Recirculation Pump

[labman]

Can anyone tell me what type of recirculation pump can be used in a true RO pushing around 500 psi. I just want to know what kind of pump can be installed into a system with that kind of pressure and hold up. I'm not looking for size just type.

[maple flats]

A pump does not care what pressure it is in as long as both inlet and outlet pressure are the same. They only question is whether the pump housing is able to handle that pressure. To recirculate from high to feed pressure you would not even need a pump, just a bypass going back to the feed pump if my understanding of fluid dynamics is correct (but my experience is in irrigation and hydronic heating, not RO's).

[Bucket Head]

Labman,

There are some good posts in this section on homemade RO's that cover the recirculation pumps. They have photo's also, although I can not remember what psi they were talking about. I'm no pump expert, but a recirc. pump in an RO has to have internals that are capable of withstanding the pressure that the high pressure pump is generating. It would have to have seals that would hold up to the 500psi you mentioned. The gpm ratings are going to be different, but the pressure limits would have to be the same.

Steve

[Teuchtar]

For my home made RO, I use a 1.5 HP Dayton 4TE68 for recirc. It has a low head, high flow curve, making it suitable for recirc. The Dow membranes manual specifies the max flow through the concentrate side. They give a corresponding pressure drop (something like 15 psi if I remember). So I had to measure that pressure drop with a differential gauge, and set a flow restriction valve to keep it within limit. A differential gauge is just pressure gauge with two ports to measure the difference in psi in two pipes. Once that pressure is set, the valve never needs to be adjusted.

My feed pump is a procon powered by 3/4 hp motor. It takes the pressure up to 150 psi.

Then the HP pump is a 1.5 Hp gould multi-stage, and it takes the pressure up to 285 psi approx. Its a low flow, high head pump. (note the distinction to recirc pump)

So the recirc pump inlet is at 285, and its discharge is at 315, then throttled across the control valve to 300 psi which is the inlet to the membranes. I think I previously mentioned concerns with seal wear at that pressure in an earlier post, but so-far after 3 seasons I've not had any trouble with the recirc pump running at this pressure. The dayton pump housing is cast stainless and very robust, so hopefully no worries.

Use of a recirc pump is an efficient way to get high flow across the concentrate side. Make sure your membrane concentrate goes back to the recirc pump inlet. Your needle valve and concentrate extraction tee get plumbed in ahead of the pump suction, and prior to the entry of raw sap from the feed pump. I say this because last year I had mine reversed, so my concentrate ended up being diluted with raw sap, which I corrected this year.

If you don't have a recirc pump, then the membranes will foul more quickly. You could avoid recirc pump entirely as Maple flats noted. But in that case you either put up with quicker fouling, or increase your feed pump capacity to make up the flow. But the second option means you would pump maybe five times the flow through the feed pump and HP pump, meaning that those pumps would both need to grow in size. Costly to purchase, lotsa amps and very costly to run (3.75 HP versus 12 HP) .

This season I tweaked the controls (the balance of concentrate to permeate flows) so that I can process a tank load of raw sap while I sleep overnight, and boil in the evaporator for 3 hours the following morning. I can process 3 tankloads in a couple days, then wash the membranes back to full output again. Maybe I could process faster, but this way allows me to get some sleep.

Overall, I don't really think the precise value of recirc flow is terribly important, provided you don't exceed the guidelines and limits published by Dow.

For folks not familiar with pump curves, thats the manufacturers way of stating pressure rise across the pump ( discharge pressure minus suction pressure ). In RO-recirc duty, both those pressures are elevated to the pressure corresponding to the HP pump discharge (in my case add 285 psi, in labmans case, 500). Rarely will the pump selection guidelines make any reference whatsoever to this elevated pressure, only the pump head rise (30 psi in my case) so you're kinda on your own. The good manufacturers have a live help line you can call and get access to their 'applications engineer'.

My Dayton pump has thick cast pump housing. Some pumps have a rather thin sheet metal housing that I thought would deform under the high pressure. Some pumps will state the absolute maximum pressure the pump should be subjected to, so this could be your guideline.