ReverseOsmosisLabReport

3 March, 2015

TO: Professor A. Drews, PhD

FROM: Group B4: Brandon Sanchez, Saman Hadavand, Janet Mok, Liliana Busanez

SUBJECT: Reverse Osmosis, Final Report

Attached is the final report and our recommendations concerning the Reverse Osmosis Lab. This

report includes a study on the effect of flow rate, pressure, and change in concentration of the feed on

the permeate flow. Whereby changing parameters such as pressure and flow rate enables change of

concentration for the feed/retentate, ROM1 and ROM2 (as well as ROM3) in series can be tested for

recovery technology in optimized design and parameters. We hope this report will satisfy the desired

expectations. If you have any questions or concerns, please contact us.

Sincerely,

Group B-4

Janet Mok: Theory and Background

Saman Hadavand: Results and Discussion

Liliana Busanez: Letter of Transmittal, Abstract, Introduction, Methods, Conclusion

Brandon Sanchez: Presentation and Tech Memo

2

Reverse Osmosis Janet Mok, Saman Hadavand, Liliana Busanez, Brandon Sanchez

The goal of the experiment was to understand the characteristics and design of a Reverse Osmosis Membrane (ROM), and in series test the system which included three membranes: ROM1, ROM2, and ROM3. Investigation on the effect of pressure and feed concentration of reverse membrane water purifying system included applying a single, double, and triple membrane system where a rejection coefficient for NaCl was found. The rejection coefficient found is 0.88, compared to the manufacturer's rejection coefficient of 0.96. The performance of the system was characterized by the flux of water, where the water permeability was 0.245 (g/s-psi-m²). The membrane’s salt mass transfer coefficient was found to be 15.249 m/s.

Advisor: Professor Aaron Drews

3 March, 2015

1

1 Introduction

Reverse Osmosis Membrane (ROM) technology is used to treat industrial wastewater or to treat

contaminated water for processes that require high-quality purified water. Semiconductor processing or

biochemical applications use reverse osmosis to optimize system performances and reduce quantity of

dissolved solids in solution.1 Other Pressure driven membrane filtration systems include microfiltration,

ultrafiltration, and nanofiltration. These systems’ application depend on pores size, and charge of solutes;

furthermore, RO membranes exclude particles such as salt ions, organics compounds, etc.1 Reverse

Osmosis as a feasible process that is implemented for a variety of solute separation techniques including

nanofiltration (NF), where separation characteristics between these two technologies are referred to as NF

membranes and are being used commercially.2 For nanofiltration applications, membranes also usually

have good rejections of organic compounds with molecular weights above 200 to 500 g/mol.2 The most

important membranes are composite membranes made by interfacial polymerization; thin film composite

of aromatic polypiperazine is an example of a widely-used nanofiltration membrane for water treatment.2

RO and nanofiltration applications include the treatment of organic containing wastewater from

electroplating and metal finishing, wood pulping to food processing industries, as well as municipal and

radioactive wastewater. Treatment of such operations utilize reverse osmosis to reject particles from

contaminated water and to reuse product water.1

Early industrial development for desalination purposes, via dual role of membrane support and

pressure applied with process pump systems, have been applied since 1961 by industrial firms developing

potential designs studying membrane modification, and feed water additives in favorable economic

projections for seawater.3 In reverse osmosis salt water is forced against membranes under high pressure

where fresh water passes through. The performance of RO membranes is the usually the measurement of

water flux and solute (NaCl) rejection for the membrane, which indicate the suitability of the membrane

for the application.1 To ensure good performance, membrane type, flow control, feed water quality,

temperature and pressure are factors that enable maximising output of water.3

2

2 Background & Theory

Osmosis is a process in which a weaker saline solution tends to migrate to a stronger saline

solution, which means reverse osmosis (RO) is essentially the process of osmosis in reverse. A RO

membrane is a semi-permeable membrane that allows the passage of water molecules, but not the

majority of dissolved salts.4 The water needs to be pushed through the reverse osmosis membrane by

applying a pressure that is greater than the osmotic pressure in order to desalinate the water in the process,

allowing pure water through while holding back a majority of the contaminants.4 Figure 1 shows a

diagram of the Reverse Osmosis process when pressure is applied on one end, the water molecules are

pushed through the semi-permeable membrane while the contaminants are not allowed through.

Figure 1: Outline of Reverse Osmosis process showing how when pressure is applied to the semi-permeable

membrane, mostly only water molecules are pushed through

Reverse Osmosis works by using a high pressure pump to increase the pressure on the salt side of

the RO and force the water across the semi-permeable RO membrane, leaving almost all of the dissolved

salts behind in the reject stream. The more concentrated the feed water, the more pressure is required to

overcome the osmotic pressure.4 The desalinated water is called the permeate or product water, and the

stream that carries the concentrated contaminants that did not pass through the RO membrane is called the

retentate or reject stream. The retentate stream goes to drain or can be fed back into the feed water supply

3

in some circumstances to be recycled through the RO system to save water. This process is shown in

Figure 2 below.

Figure 2: Diagram of a Reverse Osmosis process with labeled streams

Since the flow rates and the concentrations of the permeate and retentate were seen to remain

constant, only the feed concentration varied with time.5 The two mass conservation equations that were

then used for the feed reservoir were:

Overall Mass Balance:

𝑑!𝑑𝑡= 𝑄! − 𝑄! (1)

Salt Mass Balance:

𝑑𝑑𝑡(𝑉𝐶𝑓) = 𝑄𝑟𝐶𝑟 −𝑄𝑓𝐶𝑓 (2)

where V is the volume of the solution in the reservoir, Qr is the volumetric flow rate of the retentate being

recycled, and Qf is the outlet flow rate, which is also the feed flow rate to the RO membrane module. Cr

and Cf are the salt concentrations in the streams Qr and Qf , respectively.

In regards to the RO membrane at quasi-steady state, since all variable are assumed to be constant

except for Cf 5:

Overall Mass Balance:

𝑄𝑓 = 𝑄𝑟 +𝑄𝑝 (3)

Salt Mass Balance:

𝑄𝑓𝐶𝑓 = 𝑄𝑟𝐶𝑟 +𝑄𝑝𝐶𝑝 (4)

where Cp is the salt concentration in the permeate.

4

The rejection coefficient is expressed as the following:

𝑟 = 1− 𝐶𝑝

𝐶𝑓 (5)

The closer the rejection coefficient is to 1, the more the permeability percentage decreased. The closer it

is to 0, the more the permeability percentage increased. RO membranes have a higher rejection coefficient

after longer periods of use, which shows the stabilized rejection.4

The water flux across the RO membrane is a pressure-driven flow, and thus the water flux can be

defined as5:

𝐽𝑤 = 𝐴𝑤(∆𝑃− ∆𝜋) (6)

where ∆P is the transmembrane hydraulic pressure difference, ∆� is the osmotic pressure difference, and

Aw is the membrane water permeability. Aw was found by removing the osmotic pressure difference from

the equation and running the system with fresh water. Aw is found by this equation:

𝐴𝑤 = 𝐽𝑤/∆𝑃 (7)

The following equation was then used to solve the water flux through the membrane:

𝐽𝑤 = (𝑄𝑝𝐶𝑤)/𝑆𝑎 (8)

where Qp is the permeate flow rate, Cw is the density of water, and Sa is the active surface area of the

membrane obtained from the manufacturer.

The salt transport across the membrane is predominantly due to diffusion, and the flux of salt can

be based on the concentration gradient:

𝐽𝑠 = 𝐾𝑠(𝐶𝑓 − 𝐶𝑝) (9)

where Ks is the salt mass transfer coefficient related to the salt permeability in the membrane. This model

ignores the axial gradients and is applied to average values along the membrane.5

To determine the salt concentration in the permeate, Cp, the salt mass flux and the total permeate

flow rate was used to determine the following equation:

𝐶𝑝 = (𝐽𝑠𝑆𝑎)/𝑄𝑝 (10)

5

where Sa is the membrane surface area. It is then assumed that the permeate is a very dilute solution. As a

dilution solution, the salt passage rate, Js , can then be evaluated by the equation:

𝐶𝑝 = (𝐽𝑠𝐶𝑤)/𝐽𝑤 (11)

Finally, by rearranging equation (6) and writing it as 𝛥𝑃− (𝐽𝑤/𝐴𝑤) = 𝜑(𝐶𝑓 − 𝐶𝑝), the

osmotic pressure can be expressed as5:

𝛥𝜋 = 𝜑(𝐶𝑓 − 𝐶𝑝) (12)

3 Methods

The valve to the pump that equalizes the pressure of the inlet and outlet pressure pumps is

initially open, but closed when operating the pump. Also, f water pump for the Reverse Osmosis is dry

then the inside chamber should be filled using salt water reservoir water, and the bolt loosened then

adjusted into the pump so that it is closed tight so the pressure inside the inlet pipe is adjusted with an

appropriate water level above the inlet pipe. For the piping system the valves are slightly opened so that

liquid can flow through the system. To fill the reservoir with fresh water use the valve located against the

wall. The volume increments are noted on the reservoir for calculating the concentration.

The conductivity meter will measure the conductivity in Siemens per cm, or using the option to

change units to PPM which is done with a reference line. Constructing a calibration curve, from the

measured conductivity and PPM values using 2 grams of salt from initially 30 mL water, then diluted by

20 mL of water where conductivity is checked. The known range of PPM values from this serial dilutions

are used to convert from conductivity measurements to PPM for trial analysis. For a single membrane set-

up, the three streams include the inlet that carries the water from the feed into ROM1, second stream is

the first retentate or in a 2 membrane system the second inlet. This stream can be recycled into the

primary reservoir for a non-steady state solution. The third stream is the permeate or the “clean water”

from the RO. This will be at lower pressure and flow rate indicating less permeate than retentate, and is

drained into a clean beaker. As for the 2 or 3 ROM system, a second or third permeate, and retentate out

of the last membrane will result as a consequence. A dual membrane system is illustrated in figure 3.

6

Figure 3: Diagram of RO Membrane System showing a two membrane system in series.

4 Results and Discussion

Before starting the reverse osmosis lab, a calibration curve was generated to test the accuracy of

the measuring device by graphing a relationship between a range of concentrations of salt in a water

solvent and its conductivity measurements. Figure 4 displays the calibration curve. The equation of the

conductivity versus concentration is expressed as y= 0.0186x + 36209. This equation is useful when

wanting to convert known values of conductivity or concentration. Conversions of conductivity and

concentration were necessary in order to perform calculations later in the lab. Its useful to note that the

higher of concentration of salt in the solutions, the higher the conductivity. There could have been error in

the readings due to the mixing not being completely uniform.

7

Figure 4: Calibration curve using salt and water dilutions (PPM) for conductivity measurement

For the first part of the experiment, the water permeability (Aw) of the system was found to be

245(g/s-psi-m^2). The water permeability was found by generating a graph that showed the relationship

between a range of pressures (dP) and the water flux (Jw) through membrane 1. Figure 5 illustrates this

graph. The water flux (Jw) was calculated using equation 8 and its values are illustrated in Table-A1.

Variables that were incorporated are the surface area of the membrane (Sa) which stayed constant at

0.096m^2 and the density of water Cw= 0.988 g/ml. Ideally fresh water was suppose to be ran through the

system because water permeability is being analyzed and not the purification of water. There was error in

the system because the initial feed was not fresh or clean. This concentration of impurities in the water

could have interfered with the flow rates and pressures through the membrane giving inaccurate results.

8

Figure 5: Water Flux (Jw) vs Change in Pressure (dP); Slope of Water Permeability (Aw)

A one-membrane system was used in the next part of the experiment. The salt rejection

coefficient, r, was determined for this membrane using equation 5. The average salt rejection coefficient

was determined to be 0.879, which was lower than the manufacturers given value of 0.96. This could

have been due to damage in the laboratory equipment, minimal trials, and inaccurate recordings. The

values of the salt rejection coefficients for each trial are illustrated in Table A2. Also from the one

membrane system, the salt mass transfer coefficient, Ks, was found by generating a graph that showed the

relationship between the flux of salt across membrane 1 (Js) and the concentration difference of the

permeate and feed (Cf-Cp). The value of Ks was determined to be 15.248m/s. As analyzed in the tables,

the higher the dP, the more permeate exited from the product stream. Also the same concept works vice-

versa; the lower the dP the less permeate will exit. Pressure is a crucial factor for the salt mass transfer

coefficient. Pressure has a higher magnitude effect on the system versus the flux because Cf-Cp is smaller

in magnitude, thus has more sensitive changes. The graph is illustrated in Figure 6. This directly affects

the rejection coefficients because r= 1-(Cp/Cf). The higher the permeate, the smaller the rejection which

is dependent on pressure. Finally for the one membrane system, the osmotic pressure difference, dπ, was

determined by equation 12. The constant, Ψ , was used to calculate dπ and was determined by the plot

shown in figure 6. The value for constant, Ψ, was determined to be -0.06 ((psi-m^3)/g). The osmotic

pressure difference values are illustrated in Table A3.

9

Figure 6: Flux of Salt (Js) vs Change of Concentration; Slope of Salt Mass Transfer coefficient Ks

Figure 7: Slope of osmotic pressure constant Ψ

A two-membrane system was used next in the experiment. The rejection coefficient, r, was

determined in both system as illustrated in Table A4. Because of errors made in the lab, the system did

not run to well. This could have been to the damaged device or the broken flow meter that did not allow

accurate adjustments. The second membrane had much lower rejection than the first membrane due to

retentate being its feed. For a few trials no permeate would exit, this was due to a low pressure and flow

rate. The system was simply too weak. A three-membrane system was also observed. The third membrane

had an even smaller rejection coefficient than the second membrane observed. This was due to the even

more concentrated feed, its a similar trend. The more membranes that are added, the higher the

concentration in the feed, the lower the rejection, the less permeate will exit. The rejection coefficient for

10

the third membrane is illustrated in Table A5. Finally a single membrane system with a recycle stream

was observed. The rejection coefficient for this system is illustrated in Table A6. The feed reservoir was

so large that the results stayed relatively constant. The prediction was that the rejection coefficient would

decrease as time increased because the feed would become saltier thus increases Cf. More trials were

needed to prove this point.

5 Conclusion

The concept proved in the lab was the higher the water flux the higher the permeate. Higher

pressure enabled a higher permeate flow rate. This is illustrated in figure 6. The goal of the lab was to get

a high quantity of permeate to exit, not necessarily the quality of the permeate stream. Adding second and

third membranes to the system allowed permeate to flow out, but due to systematic errors in lab setup, the

permeate did not flow out well in the second and third membranes. The rejection coefficient for the first

membrane was the highest and decreased heavily in the second and third membranes. This was due to the

retentate being higher in salt concentration, thus creating less permeate. These values are illustrated in

tables 2,4,5, and 6. The salt mass transfer coefficient (Ks) was generated from figure 6; the pressure and

permeate streams were the most sensitive variables. Lastly, the rejection of salt concentration for the

Reverse Osmosis setup was relatively high in the membrane system indicating good performance of the

system in removing salt.

11

6 References

[1] J. Kucera. Properly Apply Reverse Osmosis. Chemical Engineering Progress. 54, 1997.

[2] EET Corporation and Williams Engineering Services Company, Inc. A Brief Review of Reverse

Osmosis Membrane Technology. 2003

[3] J. Glater. Desalination. 297-309. 1998

[4] What is Reverse Osmosis?

http://puretecwater.com/what-is-reverse-osmosis.html#salt-rejection (Accessed March 1, 2015)

[5] Chau, P.C. Reverse Osmosis with Retentate Recycle. UCSD. 1-6

https://ted.ucsd.edu/bbcswebdav/pid-520234-dt-content-rid-

10306897_1/courses/CENG176A_WI15_Zhang/Lab%20attachments%282%29/RO-recycle.pdf

(Accessed March 1, 2015)

7 Appendices

Table A1: Water Flux ( Jw) vs Pressure Change (dP)

Trials   Qp  (ml/s) dP  (psi) Jw  (g/s-­‐m^2)

1 3.5 50 36.20412298

2 4.93 75 50.99609322

3 4.17 94.3 43.13462652

4 3 50 31.03210541

5 4.17 75 43.13462652

6 5.83 94.3 60.30572484

7 3.83 94.3 39.61765457

8 3.67 94.3 37.96260895

12

Table A2: Flux of salt (Js), Rejection Coefficient of a Single Membrane

Trial Cf  (ppm) Cp  (ppm) Cf-­‐Cp  (ppm)

Qp  (ml/s)

Jw  (g/s-­‐m^2)

Js (g/s-m^2)

Rejection Coefficient

2 3546 325 3221 1.9 19.64776062

3 3546 498 3048 0.84 8.686378378

6398.318838

0.908347434

4 4666 608 4058 3.1 32.05687259

4334.485403

0.859560068

Average         19529.63781

0.869695671

        0.879201057

Table A3: Osmotic Pressure Difference

Trial Cf-­‐Cp  (ppm) dπ  (psi)

2 3221 -­‐193.26

3 3048 -­‐182.88

4 4058 -­‐243.48

13

Table A4: Rejection Coefficients of a Two System Membrane

Trial   dP  (psi) Cf(ppm) Cp1(ppm)

Cr(ppm) Cp2(ppm) rejection  coefficient  1

rejection  coefficient  2  

1 25 4666 363 5218 1402 0.922203172

0.73131468

2 25 4560 265 4820 0 0.941885965

1

3 25 4560 245 4786 2084 0.94627193

0.56456331

4 26 4570 185 4712 0 0.9595186 1

5 18 9022 960 9130 0 0.893593438

1

Table A5: Rejection Coefficient of a Three System Membrane

Trial Cf1 (ppm)

Cf3 (ppm)

Cp1(ppm)

Cp2 (ppm)

Cp3 (ppm)

Cr (ppm)

rejection coefficient 1

rejection coefficient 3

1 9135 7545 3567 4717 4080 6508 0.60952381

0.459244533

Table A6: Rejection Coefficient of a One System Recycled System

Trial 1 Cf (ppm) Cp1 (ppm) rejection coefficient 1

9152 2704 0.704545455

14

3 March, 2015

TO: Professor A. Drews, PhD

FROM: Group B-4: Brandon Sanchez, Saman Hadavand, Janet Mok, Liliana Busanez

SUBJECT: Technical memorandum regarding Cooling Tower

The purpose of the cooling tower is to apply simultaneous mass & heat transfer in order to construct a

design model for the cooling tower. We will configure the temperature set point of the cooling tower

water reservoir and let the system run until it reaches steady state. After it reaches steady state, we will

record parameters such as the temperatures of the: water inlet and outlet, inlet and outlet air dry bulb, inlet

air wet bulb, water flow rate and humidity at the bottom and top of the tower. Furthermore, we will adjust

the water and air flow rate and water inlet temperature in order to further analyze the effects of reservoir

temperature on the cooling tower. We expect to see the enthalpy of the humid air to increase as the mass

of water to air ratio increases. We also expect an increasing mass flux if the concentration difference

between the absolute humidity and mass ratio at the air-water interface is increasing. If you have any

concerns, please contact us.

Sincerely,

Brandon Sanchez