Thanks for that tip, man. I had been looking at the ones in the big box stores, but they don't include the DI step that the reefkeeper ones do that nets in near-0 ppm TDS. Never thought to look at what th reefkeepers use. I usually want to start with a blank slate on my water for many beers and have been in the market for a water system. After seeing the systems they have I am getting a killer system (Typhoon III) for the price of a RO-only system locally.

I like it because you can bypass the DI step for filtered drinking water, etc. This is important for the wife approval factor for bottled water, and making coffee. Then we can use the RO/DI water for brewing, ironing, humidifier for when the kid has a cold, etc.

I really don't think softened water will "reduce the life" of your RO; the issue I 'know of' is this: the rejection ratio of sodium ion is lower than that of Ca and Mg. This means the following...you are taking high Ca and Mg water for example, replacing it with high amounts of Na. Now you take an RO membrane and reject a lower proportion of the sodium...what are you doing? Instead of leaving some calcium and mag, you are leaving more salt which is really not a good thing for beer, more so with high SO4.

I would personally stay away from cheap units. The way RO works is by passing water over a membrane, good clean water goes through and the crap is rinsed by waste water...read: WASTE. A cheap RO may take 15 gallons of water to make 1 gallon of relatively clean water. A high end unit may be more like 5:1 waste to final. If you spend a bit more in advance (like with a watts unit for closer to $400) you will make better water and likely make up the costs in the lifetime of the unit. But the important part to me is better water, not so much the cost overall. Now, if you like crappier water with high Na then go for it.

Another concern, paper vs. thin film membrane. Paper membranes (from my limited knowledge) require chlorinated input water to keep the cellulose based membrane sanitary. TFC should have a carbon filter before the membrane since chlorine will break down the membrane. I would talk to a water professional (Kinetico or Culligan) and get a water analysis before you buy anything. Worst case, you can DIY and buy either a high end unit (Watts) or a shitty unit (eBay/Home Depot). Excuse my water arrogance but, honestly, beer is largely water and I don't like to lose control of the majority of my beer to a cop out of a unit.

Where is AJ to put me in my place and set things straight?


Now if anyone is feeding well water to an RO you really need to have a handle on your iron and crap like that before you throw it at any membrane. Crap in crap out.

Add on a permeate pump and a pressure tank and you'll never run out of water.

My kinetico unit puts out about 15-20tds water, 90-95% rejection. This kind of performance should suffice for most brewing IMO. I can pull about 10 gallons of water (3 gal storage tank) in 4 hours. I think they told me it could do about 30-50gpd but my tank is the limiting factor. How much do the reef units run? The only real issue I have is that I have whole house softening. I am probably replacing my 200-300ppm of hardness with Na; if I am only rejecting 90% it will leave 20-30ppm of salt.

RO systems are quite diverse but the units sold for home use at places like Home Depot seem adequate for production of small quantities of brewing water. I have two of the GE units and they seem to do a fine job on my well but then my well isn't that much of a challenge. With respect to softened water feed: hard water is not good for RO systems. As part of the way they work a gradient in concentration of solutes is created perpendicular to the membrane surface with high solute concentration closest to it. If the concentration of calcium carbonate gets high enough close to the surface it will precipitate onto the surface and clog the membrane. For this reason most industrial systems either inject acid into the source in order to lower pH, thus converting carbonate to bicarbonate and bicarbonate to CO2 (which passes through the membrane), or feed the system with softened water. It is true that the rejection for sodium isn't as good as it is for calcium but it's still over 90% in a typical system so there's a trade space. I have total hardness of about 2 mEq/L which, if replaced by sodium, would give 46 mg/L plus the 7 already in the water for a total of 53. Reduced 90% that would leave 5.3 which isn't that much different from what is in there now but the calcium and magnesium would be way down. Thus I do not feed my RO systems with softened water and haven't had a fouled membrane in several years of operating these things (the GE units form Home Despot). They are spec'd for a maximum of 171 ppm as CaCO3 hardness at pH 6.9. As noted above the probability of fouling will go up with hardness and pH so if you have 300 mg/L hardness and pH 8.3 you probably would want to feed through a softener. The softener would have to add 5 mEq/L or 115 mg/L sodium to replace the calcium and magnesium representing 300 mg/L hardness which at 90% rejection would leave 30 mg/L in the permeate (plus 10% of what was already in the water) which is quite a bit but 300 mg/L hardness implies a challenging water however you slice it.

Chlorine is definitely a problem for either the acetate or TFC membranes so a carbon pre-filter is an absolute requirement if you are on municipal water. Most units have such a filter but it is interesting to note that the GE unit is spec'd for up to 2.0 mg/L chlorine whereas the county water around here runs closer to 3.0 so if I were on county water I would require and additional carbon pre-filter.

As to recovery (the fraction of feed water that exits as permeate) is appreciably better than it used to be. The GE units I've been describing have a recovery of 16.7% if you don't have the storage tank hooked up and leave them on line to pressurize that. This is 6:1 as compared to 10:1 or more a few years back but it isn't great. Fancier systems have better recovery (3:1 or even better). Note that hooking up to a pressure tank seems to be quite costly in terms of efficiency. It drops to 6.7% for the GE units.

I built an RO cart out of some scrap wood and used it to house a small GE system. It's very handy and plenty effective for brewing purposes. The times I've had my water tested, it reflects about 88-90% rejection. The main problem with my municipal water is the bicarbonate level, which is around 225ppm. After passing through my RO system, it's usually down in the 20-30ppm range. I've also found that after a "break in" period, the system is able to produce about 1gal/hr - higher than the spec in the manual.

If I were to buy a new unit, one that I would want to live with for the next 10-20 years, I would go with a quick change system like this: http://www.watts.com/pro/_productsFull. ... 5596&ref=2