By Ted Kuepper, Robert Lovo & Mark Silbernagel

Summary: Part 1 of this series described the open-loop recirculation flow pattern and how it creates a no-waste membrane system by diluting and reusing the concentrate stream. Here we discuss the system’s limitations, as well as offer test results showing the advantages that make the design a viable option for reduced water waste in POU and POE applications.

Of course, all water treatment designs have limitations and a membrane system that doesn’t waste water is certainly no exception. The first limitation is the fact that the design requires a pump in order to create the recirculation flow pattern and to boost pressure. However, for systems larger than small residential units this generally does not result in an increase of equipment since a pressure boost pump is necessary for many commercial units already. The second limitation of this no-waste system design is that the quantity of water purified for the highest quality applications (drinking, cooking, etc.) must be about the same as (or less than) the quantity of water used for lower quality applications (toilets, sink faucets, etc.). Therefore, for residential and commercial water softening, a no-waste system can generally be created by softening only the hot water used in a building and can be used to soften both hot and cold water depending upon the amount of total water used. On the other hand, drinking water systems have virtually no limitations because of the relatively small amount of purified drinking, cooking and ice making water usually required at a location compared to the total quantity of water used.

Figure 2 (Figure 1 is in Part 1 in WC&P’s July issue) shows the design of a no-waste drinking water system based on the open-loop recirculation flow pattern. Because product water is delivered at municipal line pressure, this equipment can be placed where municipal water enters a building. Plus, relatively small diameter tubing (3/8-to-1/4 inch) can be routed to water fixtures wherever high quality water is desired. In other words, the consistent and relatively high pressure of the design’s product water flow eliminates the need to place equipment under each sink at locations where product water is needed. Consequently, this feature allows a single water treatment system to be placed in one location to produce drinking water for distribution throughout a building. In addition, the no-waste design is flexible enough to accommodate a conventional undersink configuration.

Figure 3 shows the design of a combination no-waste water softening and drinking water system. Decreasing the amount of water wasted with membrane-based systems will allow use of membranes for water softening applications. This configuration would be especially useful for commercial locations, such as restaurants, since it produces softened hot water necessary for dishwashing as well as demineralized water suitable for drinking, cooking and ice making.

System test configuration
A conventional reverse osmosis (RO) drinking water system installed in a residential home was tested to document its performance characteristics. The unit was a typical undersink type of installation where the feed to the RO element was connected to the cold water supply line of the kitchen faucet and the RO element concentrate line was connected to the kitchen sink drain. The RO membrane used in this system was a conventional cellulose triacetate or CTA-type of RO membrane that was chlorine tolerant. The system used a nominal 5-micron particulate filter as pretreatment. It was purchased and installed in California and included an automatic shutoff valve designed to stop water flow to the RO element. This type of valve is designed to interrupt feed flow to the membrane whenever pressure inside the product water tank rises to about two-thirds of municipal line pressure. The intent of this valve is to save water by stopping the flow of waste to drain when the product water tank is full. During the test, municipal water pressure was measured to be 62 pounds per square inch (psi). Therefore, the cut-off pressure for the RO water storage tank was approximately 41 psi.

Test results
Quantities of demineralized water created by the conventional RO drinking water system and delivered to a conventional air pre-charged product water storage tank, as well as brine produced during the same period of time, were measured to determine the system’s actual recovery. As noted earlier, in a conventional RO system, demineralized product water must initially overcome the pre-charge air pressure in the tank and overcome an ever-increasing backpressure as water fills the storage tank. Consequently, there’s an ever-decreasing driving force across a membrane in a conventional RO system as the storage tank fills (unlike a no-waste system where its driving force remains essentially constant). At the beginning of the test, the air pre-charge of the empty four-gallon water storage tank was measured to be 5 psi.

The results of this test are presented in Table 1 and show that 135.01 liters (35.7 gallons) of water were sent to drain during the production of 12.66 liters (3.34 gallons) of demineralized water with the RO system. Therefore, an average of 10.66 liters of water were wasted for each liter of RO demineralized water produced. In addition, though the recovery of the RO system started at 16 percent, each hour of operation reduced the recovery. By the time the shutoff valve stopped system operation, recovery was reduced to only 2.5 percent. Once the product water storage tank was filled, the system operated with a recovery of about 2.5 percent thereafter if modest amounts of water were removed from the product water storage tank at a time (for example, to fill a single glass of water).

An additional test was conducted to further document recovery of a residential RO system when used in a manner consistent with typical residential use. The results of this test are shown in Table 2 and are thought to indicate actual performance of an RO system installed in a residential location. As shown, when 1.0 liter of water was withdrawn from the product water storage tank, the recovery remained in the 2.5-to-4 percent range for the two hours necessary to refill the storage tank, thus creating an average recovery of 3.5 percent. The 1.0 liter of product water removed from the tank during this test is thought to be representative of the quantity of water routinely removed at a single time in residential locations. Because of the declining recovery characteristic of conventional point-of-use (POU) RO systems, it’s believed these systems typically operate at an actual average recovery of only 3-to-4 percent and consequently normally waste over 27 liters of water for each 1.0 liter of demineralized water produced throughout their operating life. This test clearly documented the effect that a conventional air pre-charged product water storage tank has on recovery of a typical RO drinking water system.

This conventional system created product water with an average total dissolved solids (TDS) measurement of 60 milligrams per liter (mg/L) while operating with a municipal feed source of 598 mg/L TDS (on average). Therefore, salt rejection for this RO membrane was 90 percent.

No-waste RO system
A no-waste RO drinking water system was installed in a residential home, shown in Figure 4, and tested to document its performance characteristics. These results are presented in Figure 5. As can be seen, the no-waste system created product water with an average TDS measurement of 39 mg/L while operating with the same municipal feed source. Therefore, salt rejection for the CTA-type of RO membrane used was 93 percent. (Please note, a CTA-type of membrane was also used in the conventional system tested.)

There was no effect on the TDS measurements taken on the cold water supply inside the house due to the reuse feature of the no-waste design (compared to the municipal water feed). This is because of the relatively large dilution factor for the RO concentrate caused by the total water used at the residential location. Measurements were taken just downstream of the no-waste drinking water system at a kitchen faucet, which was the closest water fixture to the system’s installation. The importance of the result presented in Figure 5 lies in the fact that the performance achieved with a no-waste system is identical to that expected of a conventional RO drinking water system that wastes a significant quantity of water during its normal operation.

Conventional RO drinking water systems waste a significant amount of water even when equipped with a water-saving device, such as an automatic shutoff valve. During the reported test, a conventional RO system wasted 135.01 liters (35.7 gallons) of water during the production of 12.66 liters (3.34 gallons) of product water with a membrane recovery that declined from 16 percent to 2.5 percent.

The no-waste water treatment system design described here should encourage development of POU and point-of-entry (POE) equipment that reuse water at a location through the utilization of the open-loop recirculation flow pattern. The result of this design is elimination of wastewater (with an apparent process recovery of 100 percent) while demineralizing municipal water by means of an RO membrane. In addition, the unique product water bladder incorporated in the design sets the recovery rate of the system while providing a consistent flow of RO product water at exactly municipal line pressure (without needing an additional re-pressurization pump).

The no-waste design corrects the two primary disadvantages characteristic of current POU equipment—discharge of a relatively large percentage of wastewater to drain (by RO drinking water units) and an increase in the salinity of domestic wastewater (by ion exchange water softeners). This makes a membrane-based water softener a realistic alternative for several POU applications.

About the authors
Ted Kuepper, Robert Lovo and Mark Silbernagel are members of Pacific Research Group of Ventura, Calif., and jointly developed the Zero-Waste Membrane System described in this article. Kuepper can be reached at (805) 985-3057, (805) 985-3688 (fax) or email:


Comments are closed.