Ward Labs Results: SoCal RO Water

[thefost]

I have very hard water (TDS 500-700) and use this Watts RO filter

For fun I decided to send my RO water to Ward Labs to confirm it's doing a good job. Here's the results:

pH 6.6
Total Dissolved Solids (TDS) Est, ppm 59
Electrical Conductivity, mmho/cm 0.10
Cations / Anions, me/L 0.8 / 0.7
ppm
Sodium, Na 15
Potassium, K < 1
Calcium, Ca 1
Magnesium, Mg < 1
Total Hardness, CaCO3 7
Nitrate, NO3-N 0.2 (SAFE)
Sulfate, SO4-S < 1
Chloride, Cl 20
Carbonate, CO3 < 1.0
Bicarbonate, HCO3 4
Total Alkalinity, CaCO3 3
Total Phosphorus, P 0.01
Total Iron, Fe < 0.01
"<" - Not Detected / Below Detection Limit



So a lot more sodium and chloride than I expected, but everything else looks low. Since this is a pretty popular RO system I thought a few of you might be interested in this.

 

[ajdelange]

Since you are feeding the RO unit with water from a softener (or should be) it isn't that surprising that the sodium is high but the chloride? Perhaps the brine isn't being completely rinsed from the resins.

That's certainly not too much chloride with which to brew, however.

 

[thefost]

I'm pretty sure that there is not a water softener feeding the RO unit. Not that I can do much about it, I'm an apartment dweller.


You're right though, definitely not too much chloride to brew with.

Since you are feeding the RO unit with water from a softener (or should be) it isn't that surprising that the sodium is high but the chloride? Perhaps the brine isn't being completely rinsed from the resins.

That's certainly not too much chloride with which to brew, however.

 

[mabrungard]

Both sodium and chloride have smaller ionic diameters and do pass through membranes at a bit higher rate than divalent ions like Ca and Mg. I am surprised to see them at that level and all the other ions are so low. New machine I guess?

 

[jbaysurfer]

Why "should" we be feeding our RO with softened water AJ? Is that how commercial breweries are doing it? I know of one that does it that way at one of their breweries, but their smaller pilot brewery is just unsoftened RO.

 

[thefost]

Its about a year old now. I took this sample just before changing the sediment and carbon filters, hoping to get a somewhat worst-case scenario.

Both sodium and chloride have smaller ionic diameters and do pass through membranes at a bit higher rate than divalent ions like Ca and Mg. I am surprised to see them at that level and all the other ions are so low. New machine I guess?

 

[ajdelange]

Why "should" we be feeding our RO with softened water AJ?

An RO cartridge membrane is like any other thing exposed to hard, carbonaceous (alkaline) water. Limestone will precipitate on it just as it does on a shower head or faucet. There is an additional twist in an RO system, however, and that is that the concentration of calcium and bicarbonate (and thus carbonate) ions on the 'concentrate' side of the membrane is higher than it is in the feed water so that precipitation is more likely than it would be with the feed water. Removing the calcium and replacing it with sodium means CaCO3 cannot precipitate. Something else (other than calcium carbonate) becomes the 'limiting salt'.

Is that how commercial breweries are doing it?

If their water is hard and carbonaceous and if they are smart.

I know of one that does it that way at one of their breweries, but their smaller pilot brewery is just unsoftened RO.

There you are.

 

[mabrungard]

The saturation limit for scaling ions such as Ca and Mg are often the limiting ions that force the RO machine manufacturer to set the wasting flow rate for the system. For typical home systems, the wasting rate is pre-set with an orifice that is paired to the membrane production rating. That often means that the wasting rate is huge since the manufacturer doesn't know what water quality is being used to feed their unit. For the industrial units that I specify, the wasting rate is user variable and is typically set in conjunction with membrane modeling software and knowledge of the raw water quality. If you don't use softened water, you are probably stuck with the high wasting rate in order to keep your membrane from crusting up prematurely. If you use ion-exchange softened water, then you could improve the water efficiency of your RO system by reducing the wasting rate, but that is a difficult calculation for the average home user.

With regard to why some breweries don't use softening on their raw water feed, they don't need to. I was a consultant to one of the largest brewers in the world and dealt with the nanofiltration system at one of their breweries. The raw water was very hard and brackish. They didn't soften the water. The nanofiltration system did have an acid wash system that was used to descale (clean) the membranes periodically. That provided the long life they needed and avoided creating another wastewater stream from the ion-exchange softener.

 

[ajdelange]

The saturation limit for scaling ions such as Ca and Mg are often the limiting ions that force the RO machine manufacturer to set the wasting flow rate for the system.

Precipitation thresholds are determined by an ion product. If [Ca++][CO3--]> Ks where Ks is a constant precipitation will occur.

For typical home systems, the wasting rate is pre-set with an orifice that is paired to the membrane production rating.

As an example the GE units that are sold by Home Depot have a recovery rate of 18% when used with the pressure tank. This goes up to 20% if the pressure tank is not used.

If you use ion-exchange softened water, then you could improve the water efficiency of your RO system by reducing the wasting rate, but that is a difficult calculation for the average home user.

It's actually a pretty simple calculation - easily incorporated into an Excel spreadsheet. For example, if you have water with calcium hardness equal to alkalinity then they can both be at about 125 ppm as CaCO3 before calcium carbonate becomes the limiting salt at a recovery rate of 18%. That is the upper limit at which one of the example GE units should be run with carbonaceous water. For gypseous water to the extent that all the hardness is permanent a much higher level of calcium is tolerable - hardness of 358 ppm as CaCO3 and sulfate at 342 as SO4. Clearly here the limiting salt is calcium sulfate. For any particular water chemistry the calculation has to be done for all ion combinations including, in systems with high recovery (which can be above 80% if trans membrane pressure is high enough and if some of the permeate is fed back into the feed stream) some bizarre ones like barium sulfate, silica, and calcium fluoride. Home brewers would not typically be operating at such high recoveries but a small craft brewery using a 500 GPD or larger RO system might. Home brewers do, however, use techniques, like elimination of the pressure tank and installation of a boost pump, to get higher recovery rates with the fixed orifice systems. This means softer feed water. Other tricks including reduction of the feed pH (shifts carbonate ion to bicarbonate) and the inclusion of precipitation inhibitors are used. And finally the membranes can be 'cleaned' by flushing with acid. I suppose even the small under the sink units used by home brewers could be cleaned in this way. Vinegar should do the job. In any case it is best to feed with softened water.

With regard to why some breweries don't use softening on their raw water feed, they don't need to. I was a consultant to one of the largest brewers in the world and dealt with the nanofiltration system at one of their breweries. The raw water was very hard and brackish. They didn't soften the water. The nanofiltration system did have an acid wash system that was used to descale (clean) the membranes periodically. That provided the long life they needed and avoided creating another wastewater stream from the ion-exchange softener.[/QUOTE]