reverse osmosis

[leonsmith]

Hello all after having my tank set up, getting all the inside looking ok and starting to stock fish, thought the next thing i should do is is invest in r o. I would like peoples comments on the best ones out there on the market for quality, price etc. i have a 450 ltr tank

[MatsP]

All RO units sold for hobby use are essentially the same thing - there are some minor differences. And when it comes to what's INSIDE the canisters, it's even more of "it's all the same", as there are only a small number of companies that make the thin-film membrane that is doing the actual reverse osmosis. The only main difference between different kits would be additional extras, such as pressure meters, flush taps, and various other things.

We really should have an article covering RO units, how they work and what they do - some time I will get round to doing that.

A basic RO system consists of three main components:
1. A sediment filter - this takes out particles that are in the incoming water, that would clog up the RO membrane.
2. A carbon block filter. The primary purpose of this is to remove chlorine, but it also has the side-effect of removing various other compounds. Chlorine will destroy the RO membrane, so it's well worth the effort to replace the carbon block filter on a regular basis.
3. The RO membrane itself. The RO membrane uses pressure difference to "filter" pure water out of the dirty water - in simple terms, the water is capable of passing throught the tiny holes in the membrane, but other compounds can not. This only works if there is enough pressure to push the water through those tiny holes. As a general rule, an RO unit removes about 95% of ions in the water before it reaches the membrane.

Some units (AKA 4-stage RO units) have an extra stage of De-Ionization resin, which is an ion-exchange resin. The resin is charged with hydrogen and hydroxide ions, which are "swapping place" with the ions from the water. Since Hydrogen and hydroxide ions combine into water, the ions are replaced with water, essentially. By binding the swapped ions harder than the hydrogen and hydroxide ions, the resin doesn't allow the "captured" ions back into the cleaner water. For 99% RO-using of freshwater fishkeepers, a DI stage is completely unnecessary. And if you think you believe you are one of those 1%, you are probably wrong! ;) The only reason I can see as to using a DI resin is if there is some sort if ion-pollutant in your tap-water. But if you are on "waterboard" water, then you should be fine.

All RO units "lie" about production volume... The produced water volume by the RO unit as a whole relies on pretty high pressure - more than MOST people have in their home, and it's also calculated at unusually high water temperature - something like 25'C, which is unlikely that your tap-water actually is. You also need at least 60psi/4 bar water pressure, which is possible if you are on relatively new/good water supply.

--
Mats

[RickE]

I've got an 'RO-man' unit which as Mats says is just about the same as everything else on the market. It's fine, but I do like the look of this:

http://www.new-era-aquaculture.com/images/ro-unit.jpg

[leonsmith]

Okey thanks for that mat very useful as always, the one i was going to get is made by a company called d d as thats the one my local shop stocks also they do the jbl but the filters that go inside are nearly twice as expensive and with what you say about regular carbon changes i think the dd is the better option

[MatsP]

The D D ones are supposedly good if we trust my LFS - I haven't seen one, nor used one. Mine is from RO-man.
[edited]It seems like RO-man's website is working again, but it says "closed for renovations", so I'm not sure what is going on there...[/edited]

--
Mats

[PeterUK]

closed for renovations

The site has been closed for renovations for months, I have tried to ring them a few times at different times of the day but no answer. I have noticed that RO-Man hasnt been advertising lately either also have heard on a few different forums that RO-Man has gone bust which might explain things.

I have a RO question also.

Lets say that the waste water from a RO unit is 75% (figure taken a random)

Does this mean that the 75% waste water has the original 100% of the impurities ?
IE: The waste water has MORE of the impurities than the original tap water.

Would the waste water need to be run through a HMA filter before mixing with RO to get the PH needed or am I making things more complicated than needed ?

[MatsP]

Yes, the impurities of the "input" water comes out in the waste-water. So, if we assume the RO membrane is "perfect" and thus 100% efficient on removing unwanted material [this is not entirely true, but it makes relatively little difference and makes the math MUCH simpler to follow], and we have a 30 ppm nitrate [nitrate removal is the primary reason for my RO system]. Let's further assume that we get 75% waste water, and 25% "production" (or "pure") water. The resulting waste water will contain all the nitrate from the "input" water. So that's 30 parts per million the whole 100% of output water, pure and waste. Since the 25% pure water is free of nitrate [in our assumption of a "perfect" RO membrane], the 75% waste water will contain 30 / 0.75 ppm nitrate -> 40 ppm nitrate. The same of course holds true about any other compound, wanted or unwanted, in the input water, e.g. calcium, carbonate, pesticides.

The only exception is anything adsorbed by the carbon block and collected sediment/particle filters - once that is adsorbed/collected, it is stuck there, and won't be released [at least if these do what the label says, which we assume they do, at least to a large extent].

So what effect has our assumption of a perfect RO membrane? In reality, a small proportion of whatever is in the input water will pass through the RO membrane - in my experience and particular case 5-10% of the TDS value comes out in the "pure" water - how much depends a bit on "newness" of the membrane, as well as temperature and probably pressure. So, if we make it easy again, and assume that 10% of everything is let through, we will have 90% of 30 / 0.75 in the waste water, or 36 ppm nitrate in the waste water. Not a huge difference - so it's "safe" to count on a 100% RO membrane and not loose any sleep over it - unless you are working on something that requires more precision than 10% give or take - in which case, a measurement of the value of the actual offending compound is the best approach.


The HMA filter is a carbon block filter with a special formulation to remove heavy metals [scientific facts: activated carbon adsorbs different substances depending on the size of the cavities in the structure of the carbon clumps itself - by modifying the structure, the manufacturer can decide what compounds will be highly absorbed, and which won't be]. Since the RO system has a generic carbon filter (not specifically targetting heavy metals, but capable of adsorbing those to some extent), it will remove some heavy metals. If you have problems with high levels of lead, chrome, copper or some such, then I'd say you should run it through a HMA filter. I personally don't use the waste-water in this way at all - I use the waste-water as a continuous water-change on a group of fishes that "don't bother too much about water", and since the set of tanks is 4 x 100 liter, and my RO system produces about 600 liters a week, I reckon I get about 2000 liter of waste-water a week - that's changing water 500% per week on those fish. So there is never any huge buildup of waste...

--
Mats

[Burks]

I've used the waste water from my RO unit for less demanding fish quite often, especially when I was running it like crazy for my coral propagation system. After that, I haven't used it much.

My shrimp bred like crazy in the waste water. Cleaner than it came in at least (although possibly higher in other aspects, so I figured it evened out).

[PeterUK]

Thanks, thats given me something to think about.

We really should have an article covering RO units, how they work and what they do - some time I will get round to doing that.

I second THAT ! ! !

Can we expect it sometime next week then ?