Development of Advanced Membranes for Produced Water Treatment Introduction • Produced water (PW) is oily water co-produced during oil and gas exploration and production. • Contains dispersed oils and suspended solids. Flux (L/m².h) 80 70 70 Daniel Wandera, Scott M. Husson CHEMICAL & BIOMOLECULAR ENGR. CLEMSON UNIVERSITY Ranil Wickramasinghe CHEMICAL ENGR. UNIVERSITY OF ARKANSAS Variation of Initiator Density 350 80 300 90 Results and Discussion Water Flux Measurements 90 70 Initiator density x 10-19 (g cellulose-1) 80 250 DI Water Flux (L/m².h) 60 50 Flux (L/m².h) • Estimated 21 billion barrels of PW generated from active oil and gas wells in the US in 2007. • Normally, PW is re-injected in geologically isolated wells or is surface discharged with adverse environmental effects. • Membrane filtration is an option but is affected negatively by fouling and concentration polarization. 60 60 200 50 40 150 50 30 40 40 100 20 30 30 50 10 Unmodified 0.5mM 2-BIB Unmodified PNIPAAm PNIPAAm-b-PPEGMA Unmodified-R PNIPAAm-R PNIPAAm-b-PPEGMA-R 50 60 70 80 90 100 0 0 1 2 3 4 5 6 7 0 0 10 20 30 40 0.25mM 2-BIB 1.0mM 2-BIB 50 60 70 20 20 Unmodified 10 Unmodified-R PNIPAAm - R PNIPAAm-b-PPEGMA - R 50 60 70 80 90 100 10 PNIPAAm PNIPAAm-b-PPEGMA Initiator Concentration (mM) Time (min) Objectives • Graft poly(N-isopropylacrylamide)-block-poly(oligoethylene glycol methacrylate) by surface-initiated atom transfer radical polymerization (ATRP) from ultrafiltration membranes to limit foulant accumulation and provide an easy, chemical-free way to remove any attached foulants. • Characterize the physiochemical and performance properties of modified membranes. 0 0 10 20 30 40 0 0 10 20 30 40 Time (min) Time (min) Systematically varied initiator density by changing initiator (2-BIB) conc. Direct-flow filtration with model synthetic PW (Left) and actual oilfield PW (Right) at 30 psi. Effect of varying initiator density on DI water flux, 5 kDa membranes, 30 psi. Higher Initiator density leads to lower flux. 90 Characteristics of model synthetic produced water Parameter Dispersed oil Conductivity Total Dissolved Solids (TDS) Total Organic Carbon (TOC) Value 5000 mg/L DI Water Flux (L/m².h) Variation of Polymerization Time 90 80 80 70 70 DI Water Flux (L/m².h) 1108 µs/cm 543 mg/L 82.23 mg/L 60 60 50 50 40 40 Permeate Quality Specifications for 5 kDa membranes Membrane Unmodified PNIPAAm PNIPAAm-b-PPEGMA Conductivity (µS/cm) 1038 1024 1021 TDS (mg/L) 508 502 500 7000 30 30 Experimental Methods Surface-initiated ATRP • Potential for independent control over grafting density and chain length at low polydispersity. • Straightforward to graft block copolymers. TOC (mg/L) 4.91 3.30 2.02 TOC Removal (%) 94.02 95.99 97.54 20 20 Unmodified Unmodified 1.0h PNIPAAm 0.5h PNIPAAm 2.0h PNIPAAm 0 0 0 10 20 30 40 50 60 70 10 20 30 40 50 10 1.5h PPEGMA 6.0h PPEGMA 60 70 10 3.0h PPEGMA 0 Time (min) Time (min) Modified Membrane 6000 Effect of PNIPAAm polymerization time on DI water flux, 5 kDa membranes, 30 psi. Longer time leads to lower flux. Effect of PPEGMA polymerization time on DI water flux, 5 kDa membranes, 30 psi. Longer time leads to lower flux. Unmodified Membrane 5000 Conclusions • Developed membranes for PW treatment with good anti-fouling properties that can be cleaned well by water rinse • Poor salt rejection but excellent TOC removal in all membranes • General decline in flux post modification; flux recovery better for modified membranes 0 10 20 30 40 50 60 Flux (L/m².h) 4000 Membrane Modification Regenerated Cellulose (RC) membranes Model PW (A), Unmodified 5 kDa membrane permeate (B) and Modified 5 kDa membrane permeate (C). 120 3000 2000 Unmodified 1mM 2-BIB Modified 100 1000 2mM 2-BIB Modified 0 80 Time (h) Flux (L/m².h) 60 Characterization • ATR FTIR • AFM • Water Flux Measurements Cross-flow filtration results for 1000 kDa unmodified and modified membranes using model PW at 60 psi and 50 C (Top) Cross-flow filtration results for 5 kDa unmodified and modified membranes using model PW at 40 psi and 50 C (Left). 0 10 20 30 40 50 60 70 80 90 • Modified membranes allow higher cumulative volumes of water to be processed over time prior to cleaning • Initiator density and polymerization time impact water flux, so can be used to tune membrane performance 40 20 Current Work • Developing an effective self-cleaning protocol 0 Time (h) Acknowledgements: This work was supported by the National Science Foundation (CBET 0651646). Thanks to Heath Himstedt, Colorado State University. Reference: D. Wandera et al., J. Membr. Sci. 373 (2011) 178–188