Re-circulation Explained

[Mead Maple]

Hi guys,
While all of us rookie RO enthusiasts are trying to squeeze every bit of knowledge out of you experienced operators, can we get a bit more in depth about re-circulation? We have a general idea (I think) about it but can we get more specific about how it benefits our systems and why we would opt to buy a bigger pump to be able to utilize it:confused:. I know it's been mentioned in other posts but I thought it might be good to have its own thread as an easier way for folks (like myself) to access the information. At some point I'll have a 3 ring binder full of RO information with reference tabs! Thanks everyone for your participation in all the RO threads, it's been awesome reading all this and many of us appreciate the wealth of knowledge and explanation.

[wmick]

Sorry to hijack.... I am also a rookie... but I will take a stab at answering... Take it with a grain of salt.. Hoping one of the experts will take me to task if I'm off base....

Re-circulation will aid in increasing your output concentration rates, while operating in a continuous flow fashion, rather than batching... while adhering to membrane recovery recommendations.

Membranes are designed with a specific recovery rate in mind, to prevent fouling... I understand that maple membranes are typically rated at about 15%.
Which means you should take no more than 15% of your feed-water (sap) off as permeate, per pass.
There's a pretty good explanation of why fouling happens at the link below... (It is describing Water filtration... but the concept will be the same with sap..). Read the Feedwater Chemistry section.
https://blog.harnrosystems.com/how-to-calculate-percent-recovery-and-what-affects-it-in-a-membrane-treatment-system
My simpleton way of looking at it is..... I picture the membrane like a perforated screen filter. Lots of flow across the surface of the screen will keep it wiped clean and keep the sugar moving fast and liquidized, and prevent it from landing in the holes and plugging them up as it crystallizes....

So... the other 85% of the flow needs to go somewhere... Choices might be...
... into your concentrate tank at low brix concentration, which you could boil... or run it back through the RO a second or 3rd time later. (basically recirculating in batches)
... some or all of it, back to your sap tank, in which case you are still recirculating... and will eventually bring your entire stock of sap up in Brix... as a batch.
... or reintroduce most of it back into the piping loop, immediately before your pressure pump... in which case most of the sap will get several passes through the membrane while keeping your flow high and recovery rate good, and producing a higher concentration in one operation. No batching. This can be done with the one high pressure pump... while you bleed off small volume of higher concentrated product.
... or reintroduce most of it after your pressure pump.... Same as above but--- you will need to add a dedicated recirculating pump. Haven't heard of anyone on here using this on a small rig... but might be an option if you are looking for high flow rates that a typical procon wont produce.... or just because you want to.. This system could theoretically run on less horsepower, because you'd be recirculating inside of the pressure system (where there will be very little pressure drop across the recirculating pump. This would use less energy than re-pressurizing all of your recirc volume... Smaller procon required just for the concentrate volume and a small HP circulator to move sap around in the loop.)
.... or Dont recirculate - Just keep adding more membrane housings in series... Essentially creating several "membrane passes" in one straight through operation... ending up with finished brix and not needing to recirculate at all. I think the cost of membranes and housings is what prevents this one from happening a lot.??

OK - Give it to me..... I can take it... :cool:

[maple flats]

If that's the case I have had it all wrong ("Membranes are designed with a specific recovery rate in mind, to prevent fouling... I understand that maple membranes are typically rated at about 15%.
Which means you should take no more than 15% of your feed-water (sap) off as permeate, per pass.")
I run at about 240-260 GPH on my little RO, and I remove about 3x as much permeate as I get concentrate on my first pass. That is 75% if the sap temperature is 37-38 F. I don't know where the figure of 15% came from, but it is not talking about maple sap.
After the 1st pass, I will have taken 2% sap up to 8% concentrate +/- a little depending on the temperature. Then I often recirculate the concentrate thru the RO a second or more time. When I do that I need to back off the operating pressure and then I only remove a smaller portion in permeate per minute. On my RO I often get 3 gpm of permeate and 1 gpm of concentrate if I set the pressure at about 280 PSI, If I go down to 270-275 the ratios change, Then I get slightly more concentrate per minute and slightly less permeate per minute. The total is still essentially the same. My re-circulation then might go to the opposite or nearly opposite the ratio I got the first time. Thus my little basic RO on second pass might just remove 1 gpm of additional permeate. My RO only can handle pressures up to 300 PSI, more expensive ones can be run at higher pressures and thus can remove more permeate on the first pass. Many can do between 12% and as high as over 20% in one pass, the newest ones designed for hi brix can get 35% concentrate in one pass, rather than the 8% I can get.
In my case recirculation is done 2 ways. First, on my "first pass" I open a re-circulation loop, that sends some of the flow out of the membranes back thru the high pressure pump (higher end RO's use another pump to push re-circulation back thru the membranes). The other way I re-circulate is by running my concentrate that was pushed out of the RO, back thru again.
A lot of this is going to be based on your needs and more your pocket book. An RO like mine only cost a very small % of what a high brix RO costs. It's a matter of how much sap do you need to process and how fast do you need to do it. You might think of it as balancing $ with time.

[arcticmaple8]

15 percent rejection rate means 85 percent pure water gets through and 15 percent of feed water is wasted, rejected to drain, in a water purification system. In maple the reject waste is our concentrate. Re-read your article that you linked.

[Mead Maple]

This is great, thanks guys.

So if I wanted to add a re-circulating pump to the system, how do I go about sizing that? Would it be the same as my primary HP pump or could I get away with something smaller? Also, does that line get plumbed directly into the original line with a "Y" fitting that would go into the membrane inlet?

[wmick]

15 percent rejection rate means 85 percent pure water gets through and 15 percent of feed water is wasted, rejected to drain, in a water purification system. In maple the reject waste is our concentrate. Re-read your article that you linked.

I am just learning myself... and I may have it all wrong.. but I was refering to "recovery rate" and not rejection rate.
This Lapierre Link defines Recovery as "The percentage of the water removed from the raw sap through the membrane."
http://www.sugaringequipment.elapierre.com/down/191.pdf


The typical "15% Recovery Rate" specifications is something that I have read on this forum several times...
I have not found a lot of data to show what happens if you exceed this rating.... So... at this point I am just assuming it to be valid.. I haven't looked up a lot of different membrane spec to confirm the information... But below is a link to a MES maple membrane..... You can see that its test condition rate is 15%. To be honest, I don't know why they choose 15% for testing maple membranes... When I look at Saltwater membranes they tend to test at much lower percentages.

Before I head too far down the wrong path.... I have sent an email to a membrane manufacturer, asking for some clarification on Recovery Rate. I will pass it on if they reply.

https://static1.squarespace.com/static/5a11d71de45a7cc213905883/t/5a5a5f518165f5b7c1c560af/1515872082484/MES+MS2+membrane+spec+sheet+4+inch.pdf

[wmick]

I found something of interest to the RO enthusiast and student.

Below is a link to a Maple membrane training presentation from H2O Innovations.... Good resource....
Took me a while to find an english version... It doesnt seem to be on their website...
and I've attached a snip of one slide of particular interest... for this "Recovery Rate" / "re-circulation" conversation
They are discussing things that influence Membrane Clogging.... The first one being "Non-observance of the membranes recovery rate"

20084

https://www.scribd.com/document/114735125/H2O-Innovation-Maple-Membrane-Training-2011-ENG-pdf

[wmick]

I run at about 240-260 GPH on my little RO, and I remove about 3x as much permeate as I get concentrate on my first pass. That is 75% if the sap temperature is 37-38 F. I don't know where the figure of 15% came from, but it is not talking about maple sap.
On my RO I often get 3 gpm of permeate and 1 gpm of concentrate.
First, on my "first pass" I open a re-circulation loop, that sends some of the flow out of the membranes back thru the high pressure pump (higher end RO's use another pump to push re-circulation back thru the membranes). .

Hey Maple Flats
Just trying to clarify if we are on the same page with the terminology - When they talk about the "Recovery" rate on a membrane, it is not talking about total system output... but rather... what is happening with that particular membrane at any moment in time... (total flow across the membrane includes anything that was reintroduced by recirc)... (and if you have more than one membrane in series.. then recovery is divided between them)

You mention that you are recirculating.... Do you know how much? Do you know what the total output of your pump is in GPM?
Do you have more than one membrane in series? Thats a big factor...
Your total system recovery rate is 75%.. (at 3Perm+1Conc). but your recovery % on the membrane is less if you are recirculating and/or have more than one membrane.
Curious - What does your rig entail? for pump gpm and membrane configuration ???

[arcticmaple8]

Wmick: i read your link to lappiere. Good read. I see it recomends mot concentrating past 70% in beginning and uses 75% as an example in the testing procedure. This is very typical in maple ros. I didnt see anything about 15% recovery.

[arcticmaple8]

Mead maple: hope you didnt give up on your question. The pressure pump builds your high pressure but at a lower flow rate while the recirc pump moves sap in a loop through membrane while under that high pressure but at a high flow rate. An example of a single 4 x 40 inch ro would be a high pressure pump procon 240gph(4gpm) which would require approximately 1hp to build 250psi at 4gpm. Adadd a recirc within the pressure loop and you dont need any additional pressure so a 15gpm centrifugal pump with low pressure high volume will only require approximately a 1/2 hp motor but this pump must be capable of being under 250psi... sizing your recirc is about the max flow through your membrane per manufacturer spec. An xle4040 is 15gpm i believe. With no recirc your membrane gets 4gpm from pressure pump minus the permeate flow of 1gpm maybe. So without recirc membrane will plug faster and require rinsing/wash sooner

[maple flats]

wmick, to address your questions. My pump moves just over 4 gpm. I have no measure of how much is recirculating, it utilizes a small needle valve just ahead of the main needle valve that controls the pressure of the system and I have 2- 4x40 membranes in series. My pump is belt drive off a 2HP motor.
I only get my flow numbers from observing the flow meters.
On my RO with no recirculation I get barely over 4 gpm, adding the concentrate and permeate flows. With 2% sap I can adjust those ratios from well below 1 gpm of concentrate at about 285 PSI (about .6 gpm) up to about 1.8 gpm at between 240- 250 PSI, in each case the balance of just over 4 gpm is in permeate.
I run most of the time, when inputting about 2% sap at 270-275 PSI and get about 1.1-1.2 gpm of concentrate and 2.8-2.9 gpm permeate. That is with my recirculation needle valve open about 1/2-2/3 turn (I have no idea how much is recirculating when I do this.
Just an observation, when on wash cycle, as the cycle is nearly done I get readings of about 2.6-2.8 GPM on both flow meters and at that the pressure is about 60-70 psi. All during my wash cycles I try to balance both flow meters, Early in the cycle it is not achieved, but within a couple of minutes I can balance the meters. From then on I adjust the pressure to maintain equal flows, as the membranes get cleaner the pressure needs to be set slightly lower and the totals of the 2 flow meters slowly climb until I get to between the 2.6-2.8 GPM on each.

[bowhunter]

Your explanation looks pretty accurate too me. i would add one more benefit of recirculation. Recirculation allows the sugar concentration to be more uniform across the entire membrane and therefore the permeation rate and fouling rates are more uniform. For example if you are running once through and the sugar is 2% on the inlet it end, it's about 4% on the outlet with 50% water removal. The pressure is roughly the same on the inlet and outlet ends. Because the sugar content is half as high on the inlet end, the osmotic pressure is low and the permeation rate and fouling rates are very high. The inlet end starts to foul very quickly and becomes completely plugged over a short period of time. The effective area of the membrane becomes smaller and smaller as it continues to run. I run less than 15% recovery with pretty high recirculation rates. I can run my membrane for at least 5 all day runs before it requires soap washing. This year I never had to soap wash during the season at all. I only did permeate flushes at the end of every run. The second benefit to minimal soap washing is limiting damage to the membrane. Soap washing can slowly degrade it over time.

[wmick]

Wmick: i read your link to lappiere. Good read. I see it recomends mot concentrating past 70% in beginning and uses 75% as an example in the testing procedure. This is very typical in maple ros. I didnt see anything about 15% recovery.
Yes - I find it a bit confusing. Hoping someone might clarify.

Slide #7 says this.
"According to membrane manufacturers, a membrane must operate with a maximum recovery rate of 15%.
RECIRCULATION or an increase in the NUMBER of membranes allows increasing the total recovery rate of a system without damaging the membranes."

Yet - Slide number 8 has a chart showing 75% recovery...

So - What gives??? Is the chart on slide 8 assuming recirculation? or are they operating at 5X the manufacturers recommendations? or ??

I sent the question to H20 Innovations.... Hopefully they can clarify.... The other 2 manufacturers that I emailed last week have not yet.

[wmick]

I run most of the time, when inputting about 2% sap at 270-275 PSI and get about 1.1-1.2 gpm of concentrate and 2.8-2.9 gpm permeate. That is with my recirculation needle valve open about 1/2-2/3 turn (I have no idea how much is recirculating when I do this.
Just an observation, when on wash cycle, as the cycle is nearly done I get readings of about 2.6-2.8 GPM on both flow meters
So - your "recovery rate" is maybe not quite as high as you thought...
If we assume that your wash flow is actually your total pump flow... your total flow is 2.8 + 2.8 = 5.6 gpm
and your permeate is 2.8 gpm. divided by 2 membranes = 1.4 gpm per membrane with 5.6 gpm total flow. 1.4/5.6= 25% membrane "recovery" rate.

[maple flats]

when I have my re-circulation needle valve closed, the 2 do not equal 5.6 GPM except near the end of a wash cycle. Thus that math is in error. With that needle valve closed and just the main needle valve open to give me about 275 PSI I get a total flow when I add both of 4.0-4.2 gpm. When I observe the 2.8 and 2.8 the pressure is about 40-50 PSI. Thus, with my Deer Run RO the pressure make a big difference of what total flow I get.
Back in my insulation background and my farming background with mixing chemicals to create foam insulation and with irrigation as well as spraying pesticides, as the pressure increased the flow decreased, in every case. The pumps do not have a linear performance, it is a decreasing curve.
For example, on my irrigation pump (a 13 HP centrifical 3" pump) I used to get 43 sprinklers flowing and impacting properly and got over 60 psi at the sprinklers for 60 gph/sprinkler or about 2600 gph. When I was just transfering water, with a 17' rise to refill my pond I got over 10,000 gph at about 30 PSI open discharge. In both cases the throttle was on high. Pump performance is not linear.

[maple flats]

My thought is that you are too hung up trying to calculate flow with a small portion of the data that would be needed.

[wmick]

My thought is that you are too hung up trying to calculate flow with a small portion of the data that would be needed.
You are quite right... Just trying to wrap my head around all the different variations... I do understand that pumps have a psi/flow curve at a given rpm. What type of high pressure pump does your Deer Run use? I'm a little surprised at your "curve". Vane pumps seem to be common , from what I've read on here... and procon's series 4 curve only varies by about 0.2 gpm between 50psi and 250psi ... Are other pump styles used too?
Thanks

[arcticmaple8]

Vane pumps and piston pump are positive displacement. They will give the same relstive flow at different pressure. More pres requires more hp. Multistage or other centrifugal pumps have pump curves showing flow vs psi

[wmick]

Vane pumps and piston pump are positive displacement. They will give the same relative flow at different pressure. More pres requires more hp. Multistage or other centrifugal pumps have pump curves showing flow vs psi
Yep - agreed - This is why I'm surprised at Maple Flats' flow difference at different pressures.... I'm thinking it must be a different type of pump??.

[bowhunter]

Yes. Different type pumps have very different curves or pressure/flow relationships. The ProCons are considered positive displacement and show very little difference in flow vs pressure. A centrifugal pump on the other hand is very responsive to back pressure.

To your recovery question: Without recirculation the recovery is very easy to calculate for a single membrane system. It is the permeate flow divided by the total sap flow to the membrane expressed as a %. With recirculation the membrane recovery is not as easy to understand. Lets suppose you have an single membrane RO system with recirculation that is feeding 100 gph of fresh sap into the system, and it is removing 75 gph of permeate. The system recovery is 75%, however the membrane recovery is 15% because the flow through the membrane including recirculating concentrate and fresh sap is 500 gph. So you are recirculating 400 gph of concentrate and adding 100 gph of fresh sap. Hope this helps.

[needmoremaples]

Is there a type of valve you can use to set your back pressure? In a recirculation setup I need to throttle back the pressure as sap % builds. I have a 4x40 with procon pump and 1 hp feed pump. I'd like to just set the pressure to 225 and check it like every hour. Of course I'd prefer foodgrade or at least for potable water.

[Sappy Camper]

If that's the case I have had it all wrong ("Membranes are designed with a specific recovery rate in mind, to prevent fouling... I understand that maple membranes are typically rated at about 15%.
Which means you should take no more than 15% of your feed-water (sap) off as permeate, per pass.")
I run at about 240-260 GPH on my little RO, and I remove about 3x as much permeate as I get concentrate on my first pass. That is 75% if the sap temperature is 37-38 F. I don't know where the figure of 15% came from, but it is not talking about maple sap.
After the 1st pass, I will have taken 2% sap up to 8% concentrate +/- a little depending on the temperature. Then I often recirculate the concentrate thru the RO a second or more time. When I do that I need to back off the operating pressure and then I only remove a smaller portion in permeate per minute. On my RO I often get 3 gpm of permeate and 1 gpm of concentrate if I set the pressure at about 280 PSI, If I go down to 270-275 the ratios change, Then I get slightly more concentrate per minute and slightly less permeate per minute. The total is still essentially the same. My re-circulation then might go to the opposite or nearly opposite the ratio I got the first time. Thus my little basic RO on second pass might just remove 1 gpm of additional permeate. My RO only can handle pressures up to 300 PSI, more expensive ones can be run at higher pressures and thus can remove more permeate on the first pass. Many can do between 12% and as high as over 20% in one pass, the newest ones designed for hi brix can get 35% concentrate in one pass, rather than the 8% I can get.
In my case recirculation is done 2 ways. First, on my "first pass" I open a re-circulation loop, that sends some of the flow out of the membranes back thru the high pressure pump (higher end RO's use another pump to push re-circulation back thru the membranes). The other way I re-circulate is by running my concentrate that was pushed out of the RO, back thru again.
A lot of this is going to be based on your needs and more your pocket book. An RO like mine only cost a very small % of what a high brix RO costs. It's a matter of how much sap do you need to process and how fast do you need to do it. You might think of it as balancing $ with time.

Hello Dave

I am relatively new to the Trader but have lurked for a long time. I grew up in the maple biz, but we boiled raw sap from our 4000 buckets on a 5x16, sometimes working 3 shifts, with no RO. In the olden days. I went to college, had a career, and have recently gotten back to sugaring on a much smaller scale, 275 taps this year, with ambition for no more than 500. I make candy with every bit of syrup and have a small business selling it.
I have read many posts of yours and you have imparted a great deal of thoughtful advice over the years, a very helpful resource to many. Thanks. I read the post quoted above with great interest, as I begin to think about building an RO. I was searching for info on recirculation and came across this post. I am wondering if you have any advice or modifications you would make on that system? I would like to make one similar, and would love to see a schematic (sketch) if you have one. I have questions about other parts of the system like gauges, interlocks, and a feed pump if you have one. It seemed like yours was low tech and manual operation, which for the cost of things now is what I plan to do. I don't know if you are even on the Trader anymore, as your tagline implies you have retired from maple. I hope you get this message and that you are well

Tim Bishop
Palermo, ME

[maple flats]

Sorry Tim, I'm not an engineer, I just observe what I see in my meters and gauges. Afraid I can't help.