Novel Low Fouling Nanofiltration Membranes Abstract Craig Bartels, PhD, Hydranautics, Oceanside, CA Warren Casey, PE, Hydranautics, Houston, TX Nanofiltration (NF) has become a standard process to treat mildly brackish water sources that are high in hardness, iron or organic material which forms disinfection by-products. The goal of the membrane in these processes is to selectively remove certain ions, while passing others. Designing these membranes has become more complex in recent years as regulations on the finished water have become tighter, and there is pressure to reduce operating costs. Example of such applications are the 10.5 million gallon per day (mgd) Deerfield Beach and the 40 mgd Boca Raton projects in Florida. The permeate from each of these new plants is blended with existing lime softener effluent. To meet the desired hardness of the blended water, the nanofiltration permeate hardness had to fall within a limited range, 25-80 (Deerfield) or 50-80 (Boca Raton) ppm as CaCO3. In each project the nanofiltration membrane must also reduce trihalomethane formation potential (THMFP) below 40 ppb and meet a maximum pressure requirement (90 psi for Deerfield and 80 psi for Boca Raton). In the case of Deerfield, however, there was a requirement to have high iron rejection as well. As a result, a slightly higher rejection NF membrane was required for Deerfield to achieve the iron reduction targets, while a slightly looser NF membrane was required to achieve the hardness passage requirements of the Boca Raton project. Hydranautics has developed a unique NF membrane which is made by a process that allows the rejection to be tuned to a level that can meet the specific rejection required for these new projects. For the Deerfield project the membrane was made with an average calcium chloride rejection of 86%, while for the Boca Raton project the membrane was made with 80% rejection. As expected, the permeability of the NF membrane for Deerfield is lower, 7500 gallons per day (gpd) compared to the 7800 gpd element for Boca Raton. Information will be reported on the characteristics of these unique membranes, and how they are designed for these unique applications. In addition to meeting the water quality and initial pressure requirements, these applications also have very high levels of TOC in the water, up to 20 mg/l. The high fouling potential of this water also makes it difficult for the NF membrane to maintain stable performance over the life of the membrane. The unique low fouling feature of the ESNA1-LF nanofiltration membrane makes it ideally suited for difficult organic-laden feedwaters. The ESNA1-LF membrane has operated at the Deerfield plant since 2003 and has produced product water of 0.13 ppm iron, hardness of 27.3 ppm as CaCO3, THMFP of 27 ppb, and stable operating pressure. The membranes have not required chemical cleaning during this period. The ESNA1-LF2 membrane at Boca Raton plant has been operating since November 2004. The permeate from the NF membranes has an average hardness of 75 ppm as CaCO3 and THMFP of 16 ppb. Meeting these strict water quality requirements has proven that these new low fouling membranes can be tailored, even for very large-scale plants.
Novel Low Fouling Nanofiltration Membranes NAMS 06 Chicago, Ill Craig R. Bartels, PhD Warren Casey Hydranautics Oceanside, CA
Commercial Use of Nanofiltration 37 mgd Mery sur Oise NF Plant (Desal 192 (2006) 303 314) Organic Removal TDS Passage Pesticide Removal Partial Ca Removal 9 mgd Boynton Beach NF Plant (AMTA 2005) Partial Ca Removal 7 mgd Irvine Ranch Deep Aquafer 2
New Nanofiltration Application Trends Larger Plant Sizes: 10 40 mgd Specific Permeate Hardness Target High Organic Levels = High Fouling Potential Disinfection By-Products Limits Low Pressure Operation Timing of Execution 3
New Low Fouling NF Membranes with Variable Hardness Rejection Controlled hardness removal - eliminating the need for hybrid membrane designs (LF, LF2, LF3 versions) Significantly reduces operating costs by lowering power consumption, while providing a non-aggressive low TDS permeate water Greatly reduces fouling potential caused from natural organic matter Lowest cost per gallon produced feed pressures of less than 100 psi saves energy cost 4
Membrane Surface Properties Smooth Topography ESNA1-LF Surface Surface roughness has been decreased to minimize sites for debris trapping ESPA3 Surface 5
Membrane Surface Properties: Reduced Surface Charge 10 0-10 ζ Potential -20-30 -40 ESNA1-LF ESNA1 LFC1 Surface charge similar to LFC1-50 2 3 4 5 6 7 8 9 10 ph Lower surface charge minimizes interaction with surfactants 6
ESNA1-LF Salt Rejection (Surface water @ 13 gfd, 25 C) 100.0% 97.9% 99.4% 98.5% 96.2% 96.4% 95.0% Rejection (%) 90.0% 85.0% 83.8% 89.9% 85.6% 80.0% 75.0% Na Ca Mg Fe Cl SO4 SiO2 TOC Feed Component Type 7
ESNA1-LF2: Salt Rejection (Surface Water @ 13 gfd, 25 C) 100% 97.9% 95% 93.5% 95.0% 92.5% Rejection (%) 90% 85% 83.0% 80% 77.2% 75% Na Ca Fe Cl HCO3 SO4 Ion 8
Chemistry Modification of ESNA1-LF, LF2 and LF3 to Achieve Desired Permeate Hardness 45 Hardness vs Iron Rejection 0.3 Permeate Calcium (mg/l as CaCO3) 40 35 30 25 20 15 10 5 ESNA1-LF2 ESNA1-LF ESNA1-LF3 0.25 0.2 0.15 0.1 0.05 Permeate Iron, mg/l 0 93 94 95 96 97 98 99 Membrane Nominal Calcium Rejection Calcium ph=6.5 Calcium ph=7.0 Iron ph=6.5 Iron ph=7.0 0 9
Typical Florida Feed Waters 10
Boca Raton Water Samples 11
ESNA1-LF2 Pilot Data: Low Fouling Membrane Performance Comparison Mass Transfer Coefficient, MTCw (gfd/psi) 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 Fouling of Stnd NF Feed TOC ~ 20 mg/l Stable Performance of ESNA1-LF vs other NF 0.10 0 5 10 15 20 25 30 35 40 45 50 55 60 Run Time (days) Membrane A 12 ESNA LF
Deerfield Treatment Process Holding Tank Well 5 um Cart Filters 769 gpm 322 gpm Waste 2145 gpm 1375 gpm 447 gpm Product Antiscalant Water 25 C Acid RO System 1823 gpm 482 ppm TDS ESNA1-LF Total 250 ppm Hardness 85% Recovery 10.5 mgd 285 ppm HCO3 5 Trains, 2 Stages 1.5 ppm Fe 48 Vess x 24 Vess x 7 Elem 20 ppm TOC 13 gfd Flux 13
City of Deerfield Beach, FL 14 RO System Design: 10.5 MGD of Permeate ESNA1-LF Membrane 5 trains 48-24 array of 7M tubes 85% recovery 13.0 gfd. Pretreatment: Acid Antiscalant Cartridge filtration. Operation: Start-up November 2003 Steady Performance No cleaning in two years
Deerfield Beach Separation Objective 15
Deerfield Plant Data ESNA1-LF Performance Trend * Estimated from Conductivity 16
3/24/06 12.00 11.00 10.00 9.00 8.00 7.00 6.00 5.00 4.00 3.00 2.00 1.00 0.00 Deerfield Plant Data ESNA1-LF Performance Trend Salt Passage Skid 1 Stage 1 Normalized Data ESNA1-LF Train 1 3/19/04 5/2/04 6/25/04 8/21/04 10/26/04 12/1/04 1/7/05 2/8/05 3/11/05 4/13/05 5/12/05 6/14/05 7/22/05 9/1/05 10/22/05 12/28/05 2/21/06 17 12/03/03 %
3/24/06 Deerfield Plant Data ESNA1-LF Performance Trend 12.00 11.00 10.00 9.00 8.00 7.00 6.00 5.00 4.00 3.00 2.00 1.00 0.00 Salt Passage Skid 1 Stage 2 ESNA1-LF Normalized Data Train 1 12/03/03 3/19/04 5/2/04 6/25/04 8/21/04 10/27/04 12/1/04 1/7/05 2/8/05 3/11/05 4/13/05 5/12/05 6/14/05 7/22/05 9/1/05 10/22/05 12/28/05 2/21/06 Day 18 %
3/24/06 Deerfield Plant Data ESNA1-LF Performance Trend Perm Flow Skid 1 Stage 1 Normalized Data ESNA1-LF Train 1 1400.00 1200.00 1000.00 800.00 600.00 400.00 200.00 0.00 12/03/03 3/19/04 5/2/04 6/25/04 8/21/04 10/26/04 12/1/04 1/7/05 2/8/05 3/11/05 4/13/05 5/12/05 6/14/05 7/22/05 9/1/05 10/22/05 12/28/05 2/21/06 19 gpm
3/24/06 700 650 600 550 500 450 400 350 300 250 200 150 100 50 0 Deerfield Plant Data ESNA1-LF Performance Trend Perm Flow Skid 1 Stage 2 ESNA1-LF Normalized Data Train 1 3/19/04 5/2/04 6/25/04 8/21/04 10/27/04 12/1/04 1/7/05 2/8/05 3/11/05 4/13/05 5/12/05 6/14/05 7/22/05 9/1/05 10/22/05 12/28/05 2/21/06 Day 20 12/03/03 gpm
3/24/06 50.00 45.00 40.00 35.00 30.00 25.00 20.00 15.00 10.00 Deerfield Plant Data ESNA1-LF Performance Trend Differential Pres. Skid 1 Stage 1 Normalized Data ESNA1-LF Train 1 3/19/04 5/2/04 6/25/04 8/21/04 10/26/04 12/1/04 1/7/05 2/8/05 3/11/05 4/13/05 5/12/05 6/14/05 7/22/05 9/1/05 10/22/05 12/28/05 2/21/06 Day 21 12/03/03 psi
Boca Raton Treatment Process Holding Tank Well Media Filter 5 um Cart Filters 1136 gpm 450 gpm Waste LSI = 1.9 3003 gpm 1867 gpm 686 gpm Product Water 25 C RO System 2553 gpm 466 ppm TDS No Antiscalant ESNA1-LF2 Total 265 ppm Hardness NO Acid 85% Recovery 40 mgd 265 ppm HCO3 10 Trains, 2 Stages 0.3 ppm Fe 72 Vess x 36 Vess x 7 Elem 12 ppm TOC 12.2 gfd Flux 22
City of Boca Raton, FL RO System Design ESNA1-LF2 (93% hrdnss rej) 36.8 MGD permeate 10 1 st /2 nd Stage Trains 72-36 array of 7M tubes 85% recovery 12.2 gfd ESNA-LF3 (98% hrdnss rej) 3.2 MGD permeate 2 each 3 rd Stage Trains 36-18 array of 7M tubes 50% recovery 10.7 gfd Pretreatment Multi-media filters Cartridge filters No antiscalant No acid addition 23
Boca Raton Separation Objective 24
Boca Raton Plant Data ESNA1-LF2/3 Performance Trend Salt Passage Boca Raton Unit 7 Stage 1 20 Normalized Salt Passage (%) % 18 16 14 12 10 8 6 4 2 0 9-Nov-04 17-Feb-05 28-May-05 5-Sep-05 14-Dec-05 24-Mar-06 2-Jul-06 Date 25
Boca Raton Plant Data ESNA1-LF2/3 Performance Trend Salt Passage Boca Raton Unit 7 Stage 2 24 Normalized Salt Passage (%) % 21 18 15 12 9 6 3 0 9-Nov-04 17-Feb-05 28-May-05 5-Sep-05 14-Dec-05 24-Mar-06 2-Jul-06 Date 26
Boca Raton Plant Data ESNA1-LF2/3 Performance Trend Perm Flow Norm. Boca Raton Unit 7 Stage 1 Normalized System Flow (gpm) gpm 2600 2400 2200 2000 1800 1600 1400 1200 1000 800 600 400 200 0 9-Nov-04 17-Feb-05 28-May-05 5-Sep-05 14-Dec-05 24-Mar-06 2-Jul-06 Date 27
Boca Raton Plant Data ESNA1-LF2/3 Performance Trend Perm Flow Norm. Boca Raton Unit 7 Stage 2 Normalized System gpm Flow (gpm) 550 500 450 400 350 300 250 200 150 100 50 0 9-Nov-04 17-Feb-05 28-May-05 5-Sep-05 14-Dec-05 24-Mar-06 2-Jul-06 Date 28
Boca Raton Plant Data ESNA1-LF2/3 Performance Trend Differential Pres. Norm. Boca Raton Unit 7 Stage 1 40 Normalized Press Drop (psi) psi 38 36 34 32 30 28 26 24 22 20 18 9-Nov-04 17-Feb-05 28-May-05 5-Sep-05 14-Dec-05 24-Mar-06 2-Jul-06 Date 29
Boca Raton Plant Data ESNA1-LF2/3 Performance Trend Differential Pres. Norm. Boca Raton Unit 7 Stage 2 26 Normalized Press psi Drop (psi) 24 22 20 18 16 14 12 10 9-Nov-04 17-Feb-05 28-May-05 5-Sep-05 14-Dec-05 24-Mar-06 2-Jul-06 Date 30
Analysis of Fouled Membrane from Lead Element EDAX of Membrane Surface covered with foulant. SEM of Membrane Surface covered with foulant. 3000X Ion Ca P O Fe Si C N Atomic % 8.2 5.9 29.8 0.8 0.32 42.7 8.7 Ca3(PO4)2 8.2 5.5 22 31
Boca Raton Plant Data Performance Results 32
Conclusions Novel NF membrane chemistry allows controlled variation of the hardness rejection Modification of the surface results in a smoother, less charged surface which reduces fouling High rejection of NOM materials results in low disinfection by-products and meets Federal standards 33