RO membrane - Parallel vs Series?

[niccumec]

What are the pros and cons of running membranes in series versus parallel. My BRS 150 gpd pro “water saver” version looks to be in series allowing the same water to go through 2 rejections. Would I get better overall rejection running in parallel? I am more interested in “DI saver” over “water saver”.

If in parallel, then I assume I would need to change out the restrictor to one that is 2x current?

 

[niccumec]

NM, ran across this BRS video that seems to list out the differences.

 

[WhatCouldGoWrong71]

What are the pros and cons of running membranes in series versus parallel. My BRS 150 gpd pro “water saver” version looks to be in series allowing the same water to go through 2 rejections. Would I get better overall rejection running in parallel? I am more interested in “DI saver” over “water saver”.

If in parallel, then I assume I would need to change out the restrictor to one that is 2x current?

I was told to run them like this and crank the water pressure up:

 

[MarineandReef Jaron]

If you run the membranes in series, then the waste of the first membrane will feed the second membrane. This reduces the wastewater but also lowers the overall rejection rate.

For example. My TDS is 700 and I get 97% rejection out of my membrane so with a single membrane I get 21 TDS out of the membrane going into my DI. If I used a second membrane in series then the second membrane would get fed the waste from the first membrane. It could easily be something like 2100 TDS so the second membrane would reject 97% of that and produce 63 TDS water. Since half the water is 21 TDS and half is 63 TDS the combined average out of both membranes would be about 42 TDS. This will burn through a lot more DI resin than if you fed the DI resin 21 TDS water, but you would also have lower waste water usually going from something like 3:1 waste to product to 2:1 waste to product.

What you need to decide is do you want to save water or save DI resin. If your tap TDS is very high then you likely want to run the membranes in parallel. If your water cost is high then running the membranes in series might make more sense. If you have expensive high TDS water then you need to pick your poison.

 

[niccumec]

I was told to run them like this and crank the water pressure up:

Looks like you are in parallel. Interesting. That is what I was considering.

 

[niccumec]

If you run the membranes in series, then the waste of the first membrane will feed the second membrane. This reduces the wastewater but also lowers the overall rejection rate.

For example. My TDS is 700 and I get 97% rejection out of my membrane so with a single membrane I get 21 TDS out of the membrane going into my DI. If I used a second membrane in series then the second membrane would get fed the waste from the first membrane. It could easily be something like 2100 TDS so the second membrane would reject 97% of that and produce 63 TDS water. Since half the water is 21 TDS and half is 63 TDS the combined average out of both membranes would be about 42 TDS. This will burn through a lot more DI resin than if you fed the DI resin 21 TDS water, but you would also have lower waste water usually going from something like 3:1 waste to product to 2:1 waste to product.

What you need to decide is do you want to save water or save DI resin. If your tap TDS is very high then you likely want to run the membranes in parallel. If your water cost is high then running the membranes in series might make more sense. If you have expensive high TDS water then you need to pick your poison.

That makes sense. I am more concerned with DI depletion than water conservation. Perhaps I should try parallel.

Do in need to double the restrictor rate when going from series to parallel?

 

[a.t.t.r]

Double pass to system. Clean from one to input of other…. Too lazy to do the required restrictor math tonight or if a pump would be needed but it is a thing. Can even pump 2nd units waste back into the first units feed (will further drop initial tds)


Only half joking here…. I always wanted to try it.

 

[niccumec]

If you run the membranes in series, then the waste of the first membrane will feed the second membrane. This reduces the wastewater but also lowers the overall rejection rate.

For example. My TDS is 700 and I get 97% rejection out of my membrane so with a single membrane I get 21 TDS out of the membrane going into my DI. If I used a second membrane in series then the second membrane would get fed the waste from the first membrane. It could easily be something like 2100 TDS so the second membrane would reject 97% of that and produce 63 TDS water. Since half the water is 21 TDS and half is 63 TDS the combined average out of both membranes would be about 42 TDS. This will burn through a lot more DI resin than if you fed the DI resin 21 TDS water, but you would also have lower waste water usually going from something like 3:1 waste to product to 2:1 waste to product.

What you need to decide is do you want to save water or save DI resin. If your tap TDS is very high then you likely want to run the membranes in parallel. If your water cost is high then running the membranes in series might make more sense. If you have expensive high TDS water then you need to pick your poison.

My tap TDS is only 180, but running though in Series generates 8-10 TDS, which is a low 95% rejection, which does not make sense to me as I am feeding the system with 90 psi, 70 deg relatively low TDS feed. Many sources say that once recovery gets below 96%, then it is time to change the membrane. My two Dow 75GPD membranes are new. I have run ~200 gal product through it and still getting ~8 TDS ahead of the DI.

 

[niccumec]

I was told to run them like this and crank the water pressure up:

I follow your set-up - except for the T’d clear tubes. What are you using those for?

 

[Gtinnel]

If I used a second membrane in series then the second membrane would get fed the waste from the first membrane. It could easily be something like 2100 TDS so the second membrane would reject 97% of that and produce 63 TDS water.

When you say your feed into the second membrane could easily be 2100 did you actually calculate at this value or did you just guess a number? I only ask because if I understand it right if the feed into your first membrane is 700 at a 97% rejection rate, and your waste to product ratio is 3:1 then the feed into your second membrane should be 905, unless I’ve always calculated it wrong.
I agree with your general explanation of how membranes in series work, but I think you have grossly exaggerated how high the tds in the first membranes waste would be.

 

[MarineandReef Jaron]

When you say your feed into the second membrane could easily be 2100 did you actually calculate at this value or did you just guess a number? I only ask because if I understand it right if the feed into your first membrane is 700 at a 97% rejection rate, and your waste to product ratio is 3:1 then the feed into your second membrane should be 905, unless I’ve always calculated it wrong.
I agree with your general explanation of how membranes in series work, but I think you have grossly exaggerated how high the tds in the first membranes waste would be.

I have tried using the TDS Meter on the outflow and it just errors out. I also run closer to 2-1 with a trimmed restrictor and booster pump. So mathematically it may be closer to 2000.

The bigger point is it is just ridiculous to run 2 membranes in series if you have high TDS the cost of the extra DI resin consumption will far surpass the cost of the increased water bill.

 

[PeterC99]

If you run the membranes in series, then the waste of the first membrane will feed the second membrane. This reduces the wastewater but also lowers the overall rejection rate.

For example. My TDS is 700 and I get 97% rejection out of my membrane so with a single membrane I get 21 TDS out of the membrane going into my DI. If I used a second membrane in series then the second membrane would get fed the waste from the first membrane. It could easily be something like 2100 TDS so the second membrane would reject 97% of that and produce 63 TDS water. Since half the water is 21 TDS and half is 63 TDS the combined average out of both membranes would be about 42 TDS. This will burn through a lot more DI resin than if you fed the DI resin 21 TDS water, but you would also have lower waste water usually going from something like 3:1 waste to product to 2:1 waste to product.

What you need to decide is do you want to save water or save DI resin. If your tap TDS is very high then you likely want to run the membranes in parallel. If your water cost is high then running the membranes in series might make more sense. If you have expensive high TDS water then you need to pick your poison.

This is an excellent analysis!