Hobby RO Pump Suggestion

[vordus]

I am in the process of setting up a "Hobby" RO and plan to use 2 or 3 400gpd membranes. I was wondering what pumps you guys are running. I called Aquatec and talked about the 8852-2P01-T421 160 PSI, 3/8 JG, 24 VAC. The Sales person suggested possibly going with a 5858-2LM2-B534 130psi. Any help would be appreciated.

[BartzFarm]

I know it's been nearly 6 years since this post but I was surprised that this was all that came up in my search for the 5858-2LM2-B534. I just built my first RO with this pump and was very pleased with it. I ran it at 110psi pulling 2.8amps through two 400gpd membrane. I was getting 50/50 concentrate to permeate with the concentration flowing at 4 gallons per/hour. The pump stayed at 100 Deg F. and barely broke out into a sweet.

[DRoseum]

Coronwater TYP-8900K if you want to go with more than 2 x 400 GPD membranes.

[fireant911]

Hey Dan,
My apologies for always ‘piggybacking’ on others questions to ask another question BUT it is somewhat related. In regards to the Coronwater TYP-8900K pump, can 600 gpd RO membranes be effectively employed with this pump? I saw your 2023 build incorporated five 500 gpd RO membranes in series - have you attempted anything with 600 gpd day membranes with that pump?

[BartzFarm]

Coronwater TYP-8900K if you want to go with more than 2 x 400 GPD membranes.

Thanks Dan, I'm happy to see you respond to my newbie post. I've been studying and working on your design for over a month. I attempted to purchase the Coronwater TYP-8900K twice, thru Amazon and pwangroup.net, but got scammed both times. Yes, I know I should have purchased it thru AliExpress but didn't have time to wait for delivery. Sap started flowing very early in Ohio! This is only my 3rd year of tapping. Last year I built a two pan & barrel evaporator. This year I built your RO system with only two membranes; my budget is very limited. Next year I'll expand the RO but now I'm wondering if this pump can handle it.

Before I start putting more money into this pump, like upgrading the transformer, adding a temperature limit switch and installing a fuse, perhaps you could teach me why the Coronowater 8900K is that much better? If I were to add more membranes wouldn’t that increase my output and cause less stress on the pump? Also, forgive my ignorance, but on your diagram, where does the 2.6 lpm come from? Oh, and if an 800 GPD pump is better then why not get a 1200 GPD pump for the same price? Like the Germany Seiko Global 1200 (Shenzhen Huashun Baier Electric Co., Ltd.)

On another note, perhaps someone can tell me if it’s ok to leave a 1% hydrogen peroxide in my system between runs, once/week and wait to flush before use. Again, I have a limited budget and can get plenty of H2O2 for free. I’m just confused about the cleaning process between runs.

Thanks for your response and any advice is greatly appreciated!

[DRoseum]

Hey Dan,
My apologies for always ‘piggybacking’ on others questions to ask another question BUT it is somewhat related. In regards to the Coronwater TYP-8900K pump, can 600 gpd RO membranes be effectively employed with this pump? I saw your 2023 build incorporated five 500 gpd RO membranes in series - have you attempted anything with 600 gpd day membranes with that pump?

Yes I am using a 600 gpd membrane in each of my units with 5 membranes in series. 600, then 4 x 500. Works great.

[DRoseum]

I am not familiar with the pump you mentioned or all of its technical specifications. Will need to do some research on it. The coronwater 8900 is one of the highest flowrate diaphragm booster pumps that has a continuous duty cycle.

As for BOTH the aquatec 5858 (or any 5800 series), and the 5300 series, the problem is the heat rise. You cannot run these pumps for more than 60 or 70 minutes at 100 psi or higher without them shutting off due to thermal limits.

The coronwater 8900k can run indefinitely without overheating or hitting a temp limit.

Just get an adequate transformer for the pump, something rated for 350 - 400W on 24 VDC.

You have to size overall membrane capacity and pump flow proportionally. If you increase pump flow only, you won't push any more permeate out (that rate is fixed based on membrane rating at a given operating pressure). But you will reduce recovery rate, which will limit membrane fouling. Recovery rate is defined as a % where permeate out / flow in at each membrane and should be kept as low as possible. These membranes do well around 15 - 25% max.

If you just increase membrane capacity without increasing flow, you are going to INCREASE recovery rate in the membranes towards the end in series. That will make them foul up and not perform as well.

I have a video that explains this on YouTube along with many other videos on how to flush, clean and store the systems.

https://youtube.com/playlist?list=PL6CwcmcTEWYxhYPQCMAWETg9gHWL7_RXw&si=XiAlGwwLBwh0jcUb

Between runs, no need for peroxide. Just flush with new filter and permeate with needle valve open. Use that new (now rinsed with permeate filter) for your next sap batch.

That 2.6 lpm is the flowrate of the pump at 100 - 110 psi. This is equivalent to approx 40 gallons per hour or 0.67 gpm.

[BartzFarm]

My pump does not have a thermal switch. That’s why I was thinking of attaching one with a 150°F rating. However, after running it for 4 hours at 110 psi and pulling 2.8 amps, the casing temperature only reached 100 degrees.

The spec sheet has a graph that “denotes continuous operation capability at designated pressure and current.” It also states, “At this current or lower, the pump may run continuously, unless the surrounding temperature exceeds 70°F” Wouldn’t it seem logical then, that since I’m running it outside in 40-degree temps, I should be able to run continuously at 110 psi even if I’m pulling 2.8 amps? Or am I missing something?

I didn’t think the 5300 series could run on low input. The 5800 series can do a 7-foot lift but I gravity feed it and warm the sap to around 90 degrees.

I already bought a 24 VDC 10-amp (240W) transformer when I thought I’d be getting the 8900K. Do you really need 350-400W? That seems excessive or will this pump actually exceed 10 amps? Unfortunately, I can’t use it on the 5800 because it’s 24 VAC but I do plan on getting the 8900K eventually.

Great advice on the membranes. I’m going to need some time to wrap my brain around it.

I’ve watched and read many methods of cleaning. I’ve seen most all your videos and some of them more than twice! I love your set up in the cooler, by the way. Most people use lye (RO Soap) but I’m using hydrogen peroxide. At least for the first year. My runs are a week apart and I read that if you’re not using it for more than a few days you should clean it. I just want to know if it’s safe to leave it in the system between runs or do I have to flush it out with permeate after the initial cleaning. I thought if I let it soak for several days to a week it may be beneficial (It's already been a week and I'm hoping there's no damage)

How did you find the 2.6 lpm. Was that on a spec sheet? Or did you calculate this yourself? All I could find was stats such as >4.5 LPM @80psi.

Thanks again for your help!

[carls47807]

Aquatecs will run on 24v dc as an fyi.

[BartzFarm]

Aquatecs will run on 24v dc as an fyi.

Huh, Thanks, I really had no idea. I just spent $30 on a 24 vac 3 amp transformer. I wonder why Aquatec Tech Support never mentioned that?

Haha! Just found out my order was canceled and was issued a refund.

[DRoseum]

Glad the videos help and you like the cooler setup.

10 amps may be enough but I'd prefer to have headroom well above peak load. Typically you don't want to exceed 80% of rating.

Also, careful heating your sap. Not only does increasing the temp promote more bacteria growth (below 180F) it also runs the risk of heating about the temp limits of membranes.

As for the peroxide, I wouldn't leave it in the system that long. I'd probably rinse with permeate soon after. The oxygen released by peroxide can be pretty damaging to all kinds of materials like plastics and rubbers. Short term exposure is probably fine, but would recommend it for long term. I use food grade peroxide (diluted to 3%) to clean my tubing. One year I left it in and it made many of the fittings incredibly brittle. Now I pull it dry after cleaning with my vacuum pump and don't have any issues.

If you flush the system immediately after use liberally with permeate and let it run until it pushes air out, I think you will be fine without a wash in between uses. Also, depends on temp of where it's stored. Mine stays between 34 and 38 F so it won't freeze and it's not warm enough to breed bacteria when not in use.

I don't do a soap wash every day and typically only do one per season (if it's been really warm) and at the end of the season. Pushing tons of permeate through after each use works fine. I just did a video on membrane storage this past year amd you can see how clean my membranes look in that for reference.

[BartzFarm]

Also, careful heating your sap. Not only does increasing the temp promote more bacteria growth (below 180F) it also runs the risk of heating about the temp limits of membranes.

I hadn't thought of that one. Makes sense though. I purchased the membranes you promote and they said "Recommend work pressure is 100 psi when using this product in a Maple Syrup reverse osmosis system. And we suggest heating and maintaining the raw sap at 104°F(40°C) to prevent clog on the membrane while keeping it under 113°F(45°C) to avoid damage to the membrane due to over-high temperature." I kept it around 90 degrees.

I use a 35% concentrate of peroxide diluted to 1.2% as recommend by the Membrane Solutions "Hydrogen peroxide solution, such as diluting 17 oz (0.5L), 30% H2O2 with 3.17 gal (12L) of deionized water, and then cleaning the membrane surface, is particularly effective for organic pollution."

I just got done rinsing the peroxide out. I ran it dry then replaced the sediment filter and flushed with permeate for about 30 min then ran it dry again. We got a cold spell and the sap stopped running but next weekend looks promising. Hope it's good to leave it that way till then. I keep it in my 60 degree basement. I may run permeate through it again before the weekend. Or try that carbon filter flush again. Great idea, by the way.

I think I'm starting to get the whole cleaning and maintaining process now. Thank you so much! Oh, and I already bought some Sodium Metabisulfite and starting building my PVC pipes... Genius!... Super Genius!

I got to be honest with you though, I'm still try to wrap my head around recovery rate but haven't given up yet.

[DRoseum]

No problem at all. Not sure why they suggest heating it, probably to get performance in line with benchmark tests. I don't do that and don't recommend it. All membranes performance drops significantly when temperatures of the fluid are in the 30s/40s. Expect to get about 25% of the membrane rating for permeate output. So a single 400 gpd membrane is spec'ed to produce 16 gph of permeate. For cold sap, expect 4 gph.

Recovery rate is a number that tells you how hard you are working your membrane. You want it lower. Most of these perform ideally around 15%. You can go above it, you just get more fouling (and worse performance) the higher and higher you go.

The recovery rate of a membrane = (permeate flow output of that membrane) / (inlet flow into that membrane)

So if you keep adding more membranes in series, the inlet flow (concentrate from upstream membrane) into each membrane will progressively be decreasing due to permeate outflow of all the upstream membranes. So your recovery rate will be increasing for each membrane in series. You need to size your pump flow rate so that the inlet flow into the last membrane gives you a decent recovery rate for that final membrane. A higher flow pump will just lower the recovery rate of all the upstream membranes as well. Remember, each membrane takes out a fixed flowrate of permeate (25% of its gpd rating) at the standard operating pressure. So your flow rate to each subsequent membrane decreases by that set amount.

If you go look at my diagram, the last membrane has a recovery rate of 0.25/1.5 = 16.7%.
https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEipaceO7ZdRbH9wfbmXefyiRRS_PNK8307RTnbMT36AmS idu_r7uNGqbKkAK2GiNotsqflHqLraZ5eN-teVH9ajmrP2jnjF46MO3PKtcFypJNaqUsOosbdN6bQv5N86xF1 DPj2Lyv7KLqSOSu1ethyxtvLt2EQdaa1qBKmfAKPptlgCtnWx6 oMN-WClmx02/s1452/Roseum-RO-design.jpg
Hope this helps.

[fireant911]

... I got to be honest with you though, I'm still try to wrap my head around recovery rate but haven't given up yet.

DRoseum,
I, too, am very confused about the recovery rate. This subject comes up frequently and the last time it appeared I started looking up the topic of recovery rate and went down a few rabbit holes... and unfortunately, I am still not understanding this. In reviewing your diagram, the recovery rate that applies to your last membrane makes sense; however, some of the literature that I came across gives VASTLY different calculations for the recovery rate. For example, take a look at the diagram located here: https://www.snowate.com/knowledge-calculator/knowledge/ro-system-recovery-rate-analysis.html. According to this reference, the recovery rate for a single RO with a 15% recovery would be 15% (no surprise there!). Adding a second RO membrane in series would now have a recovery rate of 27.75%. The addition of a third RO membrane in series yields a recovery rate of 38.59%. A fourth membrane increases the recovery rate to 47.5%. With five membranes in series the recovery is shown as 55.63%. The math behind this is straightforward YET I know that this recovery rate does not bode well with what the RO manufacturers recommend. Is my confusion stemming from the fact that we are interested in keeping the stuff that is normally discharged in conventional RO systems? Could the 55.63% in a five RO in series represent that 55.63% of the solution is permeate while the remaining 44.37% would, in theory be concentrate? Am I confusing the 'system' recovery rate (i.e., when considering ALL the RO's plumbed in series) instead of looking at each individual recovery rate associated with each RO membrane? What am I misunderstanding here?

[DRoseum]

Correct, there is recovery rate for each membrane, and for the entire system. Going back to my diagram, the system recovery rate for all 5 membranes in series would be = 1.25/2.6 = 48%. Which is the total permeate outflow divided by the main inlet flow.

[BartzFarm]

If you go look at my diagram, the last membrane has a recovery rate of 0.25/1.5 = 16.7%.
https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEipaceO7ZdRbH9wfbmXefyiRRS_PNK8307RTnbMT36AmS idu_r7uNGqbKkAK2GiNotsqflHqLraZ5eN-teVH9ajmrP2jnjF46MO3PKtcFypJNaqUsOosbdN6bQv5N86xF1 DPj2Lyv7KLqSOSu1ethyxtvLt2EQdaa1qBKmfAKPptlgCtnWx6 oMN-WClmx02/s1452/Roseum-RO-design.jpg

Yep, love that diagram. Very helpful thanks!

So, if I’m understanding this correctly. Your pump has a flow rate of 2.6 lpm @ 100 psi. I’m assuming this came from a spec sheet or you measured it yourself since the pump’s name plate says 4.5 lpm @ 80 psi with a 30 psi inlet. Did you time the flow @ 100 psi with zero input pressure? How do you know the flow is 2.6 lpm @100 psi?

Since the membrane, in the current conditions, can only process .25 lpm permeate the rest of the flow goes to the next membrane. So, membrane #1 has a flow rate of 0.25/2.6 = 9.6% all the way down the line to 16.7%

Ok I get it; your flow rate is calculated for each membrane. I thought it was calculated for the whole system. And by the time you get to the end of the line, there isn’t as much flow. And that’s why those filters are the first to foul, not to mention where the highest concentration is.

I see now why you suggest a pump with more flow if I’m adding more filters.

So, if I wanted to figure out the Recovery rate for my own system. First, I would need to know the flow rate of my pump at 100 psi with zero input pressure? Is this something I can figure out on my own? Like say I ran it at 110 psi and was getting equal permeate to concentrate at the rate of 8 gph or 0.50 lpm total. This would be my flow rate?

First off, I would not be getting .25 lpm permeate/membrane. It’d be getting only half that (0.125 lpm)

My flow rate would then be 0.125/.50 = 25% at the 1st membrane and 0.125/0.375 (.50-0.125) = 33% at the second membrane. This doesn’t sound very good at all. I’m either destroying my filters, or something is wrong with my calculations.

I may have also had my recirc valve cracked open which would skew my numbers a little, but I think I'm starting to get it. I guess the only thing I’m struggling with now is figuring out flow rate.

And you also got me wondering why my membranes are underperforming!

[DRoseum]

Yes, I have observed that to be the average flowrate at that pressure based upon over 8300 gallons of sap run through these units. If you know your starting volume and have a timer you can figure out your flow rate (if your recirc loop is closed). You can also measure your combined permeate & concentrate output to get the flowrate, as it should match your input.

Btw, diaphragm booster pumps do not just linearly decline in flowrate as pressure increases. They will decline linearly up to a point, and then a steeper decline (almost exponentially) as pressure continues to rise.

[fireant911]

Correct, there is recovery rate for each membrane, and for the entire system. Going back to my diagram, the system recovery rate for all 5 membranes in series would be = 1.25/2.6 = 48%. Which is the total permeate outflow divided by the main inlet flow.

Dan,
Thanks! Your help, guidance, and patience is VERY much appreciated!!! I am glad that you also posted the drawing of your system - I have been trying to read that from your website and, because of the picture compression, the details were blurry. Now I have a nice clean and easy-to-read copy.

[ADK_XJ]

@droseum How did you source your 8900 pumps? Seems like AliExpress is the only online dealer and, in addition to seeming scammy, they want as much for shipping as the whole pump....

[BartzFarm]

Dan, you have the patience of a saint. You'd make a great teacher! I really appreciate all the help. I have a much better understanding of how this works thanks to you! I'm going to take some more measurements this weekend and get a better idea of how well my system works. Thanks again!

[DRoseum]

No problem at all guys - just glad to help out.

As for the pump, yes, I bought on aliexpress and bought the 24 vdc transformer on Amazon.

Aliexpress is very legit. Yes shipping is high, but the overall price is low compared to even less powerful pumps.

[ADK_XJ]

No problem at all guys - just glad to help out.

As for the pump, yes, I bought on aliexpress and bought the 24 vdc transformer on Amazon.

Aliexpress is very legit. Yes shipping is high, but the overall price is low compared to even less powerful pumps.

Good to know, thank you - any thoughts about doubling up two of those pumps in series to drive a 4x40? Would a 4x20 be more realistic?

[DRoseum]

You could probably run 2 of them in parallel for single 4 x 40 ultra low pressure membrane (100psi) but haven't done all the math.

I have 2 identical units of 5 x 500 (actually one membrane is 600). That's 2600 gpd membrane capacity for each and I believe many 4 x 40 in membranes are approximately 2500 gpd rated.

Remember, pumps in parallel increase flowrate. Pumps in series increase pressure. These will top out on pressure and flowrate will significantly reduce at higher pressures than ~120± psi.

That's why I'd suggest parallel and low pressure membranes (not 200±psi ones).

[ADK_XJ]

You could probably run 2 of them in parallel for single 4 x 40 ultra low pressure membrane (100psi) but haven't done all the math.

I have 2 identical units of 5 x 500 (actually one membrane is 600). That's 2600 gpd membrane capacity for each and I believe many 4 x 40 in membranes are approximately 2500 gpd rated.

Remember, pumps in parallel increase flowrate. Pumps in series increase pressure. These will top out on pressure and flowrate will significantly reduce at higher pressures than ~120± psi.

That's why I'd suggest parallel and low pressure membranes (not 200±psi ones).

Ok, thanks - and those systems you describe can match the volume output of the low pressure 4*40 membrane?

[DRoseum]

Ok, thanks - and those systems you describe can match the volume output of the low pressure 4*40 membrane?

I really don't know without doing more research.

[fireant911]

@droseum How did you source your 8900 pumps? Seems like AliExpress is the only online dealer and, in addition to seeming scammy, they want as much for shipping as the whole pump....

If you are quick, there is a new listing on Ebay for the Coronwater TYP-8900 as of this morning - $55.69 + $15.35 shipping from New Jersey. I have been waiting for six weeks and they appeared this morning. Like you, I was a bit nervous about AliExpress (though folks report no issues); however, the crazy shipping $$ was the deal breaker for me.

[dummjock]

I just built my first RO and I'm only getting 25psi pressure. Any idea why? My pump is a Kemflo 110 psi and 1.0 lpm. I'm thinking that's the limiting factor with my 2 x 400 GPD membranes.

[DRoseum]

Not familiar with that pump, but those specs make me believe it's pretty undersized for running two 400 gpd membranes. 1 lpm nominal/open flow will reduce to pretty low flowrates at 100 psi. Do you have the pump curve? The aquatec 8852 reduces down from 2.6 lpm open flow to 1.2 lpm at 100 psi and that flowrate is being enough to run 2 x 400 without getting the recovery rate too high on the last membrane.

[dummjock]

Thanks! Should I go with the 8900K or the 8852. Looks like the 8900K is a better pump for a cheaper price.

[DRoseum]

Agreed - 8900 is the way to go, especially if you might ever add another membrane.

[sweet acer]

I just built a 5x400 RO with a new 8900k pump but can only about 30 psi. I adjusted the pressure screw in untill it got extreemly loud but was only able to get 80 psi. The only difference between this build and my previous 3x400 is the size of the hose. i used 3/8 hose for the 5x400. Any suggestions?

[DRoseum]

Are you adjusting the needle valve on your concentrate line to increase pressure?

Some basic things to check that I've seen before
(1) pre filter housing installed in correct flow direction
(2) are all your membrane housing ports correct? Make sure you aren't mixing up which one is permeate amd which one is concentrate.
(3) are you processing sap? Water won't get up to 100 psi

I used 3/8 in my 5 x 500 builds without any problem.

[sweet acer]

Thank you Dave. My 81 year old brain must have been foggy as I now remember the water issue. Checked everthing else and was built correctly. We have had subfreezing weather for at least 10 days so will have to wait a few more days to test it on sap.
Thanks again.

[Mean_Oscar]

Is the same pump going to get sap to a higher ultimate concentration level with a 75 or 150 gpd membrane?

[DRoseum]

The membrane gpd rating indicates how much permeate it is able to remove at a given pressure and flowrate. The 150 gpd membrane will remove more permeate resulting in a slightly higher concentration.

That's said, I think a single 150 gpd membrane on a coronwater typ-8900 pump is really undersized. How much sap do you process at a time? That pump can support up to 5 x 500 gpd in series (2500 gpd total) and process about 40 gallons of sap into 20 gallons of 2x concentrate per hour.

You want to match your pump flowrate at 100 - 120 psi to the capacity of your membranes so that each membranes recovery rate will be in the range of 15 - 20%.

Recovery rate is defined as (permeate outflow) / (inlet flow) for each membrane. The price of a 400 or 500 or 600 gpd membrane isn't that much more than a 150 (naybe 2x as much)....but it removes 3 or 4 times the amount of water per hour.

[Mean_Oscar]

Started with 2 x 75's on an 8800 and now up to 500 on same pump. Nice to get water out faster but can't tell any difference in ultimate (flow gets down to almost nothing / stalls) concentration. Would moving 5 mic housing upstream of pump help? Realize I might need a booster pump then.

[DRoseum]

There are multiple 8800 series pumps, with the 8852 having the highest flowrate. Which do you have?

An 8852 should run a single 500 gpd membrane just fine. What pressure are you running it at?

You don't want to run it so that your concentrate is nearly stopping. That means your recovery rate is really high and that's bad for the membranes. Back off the pressure to increase the concentrate flow rate.

Don't move the filter housing in front of the pump. That will really kill performance.

You could run the 500 and 2 x 75s in series as well, especially with a higher flow pump.

[Mean_Oscar]

The '500' is actually 2 each 100s and 2 each 150s. The pumps are an 8852-2J03-B424 and another the same but B423. My thinking on moving the 5 Micron filter to the intake side of the pump was that like removing a resistor in a series circuit, the voltage drop across everything else goes up. In this case, the pressure should go up. I could stop worrying about the filter housing failing and leaking: a big part of chickening out at 100 psi.