalinity are 1 drop on the drop count test. So, for example, if the drop count test showed the transition on the 9th drop, the data point is plotted for 8.5 and the error bars stretch between the values corresponding to 8 and 9 drops. The manufacturer's specification for the pH meter is ±0.1 and repeated tests of known reference buffer solutions suggest it achieves that accuracy.

Conclusion

Unsurprisingly my results confirm the report published by the water supplier. (I am sure their tests are more accurate and thorough than mine!) Qualitatively, the day-to-day variation follows the water company's description that my water is normally soft but on any specific day might be very much harder if it happens to be being drawn off one of the local reservoirs instead of Hetch Hetchy. There is another important point here that my measurements frequently cluster close to either the supplier's "minimum" or "maximum" values. This is what you would expect for water being alternately drawn from two sources. The "average" is defined by the relative frequency that each source is used and never representative of the "typical" supply. Using the average value from the annual report may not be the best estimate.

Exactly as I previously saw in Colorado, the tap water ion content is changing by a factor of five from day to day so fine tuning water profiles seems a pointless exercise without testing the actual water used on brew day. Depending on your local supply, a one-off laboratory test may not be enough to tell you about your long term supply. I do not know how typical these variations are, but they have existed at both addresses I have tested.

2022-04-28 3:02 PM

If you have comments or suggestions feel free to contact me: