Does the Permeate Pump Save Water?

by Gene Franks


The permeate pump has been a popular accessory for domestic undersink reverse osmosis (RO)  units for a number of years. It is a non-electric, water-powered device that works by isolating the RO membrane from back pressure from the storage tank.

Standard undersink RO units produce water into a small pressurized storage tank.  As the tank fills, the RO unit has to push against increasing back pressure from the storage tank. The greater the back pressure, the more inefficient the RO unit becomes. Standard undersink RO units that store water in a pressure tank never perform as well as the membrane rating. A 4:1 (four gallons to drain for each gallon of product water) membrane, when used on a standard undersink RO unit, actually performs much less efficiently because of back pressure from the storage tank. In informal tests that we did at Pure Water Products, actual efficiency performance of a standard RO in simulated home usage was around 7:1. The unit performed at 4:1 when the tank was empty but the efficiency got progressively worse as pressure built in the storage tank.

The purpose of the permeate pump is to allow the RO unit to produce water into an almost pressure-free chamber. The pump then pushes the stored water into the the pressurized storage tank. Water leaves the RO membrane housing in two streams: the purified water (permeate) and the reject or drain water (brine).  The pump uses the energy from the brine to push the permeate into the storage tank.

The pump manufacturer claims that the pump will save water and also improve the TDS  performance of the RO unit. This is discussed in detail, with manufacturer’s charts, in a 2011 issue of the Pure Water Occasional.

An earlier Pure Water Gazette piece reported an informal test done with a single home RO unit that looked at the question of whether the pump improved TDS rejection performance as claimed by the manufacturer.  The test showed no improvement in TDS performance with the pump as compared with the same unit without the pump.

The TDS test was done with a standard Black & White  undersink RO unit with the GRO 50/50 water saving membrane. The membrane performance rating is 1:1–one gallon of brine for one gallon of permeate.  The same unit, now a year older, was used to test the permeate-to-brine efficiency with and without the permeate pump.

We ran six trials: three with a standard unit without permeate pump and using the regular Payne shutoff valve, and three with an Aquatec ERP 500 permeate pump installed and the same Payne shutoff.


Standard RO membranes are set up with a ratio of about 4 (drain) to 1 (permeate).  With no tank to push back against, a 24 gpd membrane installed with a 250 ml/m drain line flow restrictor will run about 95 gallons of water to drain while producing 24 gallons of drinking water. This works out to around a 4:1 ratio. The GRO 50/50 membrane used in this test is a 50 gpd membrane, paired with a 150 ml/m brine restrictor, so the theoretical production is 50 gallons of permeate while sending 57 gallons to drain. Essentially a 1:1 ratio.  (To convert milliliters per minute to gallons per day, multiply by 0.38.)  Significantly, a smaller-than-usual storage tank was used to move things along faster. The tank is about a 7.5″ X 9″ petite version, which the manufacturer calls a “2 gallon” size, but which for practical purposes holds about one gallon.

The trial consisted simply of starting with a full tank, draining the tank to catch and measure the content, then allowing the RO unit to refill the tank while catching the brine for measurement. Three trials each were done with and without the pump. Results were surprisingly uniform. The table below reports the averages in percent recovery, which means the percentage of the water that goes into the RO unit that comes out as permeate.  Keep in mind that a standard 4:1 membrane setup without tank has a percent recovery ratio of about 20%; with a tank in a standard residential setting the percent recovery might be 12%.

The Test with GRO Membrane


Setup Total Permeate in fluid ounces Total Brine in fluid ounces Recovery Percentage  
GRO Membrane, without pump 114 (0.89 gal.) 160 (1.25 gal.) 41.6%
GRO Membrane, with pump 120 (1.07 gal.) 118 (0.92 gal.) 50.4 %

The conclusion is, then, that the permeate pump does indeed save water, even when running with the stingy GRO membrane. With a conventional 4:1 membrane, the saving is more pronounced. Permeate pump manufacturer Aquatec’s performance chart shows conventional membrane performance with the pump at about 12:1 when the storage tank is 60% full. Our tests show overall performance of conventional units averages about 7:1 under normal stop and go residential use, while the permeate pump unit with the same membrane maintains a consistent 4:1 throughout the fill cycle.

One final note about the GRO 50/50 membrane. Although it would appear from the 41% recovery that it is performing at less than the advertised 50% recovery rate, its performance is really quite remarkable. The 50% setup is the performance you get without a tank (as in a tankless “countertop” RO unit). You should get the essentially same performance with a permeate pump, whose job it is to nullify the back pressure from the storage tank. That is what we got in our test.

Here is a summary of simulated performance figures to put all this in perspective:

Setup   Approximate Recovery Percentage
 Standard membrane in undersink standard RO with tank  12%
 Standard membrane in tankless countertop RO  20%
 Standard membrane in undersink RO with permeate pump  20%
GRO membrane in tankless countertop RO 50%
GRO membrane in undersink RO without permeate pump 40%
GRO membrane in undersink RO with permeate pump 50%

Adding a permeate pump to a standard RO unit improves its recovery percentage by around 10%.  Replacing the standard membrane with a 50/50 GRO membrane increases recovery by about 30%.

Does this mean, then, that you  should always buy an RO unit with a permeate pump?  Not at all. There is more to be considered that water saving. The permeate pump makes the RO unit a more complicated device that can be a bit harder to install, harder to maintain and more difficult to troubleshoot if there is a problem. It also makes some noise, although the current version ERP-500 model is very quiet. Saving 8% or 10% in water consumption may not be worth it.

If saving water is the main objective, I suggest adding the GRO membrane first, and then the permeate pump.