Experiment & Results

Home
Introduction
Theory & Design
Experiment & Results
Conclusion
References
Members
Links
Presentations

Experimental

    This study contained four main parts.  First, the conductivity meter was calibrated. By measuring and plotting the conductivities of solutions against known concentrations of NaCl, a calibration curve was prepared. The calibration curve was used to change measurement read in milliSiemens (mS) to molar concentrations of NaCl.  After a calibration curve is obtained the RO membrane was analyzed.

    To test the performance of the reverse osmosis membrane, the first set of experiments were run at a constant concentration of 31 mmol NaCl per liter of water.  Solid NaCl was weighted and dissolved in a 19.5 L container with tap water at 76F.  A conductivity reading of the solution is taken before the run to ensure that the solution is well mixed.  If the solution was well mixed, a reading of conductivity of 22.4 mS was obtained. At the beginning of each run, the pressure is adjusted on the retentate side of the membrane to a designated level using a needle valve.  Pressures were tested between 30 and 120 psi. The pressures used are above the osmotic pressure of the solution, which is 7.84 psi. During each run 5-7 samples of the permeate were collected at 1.5 minute intervals, analyzed for purity with the conductivity meter, and recorded.  The data collected from this experiment will give an optimal operating pressure for our membrane.

    The next set of experiments was preformed with pressure held constant at 70 psi and concentration was varied between 15-60 mmol NaCl per liter.  Sampling and analysis was done as the same the previous set of experiments.  The data collected in this experiment demonstrated a relationship between the permeate concentration and inlet concentration, as the inlet concentration was varied.

    The last experiment tests the effectiveness of a UF filter in our system.  This test is done by dosing a container of 19.5 L of water with 25 g of salt to create a 31 mmol NaCl per L water.  The conductivity of the solution is then taken.  The measured value should be 22.4 mS if the solution is well mixed.  The solution was run through a UF filter and retested with the conductivity meter.  This experiment takes roughly 1 to 1.5 hours because of the extreme pressure drop across the filter.  A larger pump may also aid in increasing the flow across the filter.   

Results

A. Calibration Curve

    A line is fit to the graph in Figure 1 so that conductivity that are read can be converted to concentration of NaCl.  The line has a slope of 0.0014 so the equation for converting the reading (R[=]mS) to concentration (C[=]mol/L) is C = 0.0014 R.

Figure 1: Concentration Calibration Curve

 

B. Constant Concentration Experiments

    A linear relationship is found between the pressure and permeate flowrate, as seen in Figure 2. As a pressure above the osmotic pressure of the solution is applied to the feed, the flux across the membrane varies in direct and linear relationship with the amount of pressure applied. As the pressure increases, more water is pushed through the membrane, resulting in greater flux. Every membrane has a maximum operating pressure, above which the membrane will foul or blow out. Until the maximum pressure is reached the flux will continue to increase in proportion to the amount of force applied.

Figure 2: Permeate Flowrate vs. Pressure

    In the constant inlet feed concentration experiments the conductivity readings of the permeate were measured, converted to concentrations of NaCl and plotted as a function of pressure in Figure 3.

Figure 3: Permeate Concentration vs. Pressure at Constant Inlet Concentration

Pressure was also plotted against salt rejection percentage, as seen in Figure 4.

Figure 4: Salt Rejection % vs. Pressure at Constant Inlet Concentration

    Since the RO membranes are imperfect barriers to dissolved salts, there will always be some salt passage through the membrane. As feed pressure is increased, the salt passage is overcome as water is pushed through the membrane at a faster rate than salt can be transported. However, there is an upper limit to the amount of salt that can be rejected via increasing feedwater pressure. Above that limit, some salt flow remains coupled with the water flowing through the membrane, which decreases the permeate purity. As can be seen on the graphs, the membrane used has an optimal pressure of about 62 psi for the specific solution concentration of 31mmol NaCl per liter, and an average conductivity of 22.4 mS. The permeate concentration at this pressure is 3.2 mmol NaCl per liter, and salt rejection is 89.6%. Above an applied pressure of 62 psi, concentration of the permeate increases and salt rejection decreases.

    Comparing the obtained optimal salt rejection percentage to the 99.4% quoted by Koch Membrane Systems, it is seen that the salt rejection obtained is 9.8% less than the quoted value.  There are a few reasons for this discrepancy.  First, this membrane is roughly 3 years old. The 99.4% is a number quoted for a new membrane. As an RO unit gets older, its pores will slightly increase in size because of the extended periods of pressure applied to them.  They also build up salts that cannot be removed even with chemical cleaning.  This is why membranes are replaced every 4 to 5 years.  Another reason that the membrane was not be meeting the specifications of the optimal salt rejection percentage is that the process did not reach its steady state. In industry, membranes are run constantly for months, but because we are limited by water storage size, the experiment preformed only allow for runs of 10-12 minutes.  If this experiment is done next year, it is our suggestion that a larger tank should be obtained for a longer testing periods.

C. Constant Pressure Experiments

    In Figure 5 the variable inlet salt concentration is plotted against permeate flowrate at a constant pressure of 70 psi. The relationship seen is inversely proportional. Since osmotic pressure is a function of concentration of salts in the feedwater, an increase in the inlet salt concentration increases the osmotic pressure. Therefore the pressure needed to overcome the osmotic pressure must also increase. If feed pressure remains constant, higher salt concentration results in lower membrane water flux. The increasing osmotic pressure offsets the feedwater driving pressure.

Figure 5: Flowrate vs. Inlet Concentration

    As the water flux declines, the purity of the permeate is decreased. Figures 6 and Figure 7 show the relationships of the permeate salt concentration and salt rejection percentage as a function of inlet concentration, respectively. A high concentration of salt results in high molecular forces within the solution. Molecular forces must be overcome by applied pressure before water molecules are able to separate and pass thorough the membrane. Since the applied pressure remains constant, it is difficult to overcome the forces resulting from a high feed concentration of salt solution. Salt becomes coupled with the water and passes to the permeate side. Therefore, the salt rejection increases initially and reaches a maximum value of 89.2% at 27 mmol NaCl per liter of inlet water. After the maximum value has been reached, the percentage of salt rejection declines.

Figure 6: Permeate Concentration vs. Inlet Concentration

Figure 7: Salt Rejection % vs. Inlet Concentration

D. Ultra Filtration

    The experiment described in the Experimental Methods section was repeated several times for several NaCl concentrations, with a pretreatment UF system installed.  Each time the concentration of the solution after it had been run through the UF membrane measured the same conductivity as the inlet feed. This proves the hypothesis that the UF membrane would have no effect on the overall outcome of the process, when only salts in tap water are present. Figure 8 shows the same relationship between the inlet concentration and the concentration of the permeate was obtained as before, when the UF was absent. UF membranes are designed for trapping particle much larger than salt molecules, for example, bacteria and large organics.  If the solution could some how be dosed with bacteria and tested for bacteria count on the outlet, the true effectiveness of the UF could be tested.

Figure 8: Permeate Concentration vs. Inlet Concentration With the UF Pretreatment

Home | Introduction | Theory & Design | Experiment & Results | Conclusion | References | Members | Links | Presentations

 
2004

For problems or questions regarding this web contact Natalie Tran.
 Last updated: 06/08/04