Towards a better tinyhouse

Inventing to freedom?

Reverse osmosis water recycler thing or part of recycler system

leave a comment »

Okay. So some good news today! I looked into the possibility of building a reverso osmosis unit that can filter greywater to some degree. Reverse osmosis is good stuff. Looks like it is totes doable for $350-$700, depending how much you shop around for a good deal on the parts (and it seems to vary greatly) for the 70 liters a day needed. Increase that, but less than linearly, for more water, when you upsize the pump and the cost does not go up that fast, a 2x size membrane and pressure vessel is again not twice the cost.

Actually, you might be able to do this for even less with under the tap RO stuff. Tap water is apparently 500 ppm so that shouldn’t be a surprise, but it’s probably not going to get you a very high recovery rate (unless you use several RO modules) and would be less energy efficient. Just use a bunch of the membranes, and actually using a CA membrane might be a little better because they can withstand chlorine, so that could help reduce growth of microbes. That I think would be a substantial concern, bacteria could grow on the surface of the RO membrane, thusly clogging it. Apparently an “osmotic backflush” can help to some degree with this sort of thing, though. You just turn the pump off for 15 minutes, and water goes from the clean side, backwards to the dirty side, lifting a bunch of dirt off in the process. Scaling might or might not be a problem, that would require further investigating.

Anyway, it goes like this:

Seawater salinity: 35,000 ppm, osmotic pressure: 27 atm (394 psi) (from the wikipedia article on osmosis)
Greywater: about 1000 PPM (from my notes) osmotic pressure: 0.77 atm (11 psi) (might be higher in a situation with water saving shower heads etc. but then lower again if you don’t put food or whatever down the drain).

If recovery rate is 50% then the salinity at exit is 2x so the average pressure diff across membrane is that much lower.
Example membrane:
GPD: 150 gpd or Gallons Per Day
PSI: 800 PSI
Stabilized Rejection: 99.4 Rejection

Example Pressure vessel, surely you can get a much better price we only need 120 psi capable, which even an under the counter unit residential unit can handle, but it gives you an idea of what I’m talking about, I started looking into this under the assumption that I would need high pressure stuff, that is why I used this :

Okay, so that membrane only gets 150 gpd at a recovery rate (dirty water going in/clean water out) of 2%, which is just the test condition, not the expected operating condition. If you operate it at 50% level, basically think of this as a whole lot of RO elements in series, the dirty water goes through one, get a bit concentrated, then through the next one and is a bit more concentrated, and then the next one… anyway, by the time it gets out, it’s salinity (and therefore osmotic pressure, assuming the salt involved is the same salt, which if course it is) is twice as high.

Remember the rate of water flow across the membrane is proportional to the input pressure minus the osmotic pressure difference on the clean and dirty sides (I assume it’s 0 on the clean side). So obviously if the input pressure was 800 psi, and seawater is 394 psi you would never really be able to practically get a 50% recovery rate.

With 800 psi input with greywater though it is not a problem, going from 11 psi to 22 psi isn’t going to affect the flow rate much. But 800 psi pumps are relatively expensive (uh, I think, well pressure washers are pretty cheap, so maybe you could find a high pressure pump, would probably have to be a piston pump, that would open up interesting options for high recovery rates)

Also, the pressure drop across the membrane is much higher to begin with, so you would get that much more flow through it.

So to find out what sort of flow rate you would get with input osmotic pressure of A and output osmotic pressure of B and pump pressure of pp, and a flow constant of L (the flow rate decays exponentially with time as concentration goes up), so the initial flow rate is (pp-A)L, when is done flow rate is (pp-B)L so you would find what t is for those points and find integral between them and then divide by time difference to find average flow. But what is L, well at 394 psi input and 402 output and 800 input pump pressure (so (pp-a)=407 and (pp-b)=398) got flow of 150 gpd so just suppose L 150/400 gpd about or 0.375 gpd/psi.

So for us, at a pp of 120 and A of 11 and B of 22 (pp-A)=109 (pp-b)=98 or ~37.5 gpd. So if you only wanted 17.5 gpd you could get more than 50% recovery with this setup, actually maybe 80% or more.

An example pump (one issue is that most pumps are made for much much higher flow rates):

You still need to filter the water to get the particles out though, or they stick to the RO membrane (mind you I wonder if you could get around that with osmotic backwashing so you didn’t even need to particle filter it, but the particles might also physically clog the membrane element, as I don’t think they are made to pass particles very well, maybe you could find one that is, maybe a flat plate module).

This could be a great way to boost a rainwater collection system, greatly reducing the amount of rain water needed, or could be used in combination with an MBR (which provides low maintenance particle filtering and disposes of the wastewater that comes out of the RO unit). There are some things that can pass through an RO membrane, urea is one, but it gets probably almost all of the stuff, and urea isn’t harmful especially at these levels, and very improbable that anything else in the water is, I think. Obviously you can’t flush toxic stuff down on purpose but if you’re just showering, washing dishes, doing laundry, whatever went down the drain is probably something that you put on your skin in much larger quantities at some point, and probably already ingest some of (including laundry detergent, there is always some left on the clothes). Mobile Condo people, do you read me?

There is one thing that nags at me, which is that you have to really wonder why this is not done with those water recycling systems for sale in Australia. I have however read of it being done in cruise ships, and also that Marco Cremona guy in Malta used RO. Maybe it is maintenance issues or something, it seems like the main unknown here is the potential for membrane cloggage, but I think it is definitely worth trying. The Malta guy was a pro in RO system design, incidentally.


Written by gregor

November 20, 2010 at 19:20

Posted in Uncategorized

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: