Reverse osmosis tests

Figure 2. Schematic view of reverse osmosis test loop (I) hollow fiber membrane (2) pressure vessel (3) feed water (4) filter (5) pressure pump (6) relief valve.

This membrane demonstrated a vastly improved flux compared with the poly(piperazine isophthalamide) membrane, but its seawater salt rejection was low — in the range of 60 to 70 percent. A reverse osmosis test with a magnesium sulfate feedwater showed greater than 99 percent salt retention, however, dispelling the possibility that low sodium chloride rejections were due to defects in the polyamide barrier layer. The piperazine polyamide was soon concluded to have the following structure (see Reaction 111). 

Table 2. Effect of the Isophthaloyl Trimesoyl Chloride Ratio on the Performance of NS-300 Membranes in Reverse Osmosis Tests...

Acid Chloride Ratio Reverse Osmosis Test Results ... 

Reverse osmosis test conditions 0.5% salt concentration, 200 psi, 25 C, poly(piperazine trimesamide) membrane. 

Figure 2. Exposure of FT-30 membranes to 100 ppm chlorine in water at different pH levels. Effect on salt refection in simulated. seawater reverse osmosis tests (0) pH 1 Cn pH 5 (O) pH 8 (A) pH 12.

Acid chloride ratio8 Reverse osmosis test results ... 

Figure 5.3 Flow schematic of a high-pressure laboratory reverse osmosis test system...

Tabk 8. Performance of cellulose acetate and polybenzimidazole membranes in reverse osmosis test ... 

Generally, the salt rejections observed for these membranes in seawater reverse osmosis tests did not exceed 80 percent. This process was applied at Albany International to form composite membranes on hollow polysulfone fibers (25). Salt rejections on the hollow fiber membranes were above 98 percent at an average flux of about 1.5 gfd (2.5 L/sq m/hr) in a 12 000-hour test using 30 000 ppm seawater at 1000 psi. In other 5000-hour tests using 3500 ppm brackish water at 400 psi with addition of 100 ppm chlorine at pH 8 flux and salt rejection remained constant at 1 gfd and 98 percent respectively (Figures 8 and 12 in Reference 25 a). 

A membrane designated "Solrox" made by Sumitomo Chemical Company is closely related to the above plasma polymerized composite membranes. A 1980 report by T. Sano described the Sumitomo process (31). A support film was cast from a polyacrylonitrile copolymer containing at least 40 mole percent acrylonitrile. The support film was dried and exposed to a helium or hydrogen plasma to form a tight cross-linked surface skin on the porous polyacrylonitrile support film. Data in a U.S. Patent issued in 1979 to Sano et al showed that the unmodified support film had a water flux of 87 gfd (145 L/ sq m/hr) at 142 psi (10 kg/sq cm). After the plasma treatment a reverse osmosis test using 0.55 percent NaCl at 710 psi (4895 kPa) showed 10.5 gfd (17.5 L/sq m/hr) flux at 98.3 percent salt rejection (32). This membrane appears to fall between a conventional asymmetric membrane and a composite membrane. If the surface skin is only cross-linked, one might call it a modified asymmetric membrane. However, if the surface skin is substantially modified chemically to make it distinct from the bulk of the membrane it could be considered as a composite type. 

Reactant used to prepare Reverse Osmosis Test Data (a)... 

Properties of FT-30. The properties of FT-30 membranes have been reviewed in several publications. Therefore, only the salient features that relate to the chemistry of the barrier layer will be considered here. Reverse osmosis performance of FT-30 under seawater and brackish water test conditions was described by Cadotte et al (48) and by Larson et al (51). In commercially produced spiral-wound elements the FT-30 membrane typically gives 99.0 to 99.2 percent salt rejection at 24 gfd (40 L/sq m/hr) flux in seawater reverse osmosis tests with 3.5 percent synthetic seawater at 800 psi (5516 kPascaJJand 25°C. 

Countercurrent Reverse-Osmosis Tests. Flat-sheet 3N8 and TFC-801 membranes were characterized with and without permeate-side recirculation. Other flat-sheet membranes were not tested because of either their low ethanol-water selectivity or their tendency to degrade even in relatively dilute ethanol solutions. 

Flux decay in the reverse osmosis test has been determined and minimized using appropriate fluid dynamic regimes. Typical flux decay with DDS 800 and PA 300 are shown in Figure 5 and 6. The feed was must ultrafiltered through BMR capillary membranes. 

The six flat cast membranes were shrunk at different temperatures (from 68 to 85°C) prior to loading the membranes into the reverse osmosis test cells. This treatment adjusts the average surface pore size of each membrane so that a range of porosities could be studied. A prepressurization at a pressure of 11 720 kPa for 2 hours was used to stabilize the membranes for subsequent use at pressures of 6900 kPa or lower. (All pressures listed are gauge pressure.)... 

Most of the reverse osmosis testing was carried out at 3.5% sodium chloride concentration and 1500 psi. pressure. A number of tests were also conducted at 0.5% and 800 psi. 

The various films prepared were about 1-2 mils thick and were neither ultra thin nor prepared as asymmetric membranes. Consequently the product fluxes were low and could not easily be determined with conventional reverse osmosis test cells. Measurements were therefore carried out in a test cell first described by McKinney [8]. The test unit is uniquely suited for characterizing transport properties in dense, low flux membranes because of the small downstream dead volume and the simultaneous determination of flux and rejection in a capillary connected to the downstream receiving volume instrumented with a microelectrode to monitor down-stream salt concentration. 

---