Upgrading an RO system

[OurTown]

We are upgrading our RO system from a double membrane 4,400 GPD to a triple membrane 6,600 GPD. It is a 21 year old Propak with a 3 HP pump from Huron Valley Sales. I talked to Fred there and he said the pump is large enough for the upgrade. The feed, reject, and product water rails are all triple tapped, so I thought running a third membrane in parallel like the others would be convenient. Fred suggested a different way that they plumb their 6,600 GPD units.

Let's assume the original membranes to be #1 and #2 and call the new one #3.

Leave #1 the way it is plumbed.

On #2 keep the feed and product water hoses intact.

Run the reject hose from #2 to the feed inlet on #3.

Run the reject from #3 to the reject rail.

Remove the plug from the bottom of #2 and run the product water hose from #3 to it.

He said this way was better than just running the new one in parallel but I don't remember if he said why. Maybe it puts out lower overall reject water to product water ratio?

Wouldn't the secondary membrane clog or wear out faster because it is being fed with water that has about twice the TDS? Why run the #3 product water back through the center of the #2 membrane? Does that balance pressure?

 

[OurTown]

 

[AnalyticWash]

Leave #1 the way it is plumbed.

On #2 keep the feed and product water hoses intact.

Run the reject hose from #2 to the feed inlet on #3.

Run the reject from #3 to the reject rail.

Remove the plug from the bottom of #2 and run the product water hose from #3 to it.

He said this way was better than just running the new one in parallel but I don't remember if he said why. Maybe it puts out lower overall reject water to product water ratio?

Wouldn't the secondary membrane clog or wear out faster because it is being fed with water that has about twice the TDS? Why run the #3 product water back through the center of the #2 membrane? Does that balance pressure?

You generally always want them run in series (wastewater from 1 into second membrane, wastewater from 2 into 3rd) because you reduce waste water and increase output.

With 3HP you could probably do 4 membranes. Which you could split into two parallel systems running in series and still maintain a balanced system.

In theory, yes. You see increased wear on the downstream membranes.

That being said, run higher reject ratios and you should be fine. I have one wash with a 3 filter RO system (running in series) with 14yo membranes still putting out 0 TDS to this day. Shoot for min 2 to 1 reject to product. Make sure your system pressure is not too high. Don't go cheap on your membranes...

 

[OurTown]

You generally always want them run in series (wastewater from 1 into second membrane, wastewater from 2 into 3rd) because you reduce waste water and increase output.

With 3HP you could probably do 4 membranes. Which you could split into two parallel systems running in series and still maintain a balanced system.

In theory, yes. You see increased wear on the downstream membranes.

That being said, run higher reject ratios and you should be fine. I have one wash with a 3 filter RO system (running in series) with 14yo membranes still putting out 0 TDS to this day. Shoot for min 2 to 1 reject to product. Make sure your system pressure is not too high. Don't go cheap on your membranes...

I thought Fred at Huron Valley said that I would need to go to a larger pump for more than three membranes. Apparently, I don't understand because it looks to me if we reject 2 out of 3 gallons and keep 1 then how is that reducing wastewater? IIRC we were closer to 1:1 before. How will I know what the best pressure will be? I was told it should be between 150-200 psi. We bought the Dow Filmtech TW-30-PRO-4040 membranes and hoping those are good ones.

 

[AnalyticWash]

The wastewater reduction is running them in series vs parallel.

2:1 is very efficient for an RO system from a wastewater perspective. You can run them 1:1 but you will wear your membranes out faster. You need high reject flow to flush the membranes. If you are not already, you should be recycling all your reject wastewater.

Don't worry so much about the pressure (keep it under 200psi). Worry about the ratio.

That pump is fine for probably 5+ if you run low energy membranes. But you probably don't need that much water.

I run the XLE PRO-4040 on everything.

 

[OurTown]

The wastewater reduction is running them in series vs parallel.

2:1 is very efficient for an RO system from a wastewater perspective. You can run them 1:1 but you will wear your membranes out faster. You need high reject flow to flush the membranes. If you are not already, you should be recycling all your reject wastewater.

Don't worry so much about the pressure (keep it under 200psi). Worry about the ratio.

That pump is fine for probably 5+ if you run low energy membranes. But you probably don't need that much water.

I run the XLE PRO-4040 on everything.

I see. So does low energy mean low flow? We do not currently recycle the reject water but want to. We just need to find a simple way to use it.

 

[AquaAce]

What is everyone using the reject water for?

 

[AnalyticWash]

What is everyone using the reject water for?

I usually reject to a holding tank and then gravity feed rinse water tanks for my IBA's and/or hot water tank for self-serve/etc. The water savings are significant if you have an IBA.

 

[OurTown]

I usually reject to a holding tank and then gravity feed rinse water tanks for my IBA's and/or hot water tank for self-serve/etc. The water savings are significant if you have an IBA.

That is what I want to do but the devil is in the details as we will make over 200 gallons of reject per cycle.

 

[I.B. Washincars]

That is what I want to do but the devil is in the details as we will make over 200 gallons of reject per cycle.

But unless your auto is down, you should be using it while it's being produced. I never had a need for a gargantuan tank.

 

[OurTown]

But unless your auto is down, you should be using it while it's being produced. I never had a need for a gargantuan tank.

Yes, I remember talking to you about that at your big wash while looking at all those RO reject tanks. If we run a 2:1 ratio then we need 200 gallons but at 1:1 it would only be 100 gallons since that is about how much RO water is put into our 1,500 gallon underground tank.

 

[Keno]

A little off subject, but still related. It seems some tunnels aren't using soft, RO water anymore as their final spot free rinse anymore. Completely eliminating their RO system and the resources it uses. They are using regular city water, as they say hard water is easier to blow off the car. Just have to make sure all the water is off the car and no spots. I've heard this from a few techs, and their was a post about it on talkcarwash last week.

 

[Dick Nitro]

I see. So does low energy mean low flow? We do not currently recycle the reject water but want to. We just need to find a simple way to use it.

Industrial water plant guy here…
In my experience, low energy usually means they’re designed to allow more flow with less pressure. They have slightly wider spacing in the membrane spirals which allows good reject flow at the cost of slightly lower quality permeate water, which is nothing to fear in this application.

At the plant, we always run primary RO reject into secondary feed and then secondary reject into tertiary feed e.t.c… with all stages sharing a common permeate (demon ater) header. The more membranes in series salvages more demineralized water and reduces reject waste always.

Rule of thumb is to adjust the flow via reject outlet valve from the final stage to 25% of gross unit flow and your membranes will usually live the longest.

Monitor your pressures and you can soak fouled membranes in caustic or acid (depending) to clean them and save some $$$

cheers

 

[OurTown]

Rule of thumb is to adjust the flow via reject outlet valve from the final stage to 25% of gross unit flow and your membranes will usually live the longest.

Are you saying I should reject 3 gallons out of 4?

 

[Greg Pack]

A little off subject, but still related. It seems some tunnels aren't using soft, RO water anymore as their final spot free rinse anymore. Completely eliminating their RO system and the resources it uses. They are using regular city water, as they say hard water is easier to blow off the car. Just have to make sure all the water is off the car and no spots. I've heard this from a few techs, and their was a post about it on talkcarwash last week.

They have better more consistent drying ability than we do in IBAs. My friend's company has ripped out quite a few 6-8 membrane systems that probably cost 15-20K new.

 

[Dick Nitro]

Are you saying I should reject 3 gallons out of 4?

opposite. You reject 25% and recover 75% for your demin water tank. This is a good starting point. If you have good quality city water, you may go to 80 or 85%. All depends on how clean of water you need for your system and how crappy your feed water is.

 

[Buckeye Hydro]

I see. So does low energy mean low flow? We do not currently recycle the reject water but want to. We just need to find a simple way to use it.

No - it means they will meet factory spec output at a lower pressure. Often 100 psi membranes of the size we're talking about here are referred to as "low energy membranes," and 80 psi membranes are referred to as "extra low energy."

 

[Buckeye Hydro]

The key to minimizing concentrate flow is proper pretreatment - meaning the RO feedwater should be less than 1 gpg hardness.

 

[Buckeye Hydro]

The key to minimizing concentrate flow is proper pretreatment - meaning the RO feedwater should be less than 1 gpg hardness. and chlorine/chloramine should be pre-treated as well. Are there other constituents that must be treated/removed before the RO? Sure but they are uncommon in municipal water.

 

[OurTown]

This thing has been a bit of a struggle. I found that the inlet solenoid valve leaked past without power to it so I ordered an Asco Redhat to replace the hard to find Burkert one that was on there. Then the caps on the original two vessels were very difficult to remove. They are not the more common type that has the V-clamp but they have the almunimum split rings with the socket head cap screws tapped into them. Apparently whatever material they are made from swells after some time. I sure thought I would crack the flanges getting them out with hammers and pry bars but I didn't. After measuring them they are about .02" larger than the ID of the vessels. Crazy. Of course, you can't buy new ones now so I could have someone turn them down on a lathe, buy new vessels with the V-clamps (like the third one I added), or stuff them back in. Since I am tight for time you know what I did. I used two 3/4" pipe woodworking clamps to get them in and it was still a bit of a struggle. The next time I need to replace the membranes I'll jut get new vessels, pull the old ones out and toss them. I fired it up yesterday after replacing the hoses, gauges and adding the third membrane. The good news is that there are only two very minor leaks on the old slightly corroded brass swivel hose fittings and it puts out 2 PPM TDS which I am pleased about. Previously the best we ever got was 11. The issue is that the output is not what I was expecting. According to the flowmeter it was putting out 3.8 GPM and rejecting about 4. Then after it ran maybe 15 minutes or so it dropped to about 3.5. The reject bypass valve seems to work and I can get it to put out a lot more reject water but then the product water flow drops to around 3.0. That is a lower output than before adding the third membrane. The system pressure is running around 170 psi but the regulator doesn't seem to do anything. The only way to change it slightly downward is to adjust the reject bypass. It seems to me that it should be putting out around 5 GPM so what is going on?