Watermaker Multiple RO Vessels

[tfsilver]

I am currently building a watermaker. I can see that adding a secondary
RO vessel and filter to take the high pressure brine from the primary vessel
and process it again to get more fresh water. This seems to make sense in tat you do not have to use any more pump energy to extract more water. The only downside that I can see is the extra cost of the vessels and subsequent maintenance of 2 filters and more pickling soln. Does anyone know if I have this wrong.? From what I can see you get about 20% freshwater from the primary filter and then you get 20% more freshwater from the 80% brine reject of the primary. You should get this without any more energy input.

[Palarran]

I do think you have it wrong. The cores go in parallel to each other, not in series. If they where in series, the pressure required to push the water through the membrane would increase (double?). I would guess you would either get 0 water out, or half.

Of course, I could also be wrong.

[sailorchic34]

Actually its the same either way, not parallel or series, its just two membranes on the same pressure circuit. Yes you'll get more capacity assuming the pressure pump is sized for the total area of membrane. Its not really processed water again though as the water is split between the two membranes. (IE the water does not really care which membrane it sees). The trick is your going to need a pump large enough to supply both membranes pass through rate, plus reject water

[Tristan]

You will also have some drop of pressure on the second membrane and because higher TDS concentration down the stream, the osmotic pressure will get higher which will result in lower differential pressure and lower efficiency. Another words you will get less product from the every next added membrane.
T.

[s/v Jedi]

There is a big difference between series and parallel. I have seen self-built units that have membranes in parallel, but all suppliers will put them in series (those I have seen anyway).

The reason is that a membrane is a cross-flow filter and it really needs a high volume of water (reject) to rinse out the membrane. When you start connecting in parallel, you divide the water volume over the membranes; when you put them in series, the 2nd will still have 80% of what the first had and the 3rd will have 60% of the first (compared to 33% for parallel, so double).

With series connection, it is usually the 2nd membrane that breaks down first because it always gets a little less cross flow. This means that it is helpful when you can adapt to just using the first membrane when this happens, as the first membrane will still be good for a bit longer... it'll give you plenty time to order and wait for new membranes.

ciao!
Nick.

[Tellie]

Nicks right, manufactures of the name brand watermakers will use a series set up. It works best. Remmber it's all about pressures and flow. The pressue is easy the flow is where a pump has to be up to the job. Again, as Nick says, membranes are cross flow filters. The flow is very important to keep the membrane clean. Adding extra membranes will certainly reduce the flow. So if the pump is designed to create just enough pressure and flow for one membrane adding a second will greatly shorten the life of both membranes. But usually in home built units people tend to oversize a good quality pump so they are usually up to the job of two 40" membranes. Using a cheap pump salvaged from a cheap pressure cleaner just won't be up to the job.

[sailorchic34]

Oh concur with everything said. You'll need a higher flow to keep the membrane clean. From a parallel or serial standpoint, for small RO systems,the pressure seen in both units and amount of clean water made would be practically the same. Though the through flow rate would be 1/2 per unit in a parallel systems and you do need to keep the reject rate up in parallel. In serial both units see total flow. There is some savings in serial installation in the number of fittings/hose required to hook things up.

[Palarran]

Sorry, I read the question wrong. I was thinking tf silver wanted to take the freshwater out of the first membrane, run it through a second to make it more clean.

It does seem that the second membrane would have a much higher salinity rate supplied to it and therefore would have less output, but I don't really know. I have two membranes so I should look to see how they are piped.

[tfsilver]

The system I am building is as follows
Motor : 1/3 HP DC
Pump: Cat 2SF05SEEL
Two Filters a 20 micron and 5 micron upstream of the cat pump
This should give me a flow of .6 gpm or 36 gph of seawater to the membrane(s).
Now this is where my question comes in regarding the membrane.
I plan to use either a Filmtec SW-2514 or a SW-2521. Both these membranes have a specification of a minimum concentrate(seawater) flow of 1 gpm at 800 psi in order to flush the membrane. As you can see that I only have .6 gph at 800 psi. Logic tells me that the system will work ok with maybe not getting the secified flushing which would cause the membrane to clog earlier than expected. The membrane is also designed to operate at a lower pressure than 800 psi which could give more flow for better flushing but less permate production. I am opting for the 21" membrane as againg from logic it has more surface area for the permate to squeeze through so it should take longer to clog with the less flushing.

Do you think that the flushing is a problem at the .6 gpm at 800 psi.? Would you use the 21 inch membrane?

SW30-2521 ( Max Permate Per Day 300GPD )
SW30-2514 ( Max Permate Per Day 150GPD)

The last question is regarding a secondary membrane in series?. Below are my numbers
Primary Filter Calcs
At 36 GPH x .20 = 7.2 GPH or 172 GPD so use the SW30-2521
Seconday Filter Calcs
At 36 GPH x .80 x .20 = 5.7 GPH or 138 GPD so use the SW30-14

So the question is whether it worth the effort to add a secondary membrane to possible get another 5.7 gph max. I realize it will be less due to a pressure drop in the first membrane and higher salinity in the brine going to the secondary membrane.
I would love your comments.

[Tellie]

First off don't play around, get a bigger pump that's up to the job. A Cat or a Giant. Secondly, don't waste your money on two SW2521s. Find the room for a SW2540. The cost difference between one SW2521 and one SW2540 is just a few dollars. Membranes have a funny way of going bad at the worst of times. It will be half the cost to replace only one SW2540. Also you won't have to invest in a second membrane vessel with the extra asscociated plumbing, and you can put that money towards the better pump. Also don't even rack your brain over the difference of concentarted brine in a second membrane, the difference of product is barely measurable and a SW2540 will out perform two SW2521s to more than make up the difference.

[44'cruisingcat]

A (kind of) related question. If you have a 60litre/hour watermaker, and you replace the membrane with a 120litre/hour one, I can see that your bypass ratio would be reduced, and more fouling of the membrane would be likely. (ie, if the pump was sized to run a 60 litre/hour system)

But what if you reduced the pump output pressure so the product flow through the new membrane was back around 60 litre/hour or so?

I'm thinking it would give the same bypass ratio so should be Ok from a fouling point of view, but would reduce the load on the pump, thus it's current draw, and so give more litres per amp/hour.

Would it, or is there a problem I haven't considered?

[Tellie]

You might be confusing the pressures required for RO and the pressures required to cross flow the membrane. It's a little confusing to explain. Think for ease of math a 10 to 1 ratio. To create one gallon of product water it will take 10 gallons of sea water to do so. One gallon of fresh goes to your tank while 9 gallons go overboard as brine. Basically it takes 800psi for a good balance of production and longevity of the membrane. The real ranges of pressure a membrane will work under is between 700-1200psi. Under the 800psi range it will work but production will drop off. Higher than 850psi and in most typical boat watermakers the product will increase somewhat but sooner rather than later the membrane will wear down and shorten its life and total life expectancy of product output. In this pressure range 16gph is a good number to expect. Now there is also the pressure needed to drive the 9 gallons of water over the membrane surfaces to keep them properly washed so they don't quickly foul. Reducing the pressures on the pump will reduce the cleaning flow action. The way membranes are made they are rolled kind of like a roll of paper towels. The pressures and flow are directed through the end. There has to be enough pressure to force the water through the end of the membrane all the way to the other end without even considering the pressures needed to start the RO process within the membarne itself. If you reduce the pump pressure it has a harder time getting the 9 gallons of water through thus reducing the flow. If the membrane manufacturer where to roll the membrane even tighter it would take even more pressure to force the same amount of water through. But they design a good balance of coiling taking into consideration the amount of pressure to start the RO process through the membranes surface. So remember it takes pressure to push water and pressure to start the RO process. Two different actions combined. So there is the balance you upset with reducing pump pressure. Taking it to the end degree, think of it this way, you could simply block off the brine discharge side of the membrane vessel, fill it with salt water and pressurize it too 800psi you'll make water until the pressure equalizes(Reverse Osmosis) and you introduce more salt water and more pressure, but without the exit for the brine the membrane will quickly clog. In essence that is what you are doing on a slower scale by lowering pump pressure. Adding a second or larger membrane has the opposite effect and flow must be increased and pressure maintained to meet the added membrane surface and balance. That is why a series conection works somewhat better than a parallel plumbed set of membrane vessels. This is one of the many areas DIY builder get into trouble with building their units. In the quest for either the most economical amp wise watermaker or the perverbial fire hose watermaker they get into trouble and their units either have to run for extremly long times or they have to run long enough to keep the membranes healthy and dump water over board. Usually they don't understand either and though their watermakers seem to run fine for awhile, down the road membrane replacements get costly and reconfigurations add up as well. I've built a few watermakers early on and learned that what seems to make sense on paper and on the Internet doesn't always work out in the real world. But ain't it the way with most things.

[44'cruisingcat]

Thanks for the detailed answer Tellie. I'm not sure I worded the question very well.

Let's say I have a 60LPH watermaker, with a 10:1 bypass ratio. So the pump delivers 660LPH at 800 psi.

I'm wondering if I changed the membrane to one rated at 120LPH at 800psi, but ran it at a lower pressure, so it was still only producing 60LPH of product, so it would still be running at a 10:1 bypass ratio, because the pump is still delivering 660LPH, but at maybe 650-700psi. (the bypass ratio might possibly be slightly more, since the pump tends to run a little faster at lower pressure.)

But by running at the lower pressure it might use a little less power. (I've observed this - at lower pressure the pump does run faster and draws less amps.)

[goboatingnow]

Originally Posted by 44'cruisingcat

Thanks for the detailed answer Tellie. I'm not sure I worded the question very well.

Let's say I have a 60LPH watermaker, with a 10:1 bypass ratio. So the pump delivers 660LPH at 800 psi.

I'm wondering if I changed the membrane to one rated at 120LPH at 800psi, but ran it at a lower pressure, so it was still only producing 60LPH of product, so it would still be running at a 10:1 bypass ratio, because the pump is still delivering 660LPH, but at maybe 650-700psi. (the bypass ratio might possibly be slightly more, since the pump tends to run a little faster at lower pressure.)

But by running at the lower pressure it might use a little less power. (I've observed this - at lower pressure the pump does run faster and draws less amps.)

There is a lower limit to pressure and permeate production falls away quite drastically as you approach it ( see Dows spec sheetsb) Also you need to maintain the 1gpm for filmtec membranes to ensure long life

Note that since its a back pressure system. The system pressure is not a function of the pump. The pump speed only controls flow.

Dave

[ImaginaryNumber]

Originally Posted by Tellie

...This is one of the many areas DIY builder get into trouble with building their units. In the quest for either the most economical amp wise watermaker or the proverbial fire hose watermaker they get into trouble and their units either have to run for extremely long times or they have to run long enough to keep the membranes healthy and dump water over board. Usually they don't understand either and though their watermakers seem to run fine for awhile, down the road membrane replacements get costly and reconfigurations add up as well. I've built a few watermakers early on and learned that what seems to make sense on paper and on the Internet doesn't always work out in the real world. But ain't it the way with most things.

Tellie,
Can you recommend a good set of plans for the DIY builder?