hodorskib's Small Scale RO Build

[carls47807]

Great info hags. Indeed, the membranes are getting worked in parallel.

It is important to remember that a 400gpd tfw membrane is only rated at 25% capacity at 32f. A 400gpd membrane should flow 100gpd of water, or .07gpm. 3 should give .21gpm. If you produce 12 gallons of water per hour that comes out to .2 gallons per minute which gives you 100% recovery.

Membrane fouling from sugar solutions has been found to be easily removed by frequent permeate rinses. It isn't ionic, and doesn't stick to the membranes very well.

The aquatec 8855 pairs well with 2 400gpd membranes (it lies somewhere between a 300 and 400gpd pump).

Carl

[hags]

That's a great question. I just called Aquatec today and spoke with one of their techs about the 8800 series. He said they don't spec the in feed vacuum or suction capability but should have around 6 feet of head. I will be doing a gravity feed system and want the filter before for two reasons, first most of the 5u filters are not designed for 125-160psi, and secondly I didn't want to place a strainer in front of the pump to protect it. His advised testing it to see if it would work. Has anyone else here tested with the 8851/8852?

I would agree the 5u filter should not be subject to 125 psi and before the booster pump is probably better. Gravity feed should create some feed pressure, depending on the setup, or a cheap low pressure inlet pump could be added. Most of the booster pumps have an inlet pressure max. of 30 psi. Flojet makes a 1GPM, 25 psi water pump for RVs. Since the pressurized system is only producing 12 GPH output your low pressure feed pump would always stay ahead.

hags

[eustis22]

now my hairs hurt

[carls47807]

An aquatec 8800 will easily self prime through a 5 micron filter. When it doesn't you'll know it needs to be changed.

The 125psi really isn't a problem on the filter. Unless it is entirely plugged the filter is actually seeing very little pressure differential (125 in and 125 out). They can handle a 80psi pressure differential, and if you had that you wouldn't have enough pressure to produce any permeate so you would have replaced it long ago.

The filters are wound to have inlet pressure, not suction. As they get plugged they squeeze tighter to decrease the probability of particles getting through. If you suck through them they probably wouldn't perform the same and you could lose your 5 micron absolute filtering capabilities. Plugged filters collapse under suction but hold their shape under pressure.

[hodorskib]

hags - you are correct on everything you said. I do push the membranes harder than is recommended. My family and my job commitments allow me only a few hours a day to collect and boil so I try and make the most of them. Over the past 6 years running the system the membranes can last between 2-3 years with flushing and washing on a regular basis. With that said you can purchase the membranes for under 25 dollars each which means 150 dollars every two years to save 75-80% on my boiling time and fuel costs is well worth it for me. I am running two setups this year to see how they compare one:
6 - 150gpd membranes in series as I have for the past two years
the second is:
3 - 400gpd membranes in series

After today I will have processed over 1,000 gallons of sap (averaging 1.2% sugar) to 7.5-8% and have found that they are about equal. I also test the permeate with a digital brix meter during each run and it always registers .01 or .00. The 6 membrane system seems to be doing a little better though. After each run I flush with permeate for about 10 minutes change the water filter and then do a warm wash cycle using RO soap (ph 11 - which is what the membranes are rated for) for about an hour. Then I flush with permeate for at least another hour (testing with a ph meter to make sure the ph level is the same as the permeate). This has kept both systems running as the first day I turned them on. At the end of the season I will will do a wash cycle and then a cycle with citric acid and flush for several hours then store them in a pvc holder with the preservative. If the weather holds I should process between 2,500 and 3,000 gallons of sap through them this year.

The concept behind this system is to provide an economical way for the backyard producer to use and benefit from an RO that they can put together themselves I hope this information has been helpful.

[hags]

Hey hodorskib,

Thanks for the reply and you have done an awesome job! Way above the call of duty to bring all this to the community and answer questions and in testing new ideas. This is the part of homemade equipment and maple syrup processing that I like most. I've run some new numbers and I think I see why the (3) 400 GPD membranes are possibly fouling. Now that I have the info. on your (6) 150 GPD setup I'll probably go back and run the numbers on that too. Using the information from Carl and re-thinking the process I can see why your original system has worked so well.

Thanks so much for sticking with it for 6 years! hags

[hags]

Here's a chart to decipher hodorskib. It looks like your current setup using (5) 150 GPD membranes is most efficient and easiest on the membranes at the 6:1 reduction you're currently running. At super cold sap you're a bit short of membranes if your using 6, but as the sap warms up, 4 to 5 membranes is plenty.
Note what happens when using the 400 GPD membranes and why you may be having issues with that setup.

I'm not sure why the chart is so small, but it's bigger if you click on it.

Aqua Tec 8855 6-1.jpg

[jdircksen]

Great info hags. Indeed, the membranes are getting worked in parallel.

It is important to remember that a 400gpd tfw membrane is only rated at 25% capacity at 32f. A 400gpd membrane should flow 100gpd of water, or .07gpm. 3 should give .21gpm. If you produce 12 gallons of water per hour that comes out to .2 gallons per minute which gives you 100% recovery.

Membrane fouling from sugar solutions has been found to be easily removed by frequent permeate rinses. It isn't ionic, and doesn't stick to the membranes very well.

The aquatec 8855 pairs well with 2 400gpd membranes (it lies somewhere between a 300 and 400gpd pump).

Carl

How would I know if my membranes are fouled?
I made a bad batch of syrup yesterday and I think possibly my RO is to blame. We had a warm spell and the trees stopped flowing sap for about 10 days. I ran the RO once during those 10 days to flush it. I thought my season was over but the sap started flowing again. Yesterday's batch went through the RO and while boiling it smelled off. It tastes weird too. I don't know how to describe it.

Today I'm boiling again except not using the RO. So far during the boil I don't smell anything funky. It seems pretty normal to me.

So....how do I clean my RO? I did notice yesterday that when I ran it, it stayed at 60psi for the first 5-10 gallons of sap. Then it went up to 100psi which is where it usually is.

[jwith]

6 - 150gpd membranes in series as I have for the past two years
the second is:
3 - 400gpd membranes in series

hodorskib, Thanks for all the helpful information, developmentand testing of different rigs. After all the feedback I decided to go with that 5-150 GPD in series over the 400GPD series or parallel. After looking over hags numbers and running my own I think for the pump I have it is much better suited for 150GPD. It will also operate closer to the membrane performance specifications. Plus, I'd prefer to run the system longer and have a higher concentration, than to run it fast and have to cycle it multiple times.

Here's a chart to decipher hodorskib. It looks like your current setup using (5) 150 GPD membranes is most efficient and easiest on the membranes at the 6:1 reduction you're currently running. At super cold sap you're a bit short of membranes if your using 6, but as the sap warms up, 4 to 5 membranes is plenty.
Note what happens when using the 400 GPD membranes and why you may be having issues with that setup.

hags, thanks for your analysis. This was helpful in sizing my system.

I'm sure you were just doing a first order analysis, but it appears you are assuming equal permeate, which assumes equal pressure and equal TDS over all membranes. I would think the TDS would go up as you go further down the series causing the osmotic pressure to rise and therefore cause different recovery over each membrane. Maybe this is not noticeable, I am no expert on RO systems by any means, just interested in the math and science behind it. Overall I realize a lot of these numbers are theoretical, especially since most of RO membranes are spec'd for NaCl not sugars. It would be interesting to calculate the actual recovery of each membrane.

[carls47807]

Hey jdircksen,

Was your syrup stringy?

Could be from nasties in your ro, or from old sap sitting around.

Was it a butterschothy flavor? It could have been diacetyl or metabolism off flavor. This can be a big problem when first introducing RO's. Because your sugar concentration is higher, if you don't start making larger batches the off flavor and smelly aromatics don't have time to reduce.