RO throughput

[seangee]

Its been 8 months since I bought a 50GPD unit. Initially this was taking around an hour to produce 10 litres of water (or about 6 min/L). This has gradually got slower and had dropped down to 10 min/L. I wasn't too concerned because all the research I have done suggests the flow is slower when the water is colder. The water quality (as TDS) has actually improved over time. Initially it was 14 and has recently been down to 5, so I did not suspect the membrane was clogged.

Last night I swapped the membrane for a 100GPD version. What a difference! My TDS has gone back up to 14 (tap is over 300) but I am getting 3 min/L. I am not searching for 0TDS or 100% purity. My tap water has 50ppm nitrates and is very hard. My RO water has 0 nitrates, 0KH and 0GH as you would expect. I did not swap out the pre and post filters as I wanted to compare like for like.

As a bonus I have purchased a second membrane housing and will connect this in series (to re-filter the waste water). In theory this should double the throughput again - and reduce the waste.

Hmmm maybe I can consider another tank now. At around 200 litres per week it was taking too much time to replenish - but if I am able to get 4x the throughput ...

 

[AbbeysDad]

Sounds good! I still use my DIY Tap Water Filter (filled with API Nitra-Zorb) to filter out nitrates. I follow-up by using an inline carbon filter after the tap water filter (that I change every 2000g). Happily it just keeps chugging away with salt water recharges about every 200g. The same resin has yielded over 8500g so far!
I ruled out RO as I'd need an additional pump for high enough psi to force water through the membrane as my well system is on a 25-50 controller. Then there's the 4g waste water for every 1g RO and adding back minerals.
Still....glad it's working (and working better) for you.

Note: I'm not sure how effective it will be to re-filter the waste water since that water wouldn't go through the first membrane But, "nothing ventured, nothing gained". Often our greatest successes in the hobby is the result of trial and error...if even in the face of any internet 'nay sayers'.

(I am continuing to experiment with anoxic biocenosis clarification baskets having just put a 40g sump/refugium on my 110g stock tank.)

 

[seangee]

Note: I'm not sure how effective it will be to re-filter the waste water since that water wouldn't go through the first membrane But, "nothing ventured, nothing gained".

Indeed. Have seen a few good reports of this. The intent of the second filter is to refilter the clean water but you just swap the inlet. I <blush> have mislaid the wrench so had to order a new one and just put the new housing in place of the old one. The design of the add on is really simple. There is a pressure valve on the waste outlet so when enough pressure builds up it pushes water through the membrane and expels any surplus. Assuming the original design is the same it seems feasible that a lot of the expelled water has not actually tried to go through the membrane.

I did spot an ad for a flow through (tankless) 400GPD system that claims it produces 1G waste for every 3G of RO produced. It also has a reverse flush feature which (in theory) should extend the usable life of the membrane. Sadly it costs about 4x as much as my system and I have just bought 2x100GPD membranes, filter cartridges and the additional housing. Maybe next time around . In the meantime if I get 6 months out of what I have it will work out cheaper than buying RO for 2 weeks (ignoring the cost of water).

 

[seangee]

Hmm. Weather is a bit inclement this morning (even the labrador refuses to go out in it) so I spent a bit of time looking at the flow through units (direct flow in marketing speak). It seems they all adopt the same principle I suggested in post #1, typically using 2 or 3 high capacity membranes. For the extra money you seem to get a bit more instrumentation and an electronic processor. My flow control is all mechanical (or electro-mechanical in the case of the booster pump & solenoid).

Hopefully for all that extra cash they have actually designed the systems to balance / match flow rates to give optimum performance rather than just bolting on an extra housing as I will be doing, but I won't hold my breath on that

@AbbeysDad, you may be right about the reduced filtration efficiency, almost all the ones I have seen include a DI stage at the end. I'll update in a week or 2 once I have bolted on the new housing. In the meantime it was a pleasure getting my final 30 litres done in 90 minutes yesterday compared to the 5 hours that would have taken a week ago.

 

[seangee]

So far so good. Just plumbed in the second membrane and I am now down to 2 minutes / litre. Its currently flushing through but the TDS is already down to 20 which I suspect is good enough.

So I improved my throughput by an additional 50% of clean water - presumably at the cost of the same amount of water going in as it is the waste being pushed throgh the second membrane. Plumbing was a bit trial and error as the two filters work in different ways. One has a flow restrictor on the waste and the other on the clean water. Tomorrow I will try swapping the order around as I expect that may give an improvement as it will effectively increase the pressure on the second pass. Time will tell. If not I am still getting 30 liters per hour compared to the original 10 for very little additional outlay - and I am wasting less water.

 

[seangee]

Final round of tube swapping complete. As I hoped I did get a slight improvement but as I suspected not as much as is theoretically possible. The improvement is about 10% so now I get around 33 litres per hour with no discernable drop in water quality. SInce the input is pumped I don't believe this has anything to do with time of day or input pressure.

 

[seangee]

In case you wonder why I am making such a big deal of this...
This morning was a late start and I did a water change of 120 litres on the community tank. I have just finished replacing the water I used.
Late yesterday afternoon I changed the 2 smaller tanks (90 litres total) and replenished the water before I went to bed last night.

That is my usual water change routine. Normally I am able to finish filling my final jerry can before leaving for work on Monday morning, as long as I don't go out on the weekend. Since I typically do go out on weekends its Tue or Wed before I have replaced all my water, only to start again on Friday.

 

[seangee]

Well its been a while...
Thought I had updated this thread to say I had abandoned the project.
The way I had it was working perfectly while the water was running. But because I was using a tank with a shut off valve the different control mechs caused a problem. Basically once the tank was full all the water was shunted down the waste channel so I was producing RO 24/7 and sending almost 100% of it to waste. Probably could have found a solution ...

Anyway it seems my initial thinking was on the right track
https://www.directwaterfilters.co.u...direct-flow-reverse-osmosis-pumped-system-LCD
I preferred this one of the ones they have (some with higher throughput and some cheaper) because having 3 membranes means that only 1/3 of the water goes to waste according to their blurb. Much better than the 3/4 I pump to waste at present. And at 1.5 litres per minute I can fill my 10 litre jerry cans in around 7 minutes, compared to the 40 minutes they take now, and get back the space used by the storage tank.

 

[mbsqw1d]

What are your thoughts on using rainwater to supplement the water source?

 

[seangee]

I would if I could. Besides the practicality of collecting and storing enough in a suburban environment with a low rainfall ( it rains frequently but there is not that much of it), I live midway between the busiest and third busiest airports in the world. I haven't tested it - but I am not sure that I would want to drink that rainwater without filtering it

 

[seangee]

Changed membranes today even though the claimed life is 18-24 months, as well as the filter cartridges. There was no increase in TDS, I simply changed because it came up in my calendar. I last changed the cartridges in September. I have seen a noticeable drop in flow over Winter which is expected. This time I used a 150GPD membrane (compared to 100GPD).

Since the flow is fairly constant from 1 week to the next I use a timer. Today I simply took usual time / 3 x 2. When the timer went off the water was at the same mark I usually fill it to (in Winter I leave space for boiling water to get the temp right.) So the flow rate does appear to improve linearly with the membranes rated flow rate and the water quality is identical. My system is pumped - this may not hold true in an unpumped systen if there is insufficient pressure.

200GPD membrane on order for next year . That''s as far as I can go without changing the cannister. The cannister itself is cheap but 400GPD membranes currently cost twice as much as 200GPD - and at 100GPD it was fast enough for me in summer but a bit frustrating in WInter when it takes 20% longer to produce the same amount of water. Now its coming out faster than in summer so I am happy - and next years improvement will be a bonus.