Help with RO water analysis

[Peebee]

Not really an essential subject (many brewers just assume RO water is devoid of dissolved ions), but it would help me if I could get a better alkalinity analysis for RO water.

I'm only interested in municipal source supplies, not exotic private bore hole sources. So only calcium, magnesium, sodium, chloride, and sulphate. Alkalinity is calculated, "as bicarbonate" or "mEq/l". Anything else is ignored as being too trivial in quantity (Nitrate and Potassium might not be, but that's handled in other ways).

Currently the options are; everything zero (including alkalinity) or a "contrived" analysis of Ca=1ppm, Mg=0, Na=8, Cl=1, SO4=4, and alkalinity ("as bicarbonate") as 16ppm. If I put that last "analysis" in a spreadsheet:

Alkalinity calculates to a slightly higher figure, but 1.5 ppm (as bicarbonate) is neither here nor there. "97%" is the percentage of ions filtered out: Possibly too high for a decant "home" filter, but on the low side for the more substantial commercial/industrial units. The orange ion concentrations are the source (unfiltered), the value below the resultant RO water concentrations (milligrams per litre or parts-per-million). (The percentage can be dropped to 70% ... well, some people don't change their filters too often!).

The question! The 97% is applied to all ions; they are all treated the same. But I've a suspicion that RO membranes are a little more biased; they filter out some ions more easily than others? And the biases might be significant? What say you?

Thanks

 

[hottpeper13]

I don't have a meter so don't have a clue as to how many gallons go thru my RO system, so I use a TDS meter and change out the filters when it gets around 20 tds. Before I got the TDS meter, if 4 oz of acidulated malt didn't get my Pilsner to 5.2 pH, I'd change the filters.

 

[MaxStout]

I use a tds meter, as well. A good $15 investment to check the health of my filters. If the rejection rate drops below about 90% it's time to replace.

 

[Peebee]

Thanks! Only three of you, I had imagined this subject kicking off a more vocal debate? But (personally) this query is being very educational (I have no need to use "RO" water so my knowledge concerning it was a bit thin).

I haven't been sitting on my hands hoping someone will come up with all my answers. Seems the home systems are not automatically less "efficient" ("what gets filtered"-wise) compared to commercial/industrial systems, just considerably slower (size). And I'm not so sure commercial/industrial systems produce less reject water now? I dug this up, it's from Reverse Osmosis problems removed table - www.uk-water-filters.co.uk, but I don't trust stuff from people trying to sell me things. Still, the table is from an (unknown) third party (summarised below):

Based on just this table (insufficient to draw trustworthy conclusions from) "97%" rejection is way too high ... 92% seems a better top figure? The "RO" people say higher, but they are including everything, and as homebrewers we're only really interested in ionic content ... hence, I'm suggesting 92% rejection tops. Of the dimmed-out ones above: Nitrates and Potassium were only included because they are commonly found in drinking water ... beer-making doesn't need 'em. Phosphate and Silicate (I've hardly come across silicate) are a source of alkalinity, but beer can do without them. Chlorine and Chloramine sharply reduced ... no tears to shed for them, or for the total removal of bacteria/viruses.

The four remaining we'll want, and will be added to creating a water profile for beer making. No Calcium and Magnesium? Apparently, those element trash RO filter membranes: So, this RO filter retailer (UK WaterFilters) suggests a "water softener" before the RO Filter, but doesn't build one in. Odd behaviour. Are they normally built-in, or do people know they are to be in-place?

"Water Softeners" seems daft; they are the opposite to what beer needs, removing Calcium and replacing it with Sodium! But the RO system will be removing the Sodium anyway; it's just a temporary fix to keep Calcium away from the RO membrane. Assuming sodium chloride salt is used for ion exchange ... the potassium salt might be used. There are "salt-free" water softeners, but they are really a misnomer and are truly water "conditioners" using other means (including electrical and magnetic) to "rearrange" the molecules (the calcium and magnesium remains where it is). I'm not at all sure these "water conditioners" are of any value at saving RO membranes from Calcium?

 

[Buckeye_Hydro]

I don't have a meter so don't have a clue as to how many gallons go thru my RO system, so I use a TDS meter and change out the filters when it gets around 20 tds. Before I got the TDS meter, if 4 oz of acidulated malt didn't get my Pilsner to 5.2 pH, I'd change the filters.

Remember that a TDS meter tells you absolutely nothing about when to change your prefilters (prefilters are all the filters that touch the water before the water reaches the RO membrane).

From the FAQ's on our website:
A good rule of thumb is to replace your sediment filter and carbon block after six months. A more precise way to maximize the usable life of these two filters is to use a pressure gauge to identify when pressure reaching the membrane starts to decline. This is your indication one or more of the prefilters (all the filters that touch the water before it reaches the RO membrane) is beginning to clog.

Also be cognizant of the chlorine capacity of the carbon block. A good 0.5 micron carbon block for example will remove much of the chlorine from 20,000 gallons of tap water presented at 1 gpm. Some original equipment suppliers commonly provide carbon cartridges rated at only 2,000 to 6,000 gallons. Remember that all the water you process, both "waste water" and purified water, goes through the carbon block.

Regarding your RO membrane, use your total dissolved solids (TDS) meter to measure, record, and track the TDS (expressed in parts per million [ppm]) in thwo places: 1) tap water, 2) after the RO.

The TDS in your tap water will likely range from about 50 ppm to upwards of 1000 ppm. Common readings are 100 to 400 ppm. So for sake of discussion, let's say your tap water reads 400 ppm. That means that for every million parts of water, you have 400 parts of dissolved solids. How do we go about getting that TDS reading down to somewhere near zero?

If you do some experimenting with your TDS meter, you'll note that your sediment filter and carbon block do very little to remove dissolved solids. So with your tap water at 400 ppm, you can measure the water at the “in” port on your RO membrane housing and you'll see it is still approximately 400 ppm.

The RO membrane is really the workhorse of the system. It removes most of the TDS, some membranes to a greater extent than others. For instance, 100 gpd Filmtec membranes have a rejection rate of 98% (i.e., they reject 98% of the dissolved solids in the feed water). So the purified water coming from your 100 gpd membrane would be about 8 ppm (a 98% reduction). The lifespan of an RO membrane is dependent upon how much water you run through it, and how “dirty” the water is. Membranes can function well for a year, two years, or more. To test the membrane, measure the TDS in the water coming into the membrane, and in the purified water (permeate) produced by the membrane. Compare that to the membrane’s advertised rejection rate, and to the same reading you recorded when the membrane was new. Membranes also commonly produce purified water more slowly as their function declines.

Additionally, don’t forget to sanitize the entire system at least once per year, and wash and lube your housing o-rings with food-grade silicone grease every filter change.

Russ

 

[Peebee]

Diagram from the OP corrected to comply with lower rejection rate. Remember, I'm applying a RO rejection rate of ionic compounds only, not total solutes which we (as brewers) are not so interested in. And the rejection rates do differ for different elements/compounds, but not enough to deviate from a flat rate (for the small handful of compounds we're interested in), which is an answer to my original question.

The example "RO Water" I'm using is the same as before (as recorded in Martin Brungard's "Bru'n Water" v.5.5). I guess the high (relatively!) Sodium suggests ion-exchange "water softening" prior to RO treatment is being illustrated?

This is using data from that "unknown" source as published by "UK WaterFilters", a commercial company so I have to treat the data as possibly suspect? But it does add up rather nicely! This is suggesting an "optimum installation" rejection rate of the chosen compounds is 92% (it can be set less than this, and up to 95%).

Orange quantities are source water, below these are grey, bold, figures for the resultant RO Water.

(Note: I haven't let on what that snippet of spreadsheet is from yet ... "work-in-progress". As it's me, that really means "pharting-about-in-progress").

 

[Peebee]

Hmmm ... no more comments? I did say I'd created an "answer to my original question", but I'd be surprised if it was the "right" one.

I couldn't use @Buckeye_Hydro's figure of 98% rejection rate because it was too general and really only relevant to a specific product, whereas the information I'd gleaned could be targeted for rejection rates and (claimed?) to be "typical". Hence, I'm sticking with 92% for now: 98% does seem particularly good though . Does your (@Buckeye_Hydro) company Website have similar information tucked away? I couldn't find any.


I'm very new to this "RO Water" game and have no need of it here at home (water analysis: Ca:22, Mg:1, Na:7 Cl:12 SO4:45, HCO3:9). And I'm struggling with the unintuitive nature of "RO" (semi-permeable) membranes. It's all at a "molecular" level; "rejection" figures get lower with decreasing efficiency (i.e. more passes through the membrane ... is that due to microscopic holes and rips?) whereas I'm more used to filters getting clogged up and letting less pass through.