RO Machine Problem or Brun Water Problem?

[TheHopfather]

For the last couple of batches I've brewed, at least the ones I've bothered checking the pH on, I've found that the pH estimates provided by Brun Water have been off, my pH is ending up higher than predicted. For example here is what I've done with my water for an IPA I brewed last week;

5.30 kg (11.7 lbs) - 2 Row
0.45 kg (1.0 lbs) - Munich
0.40 kg (0.9 lbs) - Wheat Malt
0.20 kg (0.4 lbs) - Crystal 40

20.5L (5.4gal) RO mash water
3.9g gypsum, 1.2g calcium chloride, 0.2g canning salt
3.3mL lactic acid

Ca - 67 | Na - 12 | SO4 - 107 | Cl - 48
Est mash pH - 5.20

My measured pH was actually 5.58 pH. This was a room temp sample taken 15 min into the mash with a freshly calibrated meter which was reading the calibration solutions spot on. I am currently using Brun Water v4.1.

I decided to work backwards in Brun Water to see if I could figure out what was going on. The only way I can make my measured pH jive with what Brun Water was putting out is if I adjust the RO water values to have a bicarbonate value of 150 ppm versus the default of 16 ppm. If this is actually correct I am going to have to conclude that my local RO machine is not working correctly. I'm posting this as a sanity check, based on the numbers posted above should 3.3mL of 88% lactic acid get me down near 5.2 with actual RO water?

 

[bleme]

Did you try testing pure RO water?

 

[ajdelange]

Some of the people who operate RO machines add minerals back into the water so that it doesn't have that flat taste that pure and nearly pure water have.

You can
1)Ask the operator of the machine if it is adding any minerals after RO filtration
2)Get a cheapie TDS meter from any of hundreds of sources on the net and check the water approximate TDS
3)Get an estimate of the alkalinity of the water with an alkalinity test kit or by finding out how much acid it takes to lower the pH of a gallon to 4.5 or so. It should take only a tiny amount.

You can't expect Bru'n water (nor any other spreadsheet or calculator) to give you a mash pH prediction that is spot on but then the error shouldn't be as big as 0.3 pH either.

 

[mabrungard]

Having a TDS meter is imperative when using RO water. That is the only way to assure that the water is reasonably demineralized. More than likely, the RO source is not as pure as expected.

 

[Silver_Is_Money]

If my RO water measures roughly 50 TDS is that a concern? The well that feeds it (via a salt and resin bed type softener inbetween) measures about 820 TDS. What ballpark might my RO's baseline alkalinity and RA be playing in?

My RO's pH is 6.08.

 

[madscientist451]

Just curious, how does the beer actually taste?

 

[Silver_Is_Money]

Just curious, how does the beer actually taste?

Not sure if this was directed to the OP or myself. Don't want to be changing the direction of someone else's thread. That said:

I have my first batch using a combination of my well water (25%, acidified to kill high alkalinity) and my newly available home RO water unit (75%) sitting in the fermenter and clearing presently. Mashed at pH = 5.33 (meter verified). Sparge water acidified to pH 5.50 (meter verified).

I was a routine home brewer until 19 years ago when I moved to a new home and job, lost interest and stopped brewing. I recently retired, and now with time on my hands again, I have rekindled my brewing interest. This is my first batch in 19 years.

 

[ajdelange]

If my RO water measures roughly 50 TDS is that a concern? The well that feeds it (via a salt and resin bed type softener inbetween) measures about 820 TDS.

Your rejection is approximately 1 - 50/820 = 94%. That's not too bad but with a new system at least we might hope for a few percent more i.e. 97 - 99% with the corresponding TDS at or below 25. If this is an old system it might be time to think about replacing the the membrane cartridge.

What ballpark might my RO's baseline alkalinity and RA be playing in?

That's impossible to tell. The number you get from the meter is only accurate if everything in the water is NaCl. Ignoring that for the moment and supposing that you had 820 mg/L NaHCO3 the alkalinity would be 14 mEq/L (700 ppm as CaCO3) and the RA (which isn't really important) the same. Were there 820 mg/L Ca(HCO3)2 then the alkalinity would be 10.1 mEq/L (500 ppm as CaCO3) and the RA about 8.7 mEq/L. Were the entire 820 mg/L made up of calcium and sulfate then the alkalinity would be 0 and the RA negative. These are some extreme cases so obviously what you have will lie somewhere within these bounds.

My RO's pH is 6.08.

That's normal for RO water.

 

[Silver_Is_Money]

It's new, but it is a budget RO unit. I ran a series of GH/KH (liquid drops) tests on the well water and when running it through the Kaiser Water spreadsheet the average of my results indicate:

Total Hardness = 756.4
Ca = 212.4
Mg = 55.2
Alkalinity = 436.1
Bicarb = 532
RA = 252

I tried using this same test kit on the RO water, but its resolution is not good and I can't pin down definitive results. Can I make a wild ballpark first assumption that the RO has about 6% of the well water, or:

Total Hardness = 45.4
Ca = 12.7
Mg = 3.3
Alkalinity = 26.2
Bicarb = 32.0
RA = 15.1

 

[JonM]

So, it's probably not a problem with either Bru'n water or with the RO system. It could be that the grains you got had a different Lovibond than that which was listed, or (an error I've certainly made in the past) maybe you entered in the wrong Lovibond values or made another data entry error.

 

[ajdelange]

Can I make a wild ballpark first assumption that the RO has about 6% of the well water?

That's what I'd do.

 

[Silver_Is_Money]

That's what I'd do.

Thank you!

 

[Silver_Is_Money]

Having a TDS meter is imperative when using RO water. That is the only way to assure that the water is reasonably demineralized. More than likely, the RO source is not as pure as expected.

What should typical RO water read on a TDS meter?

I don't have a TDS meter, but I do have a Hanna meter that reads conductivity in units of MicroSiemens (uS/Cm). I just got a reading of 116 for our homes RO unit using this meter. I had initially assumed that TDS is roughly half of uS/Cm, but some literature I've seen places TDS at 0.64 x uS/Cm, and some says TDS = 0.71 x uS/Cm, and other sources say different correlation factors from these, so it seems that they do not necessarily have a linear correlation, and since everyone has a different answer, who knows what factor is correct. I guess I'll just have to shell out the $'s to acquire a TDS meter so I can pin down my TDS and determine how to correlate this to my uS/Cm instrument.

 

[ajdelange]

No, don't do that. Conductivity is fine. A TDS meter measures conductivity and then multiplies by a constant suited to the intended use of the meter, usually NaCl.

Just check your water pre and post RO. The post RO reading should be about less than 5% of the pre.

 

[Silver_Is_Money]

No, don't do that. Conductivity is fine. A TDS meter measures conductivity and then multiplies by a constant suited to the intended use of the meter, usually NaCl.

Just check your water pre and post RO. The post RO reading should be about less than 5% of the pre.

The well water reads 1,280 u/S/Cm.

 

[Silver_Is_Money]

On my Hanna meter a solution of concentration 48 mg/L NaCl reads as 100 uS/Cm. Would you know what reading a TDS meter gives for a solution with that same concentration of NaCl dissolved in DI water?


Edit: I contacted Hanna directly, and the correct factor for my unit is 0.50, so my reading of 116 uS/Cm converts to 58 TDS.

 

[ajdelange]

Hach's 1000 uS/cm sodium chloride conductivity standard contains 491 mg/L NaCl so a solution of 48 mg/L NaCl should read 97.8 uS/cm and the factor for conversion between conductivity and TDS for sodium chloride is 0.491 (pretty close to what they told you).

The fact that 116*0.491 = 56.956 does not mean that your RO water has a TDS of 57 mg/L. It only means that a solution of sodium chloride with a conductivity of 116 would contain 57 mg/L NaCl.

Given that the water entering the RO units has conductivity of 1280 uS/cm we would hope that 95% of all the ions that contribute to conduction would be removed reducing the conductivity to 5% or less of what it is going into the RO unit which in this case would be 64 uS/cm. Clearly your rejection is closer to 100(1 - (116/1280)) = 91% than 95%.

 

[mabrungard]

While the NaCl solution is the typical standard, the 442 Natural Water standard may be more applicable for our low TDS waters. However, the meter has to be set up to use that standard. Not all meters can do that.

 

[ajdelange]

Sure they can. Again referring to the Hach calibrating solution it is 3000 mg/L TDS and has conductivity of 3960 uS/cm. The factor is thus 3000/3960 = 0.757576 and he computes the TDS of his 116 uS/cm water as 116*0.7575 = 87.9 instead of 57 mg/L using the NaCl factor. Do we think 40% Na2SO4, 40% NaHCO3 and 20% NaCl represents his actual water better than 100% NaCl? No calcium in either of the formulations. Do we care? What is important for checking the rejection of the RO unit is the ratio of the before and after measurements in whatever units.

Suppose the meter only reads TDS. It doesn't matter what the factor is as we only want the ratio and the factor cancels out. If we want the 422 TDS as opposed to the NaCl and we know the factor our meter uses (and we hope the manufacturer of any meter, no matter which part of China it came from, will tell us what that is) then we divide TDS by that factor (converting back to conductivity units) and then multiply by 0.7575 to get 422 TDS units.

 

[Buckeye_Hydro]

I agree with you both.

Not all TDS meters, right out of the box, have a select-able conversion factor. Most retail level meters are NaCl. For instance, our COM100 has a select-able conversion, but our AP1, or TDS3 do not.

And

What we're interested in here is the ratio of permeate TDS or conductivity, to feedwater TDS or conductivity.

Russ

 

[dmr]

For the last couple of batches I've brewed, at least the ones I've bothered checking the pH on, I've found that the pH estimates provided by Brun Water have been off, my pH is ending up higher than predicted. For example here is what I've done with my water for an IPA I brewed last week;

5.30 kg (11.7 lbs) - 2 Row
0.45 kg (1.0 lbs) - Munich
0.40 kg (0.9 lbs) - Wheat Malt
0.20 kg (0.4 lbs) - Crystal 40

20.5L (5.4gal) RO mash water
3.9g gypsum, 1.2g calcium chloride, 0.2g canning salt
3.3mL lactic acid

Ca - 67 | Na - 12 | SO4 - 107 | Cl - 48
Est mash pH - 5.20

My measured pH was actually 5.58 pH. This was a room temp sample taken 15 min into the mash with a freshly calibrated meter which was reading the calibration solutions spot on. I am currently using Brun Water v4.1.

I decided to work backwards in Brun Water to see if I could figure out what was going on. The only way I can make my measured pH jive with what Brun Water was putting out is if I adjust the RO water values to have a bicarbonate value of 150 ppm versus the default of 16 ppm. If this is actually correct I am going to have to conclude that my local RO machine is not working correctly. I'm posting this as a sanity check, based on the numbers posted above should 3.3mL of 88% lactic acid get me down near 5.2 with actual RO water?

No one mash pH calculator is going to be correct for every grist/salt/acid combination. I ran your data through both Brewer's Friend and MpH. These two calculators estimate mash pH values of 5.43 and 5.48, respectively. Cheers!

 

[TheHopfather]

Thanks for all the insight on this problem everyone. I ended up contacting the RO machine company, they opened up a maintenance ticket and serviced the machine. I brewed up a pilsner using the "new" water and came much closer to my predicted pH. I was shooting for 5.30 in Brun Water and ended up with 5.40, I can live with that.

Only problem with that is my efficiency went through the roof as a result. I was aiming for an SG of 1.048 and ended up with a hydro reading of 1.058, whoops. At least now I feel pretty confident that my RO water was the issue and can go from here.

Thanks again.